1
USING THE IJG JPEG LIBRARY
3
Copyright (C) 1994-1998, Thomas G. Lane.
4
This file is part of the Independent JPEG Group's software.
5
For conditions of distribution and use, see the accompanying README file.
8
This file describes how to use the IJG JPEG library within an application
9
program. Read it if you want to write a program that uses the library.
11
The file example.c provides heavily commented skeleton code for calling the
12
JPEG library. Also see jpeglib.h (the include file to be used by application
13
programs) for full details about data structures and function parameter lists.
14
The library source code, of course, is the ultimate reference.
16
Note that there have been *major* changes from the application interface
17
presented by IJG version 4 and earlier versions. The old design had several
18
inherent limitations, and it had accumulated a lot of cruft as we added
19
features while trying to minimize application-interface changes. We have
20
sacrificed backward compatibility in the version 5 rewrite, but we think the
21
improvements justify this.
28
Functions provided by the library
29
Outline of typical usage
34
Mechanics of usage: include files, linking, etc
36
Compression parameter selection
37
Decompression parameter selection
40
Compressed data handling (source and destination managers)
42
Progressive JPEG support
44
Abbreviated datastreams and multiple images
46
Raw (downsampled) image data
47
Really raw data: DCT coefficients
51
Library compile-time options
52
Portability considerations
53
Notes for MS-DOS implementors
55
You should read at least the overview and basic usage sections before trying
56
to program with the library. The sections on advanced features can be read
57
if and when you need them.
63
Functions provided by the library
64
---------------------------------
66
The IJG JPEG library provides C code to read and write JPEG-compressed image
67
files. The surrounding application program receives or supplies image data a
68
scanline at a time, using a straightforward uncompressed image format. All
69
details of color conversion and other preprocessing/postprocessing can be
70
handled by the library.
72
The library includes a substantial amount of code that is not covered by the
73
JPEG standard but is necessary for typical applications of JPEG. These
74
functions preprocess the image before JPEG compression or postprocess it after
75
decompression. They include colorspace conversion, downsampling/upsampling,
76
and color quantization. The application indirectly selects use of this code
77
by specifying the format in which it wishes to supply or receive image data.
78
For example, if colormapped output is requested, then the decompression
79
library automatically invokes color quantization.
81
A wide range of quality vs. speed tradeoffs are possible in JPEG processing,
82
and even more so in decompression postprocessing. The decompression library
83
provides multiple implementations that cover most of the useful tradeoffs,
84
ranging from very-high-quality down to fast-preview operation. On the
85
compression side we have generally not provided low-quality choices, since
86
compression is normally less time-critical. It should be understood that the
87
low-quality modes may not meet the JPEG standard's accuracy requirements;
88
nonetheless, they are useful for viewers.
90
A word about functions *not* provided by the library. We handle a subset of
91
the ISO JPEG standard; most baseline, extended-sequential, and progressive
92
JPEG processes are supported. (Our subset includes all features now in common
93
use.) Unsupported ISO options include:
94
* Hierarchical storage
96
* Arithmetic entropy coding (unsupported for legal reasons)
98
* Nonintegral subsampling ratios
99
We support both 8- and 12-bit data precision, but this is a compile-time
100
choice rather than a run-time choice; hence it is difficult to use both
101
precisions in a single application.
103
By itself, the library handles only interchange JPEG datastreams --- in
104
particular the widely used JFIF file format. The library can be used by
105
surrounding code to process interchange or abbreviated JPEG datastreams that
106
are embedded in more complex file formats. (For example, this library is
107
used by the free LIBTIFF library to support JPEG compression in TIFF.)
110
Outline of typical usage
111
------------------------
113
The rough outline of a JPEG compression operation is:
115
Allocate and initialize a JPEG compression object
116
Specify the destination for the compressed data (eg, a file)
117
Set parameters for compression, including image size & colorspace
118
jpeg_start_compress(...);
119
while (scan lines remain to be written)
120
jpeg_write_scanlines(...);
121
jpeg_finish_compress(...);
122
Release the JPEG compression object
124
A JPEG compression object holds parameters and working state for the JPEG
125
library. We make creation/destruction of the object separate from starting
126
or finishing compression of an image; the same object can be re-used for a
127
series of image compression operations. This makes it easy to re-use the
128
same parameter settings for a sequence of images. Re-use of a JPEG object
129
also has important implications for processing abbreviated JPEG datastreams,
132
The image data to be compressed is supplied to jpeg_write_scanlines() from
133
in-memory buffers. If the application is doing file-to-file compression,
134
reading image data from the source file is the application's responsibility.
135
The library emits compressed data by calling a "data destination manager",
136
which typically will write the data into a file; but the application can
137
provide its own destination manager to do something else.
139
Similarly, the rough outline of a JPEG decompression operation is:
141
Allocate and initialize a JPEG decompression object
142
Specify the source of the compressed data (eg, a file)
143
Call jpeg_read_header() to obtain image info
144
Set parameters for decompression
145
jpeg_start_decompress(...);
146
while (scan lines remain to be read)
147
jpeg_read_scanlines(...);
148
jpeg_finish_decompress(...);
149
Release the JPEG decompression object
151
This is comparable to the compression outline except that reading the
152
datastream header is a separate step. This is helpful because information
153
about the image's size, colorspace, etc is available when the application
154
selects decompression parameters. For example, the application can choose an
155
output scaling ratio that will fit the image into the available screen size.
157
The decompression library obtains compressed data by calling a data source
158
manager, which typically will read the data from a file; but other behaviors
159
can be obtained with a custom source manager. Decompressed data is delivered
160
into in-memory buffers passed to jpeg_read_scanlines().
162
It is possible to abort an incomplete compression or decompression operation
163
by calling jpeg_abort(); or, if you do not need to retain the JPEG object,
164
simply release it by calling jpeg_destroy().
166
JPEG compression and decompression objects are two separate struct types.
167
However, they share some common fields, and certain routines such as
168
jpeg_destroy() can work on either type of object.
170
The JPEG library has no static variables: all state is in the compression
171
or decompression object. Therefore it is possible to process multiple
172
compression and decompression operations concurrently, using multiple JPEG
175
Both compression and decompression can be done in an incremental memory-to-
176
memory fashion, if suitable source/destination managers are used. See the
177
section on "I/O suspension" for more details.
186
Before diving into procedural details, it is helpful to understand the
187
image data format that the JPEG library expects or returns.
189
The standard input image format is a rectangular array of pixels, with each
190
pixel having the same number of "component" or "sample" values (color
191
channels). You must specify how many components there are and the colorspace
192
interpretation of the components. Most applications will use RGB data
193
(three components per pixel) or grayscale data (one component per pixel).
194
PLEASE NOTE THAT RGB DATA IS THREE SAMPLES PER PIXEL, GRAYSCALE ONLY ONE.
195
A remarkable number of people manage to miss this, only to find that their
196
programs don't work with grayscale JPEG files.
198
There is no provision for colormapped input. JPEG files are always full-color
199
or full grayscale (or sometimes another colorspace such as CMYK). You can
200
feed in a colormapped image by expanding it to full-color format. However
201
JPEG often doesn't work very well with source data that has been colormapped,
202
because of dithering noise. This is discussed in more detail in the JPEG FAQ
203
and the other references mentioned in the README file.
205
Pixels are stored by scanlines, with each scanline running from left to
206
right. The component values for each pixel are adjacent in the row; for
207
example, R,G,B,R,G,B,R,G,B,... for 24-bit RGB color. Each scanline is an
208
array of data type JSAMPLE --- which is typically "unsigned char", unless
209
you've changed jmorecfg.h. (You can also change the RGB pixel layout, say
210
to B,G,R order, by modifying jmorecfg.h. But see the restrictions listed in
211
that file before doing so.)
213
A 2-D array of pixels is formed by making a list of pointers to the starts of
214
scanlines; so the scanlines need not be physically adjacent in memory. Even
215
if you process just one scanline at a time, you must make a one-element
216
pointer array to conform to this structure. Pointers to JSAMPLE rows are of
217
type JSAMPROW, and the pointer to the pointer array is of type JSAMPARRAY.
219
The library accepts or supplies one or more complete scanlines per call.
220
It is not possible to process part of a row at a time. Scanlines are always
221
processed top-to-bottom. You can process an entire image in one call if you
222
have it all in memory, but usually it's simplest to process one scanline at
225
For best results, source data values should have the precision specified by
226
BITS_IN_JSAMPLE (normally 8 bits). For instance, if you choose to compress
227
data that's only 6 bits/channel, you should left-justify each value in a
228
byte before passing it to the compressor. If you need to compress data
229
that has more than 8 bits/channel, compile with BITS_IN_JSAMPLE = 12.
230
(See "Library compile-time options", later.)
233
The data format returned by the decompressor is the same in all details,
234
except that colormapped output is supported. (Again, a JPEG file is never
235
colormapped. But you can ask the decompressor to perform on-the-fly color
236
quantization to deliver colormapped output.) If you request colormapped
237
output then the returned data array contains a single JSAMPLE per pixel;
238
its value is an index into a color map. The color map is represented as
239
a 2-D JSAMPARRAY in which each row holds the values of one color component,
240
that is, colormap[i][j] is the value of the i'th color component for pixel
241
value (map index) j. Note that since the colormap indexes are stored in
242
JSAMPLEs, the maximum number of colors is limited by the size of JSAMPLE
243
(ie, at most 256 colors for an 8-bit JPEG library).
249
Here we revisit the JPEG compression outline given in the overview.
251
1. Allocate and initialize a JPEG compression object.
253
A JPEG compression object is a "struct jpeg_compress_struct". (It also has
254
a bunch of subsidiary structures which are allocated via malloc(), but the
255
application doesn't control those directly.) This struct can be just a local
256
variable in the calling routine, if a single routine is going to execute the
257
whole JPEG compression sequence. Otherwise it can be static or allocated
260
You will also need a structure representing a JPEG error handler. The part
261
of this that the library cares about is a "struct jpeg_error_mgr". If you
262
are providing your own error handler, you'll typically want to embed the
263
jpeg_error_mgr struct in a larger structure; this is discussed later under
264
"Error handling". For now we'll assume you are just using the default error
265
handler. The default error handler will print JPEG error/warning messages
266
on stderr, and it will call exit() if a fatal error occurs.
268
You must initialize the error handler structure, store a pointer to it into
269
the JPEG object's "err" field, and then call jpeg_create_compress() to
270
initialize the rest of the JPEG object.
272
Typical code for this step, if you are using the default error handler, is
274
struct jpeg_compress_struct cinfo;
275
struct jpeg_error_mgr jerr;
277
cinfo.err = jpeg_std_error(&jerr);
278
jpeg_create_compress(&cinfo);
280
jpeg_create_compress allocates a small amount of memory, so it could fail
281
if you are out of memory. In that case it will exit via the error handler;
282
that's why the error handler must be initialized first.
285
2. Specify the destination for the compressed data (eg, a file).
287
As previously mentioned, the JPEG library delivers compressed data to a
288
"data destination" module. The library includes one data destination
289
module which knows how to write to a stdio stream. You can use your own
290
destination module if you want to do something else, as discussed later.
292
If you use the standard destination module, you must open the target stdio
293
stream beforehand. Typical code for this step looks like:
297
if ((outfile = fopen(filename, "wb")) == NULL) {
298
fprintf(stderr, "can't open %s\n", filename);
301
jpeg_stdio_dest(&cinfo, outfile);
303
where the last line invokes the standard destination module.
305
WARNING: it is critical that the binary compressed data be delivered to the
306
output file unchanged. On non-Unix systems the stdio library may perform
307
newline translation or otherwise corrupt binary data. To suppress this
308
behavior, you may need to use a "b" option to fopen (as shown above), or use
309
setmode() or another routine to put the stdio stream in binary mode. See
310
cjpeg.c and djpeg.c for code that has been found to work on many systems.
312
You can select the data destination after setting other parameters (step 3),
313
if that's more convenient. You may not change the destination between
314
calling jpeg_start_compress() and jpeg_finish_compress().
317
3. Set parameters for compression, including image size & colorspace.
319
You must supply information about the source image by setting the following
320
fields in the JPEG object (cinfo structure):
322
image_width Width of image, in pixels
323
image_height Height of image, in pixels
324
input_components Number of color channels (samples per pixel)
325
in_color_space Color space of source image
327
The image dimensions are, hopefully, obvious. JPEG supports image dimensions
328
of 1 to 64K pixels in either direction. The input color space is typically
329
RGB or grayscale, and input_components is 3 or 1 accordingly. (See "Special
330
color spaces", later, for more info.) The in_color_space field must be
331
assigned one of the J_COLOR_SPACE enum constants, typically JCS_RGB or
334
JPEG has a large number of compression parameters that determine how the
335
image is encoded. Most applications don't need or want to know about all
336
these parameters. You can set all the parameters to reasonable defaults by
337
calling jpeg_set_defaults(); then, if there are particular values you want
338
to change, you can do so after that. The "Compression parameter selection"
339
section tells about all the parameters.
341
You must set in_color_space correctly before calling jpeg_set_defaults(),
342
because the defaults depend on the source image colorspace. However the
343
other three source image parameters need not be valid until you call
344
jpeg_start_compress(). There's no harm in calling jpeg_set_defaults() more
345
than once, if that happens to be convenient.
347
Typical code for a 24-bit RGB source image is
349
cinfo.image_width = Width; /* image width and height, in pixels */
350
cinfo.image_height = Height;
351
cinfo.input_components = 3; /* # of color components per pixel */
352
cinfo.in_color_space = JCS_RGB; /* colorspace of input image */
354
jpeg_set_defaults(&cinfo);
355
/* Make optional parameter settings here */
358
4. jpeg_start_compress(...);
360
After you have established the data destination and set all the necessary
361
source image info and other parameters, call jpeg_start_compress() to begin
362
a compression cycle. This will initialize internal state, allocate working
363
storage, and emit the first few bytes of the JPEG datastream header.
367
jpeg_start_compress(&cinfo, TRUE);
369
The "TRUE" parameter ensures that a complete JPEG interchange datastream
370
will be written. This is appropriate in most cases. If you think you might
371
want to use an abbreviated datastream, read the section on abbreviated
374
Once you have called jpeg_start_compress(), you may not alter any JPEG
375
parameters or other fields of the JPEG object until you have completed
376
the compression cycle.
379
5. while (scan lines remain to be written)
380
jpeg_write_scanlines(...);
382
Now write all the required image data by calling jpeg_write_scanlines()
383
one or more times. You can pass one or more scanlines in each call, up
384
to the total image height. In most applications it is convenient to pass
385
just one or a few scanlines at a time. The expected format for the passed
386
data is discussed under "Data formats", above.
388
Image data should be written in top-to-bottom scanline order. The JPEG spec
389
contains some weasel wording about how top and bottom are application-defined
390
terms (a curious interpretation of the English language...) but if you want
391
your files to be compatible with everyone else's, you WILL use top-to-bottom
392
order. If the source data must be read in bottom-to-top order, you can use
393
the JPEG library's virtual array mechanism to invert the data efficiently.
394
Examples of this can be found in the sample application cjpeg.
396
The library maintains a count of the number of scanlines written so far
397
in the next_scanline field of the JPEG object. Usually you can just use
398
this variable as the loop counter, so that the loop test looks like
399
"while (cinfo.next_scanline < cinfo.image_height)".
401
Code for this step depends heavily on the way that you store the source data.
402
example.c shows the following code for the case of a full-size 2-D source
403
array containing 3-byte RGB pixels:
405
JSAMPROW row_pointer[1]; /* pointer to a single row */
406
int row_stride; /* physical row width in buffer */
408
row_stride = image_width * 3; /* JSAMPLEs per row in image_buffer */
410
while (cinfo.next_scanline < cinfo.image_height) {
411
row_pointer[0] = & image_buffer[cinfo.next_scanline * row_stride];
412
jpeg_write_scanlines(&cinfo, row_pointer, 1);
415
jpeg_write_scanlines() returns the number of scanlines actually written.
416
This will normally be equal to the number passed in, so you can usually
417
ignore the return value. It is different in just two cases:
418
* If you try to write more scanlines than the declared image height,
419
the additional scanlines are ignored.
420
* If you use a suspending data destination manager, output buffer overrun
421
will cause the compressor to return before accepting all the passed lines.
422
This feature is discussed under "I/O suspension", below. The normal
423
stdio destination manager will NOT cause this to happen.
424
In any case, the return value is the same as the change in the value of
428
6. jpeg_finish_compress(...);
430
After all the image data has been written, call jpeg_finish_compress() to
431
complete the compression cycle. This step is ESSENTIAL to ensure that the
432
last bufferload of data is written to the data destination.
433
jpeg_finish_compress() also releases working memory associated with the JPEG
438
jpeg_finish_compress(&cinfo);
440
If using the stdio destination manager, don't forget to close the output
441
stdio stream (if necessary) afterwards.
443
If you have requested a multi-pass operating mode, such as Huffman code
444
optimization, jpeg_finish_compress() will perform the additional passes using
445
data buffered by the first pass. In this case jpeg_finish_compress() may take
446
quite a while to complete. With the default compression parameters, this will
449
It is an error to call jpeg_finish_compress() before writing the necessary
450
total number of scanlines. If you wish to abort compression, call
451
jpeg_abort() as discussed below.
453
After completing a compression cycle, you may dispose of the JPEG object
454
as discussed next, or you may use it to compress another image. In that case
455
return to step 2, 3, or 4 as appropriate. If you do not change the
456
destination manager, the new datastream will be written to the same target.
457
If you do not change any JPEG parameters, the new datastream will be written
458
with the same parameters as before. Note that you can change the input image
459
dimensions freely between cycles, but if you change the input colorspace, you
460
should call jpeg_set_defaults() to adjust for the new colorspace; and then
461
you'll need to repeat all of step 3.
464
7. Release the JPEG compression object.
466
When you are done with a JPEG compression object, destroy it by calling
467
jpeg_destroy_compress(). This will free all subsidiary memory (regardless of
468
the previous state of the object). Or you can call jpeg_destroy(), which
469
works for either compression or decompression objects --- this may be more
470
convenient if you are sharing code between compression and decompression
471
cases. (Actually, these routines are equivalent except for the declared type
472
of the passed pointer. To avoid gripes from ANSI C compilers, jpeg_destroy()
473
should be passed a j_common_ptr.)
475
If you allocated the jpeg_compress_struct structure from malloc(), freeing
476
it is your responsibility --- jpeg_destroy() won't. Ditto for the error
481
jpeg_destroy_compress(&cinfo);
486
If you decide to abort a compression cycle before finishing, you can clean up
487
in either of two ways:
489
* If you don't need the JPEG object any more, just call
490
jpeg_destroy_compress() or jpeg_destroy() to release memory. This is
491
legitimate at any point after calling jpeg_create_compress() --- in fact,
492
it's safe even if jpeg_create_compress() fails.
494
* If you want to re-use the JPEG object, call jpeg_abort_compress(), or call
495
jpeg_abort() which works on both compression and decompression objects.
496
This will return the object to an idle state, releasing any working memory.
497
jpeg_abort() is allowed at any time after successful object creation.
499
Note that cleaning up the data destination, if required, is your
500
responsibility; neither of these routines will call term_destination().
501
(See "Compressed data handling", below, for more about that.)
503
jpeg_destroy() and jpeg_abort() are the only safe calls to make on a JPEG
504
object that has reported an error by calling error_exit (see "Error handling"
505
for more info). The internal state of such an object is likely to be out of
506
whack. Either of these two routines will return the object to a known state.
509
Decompression details
510
---------------------
512
Here we revisit the JPEG decompression outline given in the overview.
514
1. Allocate and initialize a JPEG decompression object.
516
This is just like initialization for compression, as discussed above,
517
except that the object is a "struct jpeg_decompress_struct" and you
518
call jpeg_create_decompress(). Error handling is exactly the same.
522
struct jpeg_decompress_struct cinfo;
523
struct jpeg_error_mgr jerr;
525
cinfo.err = jpeg_std_error(&jerr);
526
jpeg_create_decompress(&cinfo);
528
(Both here and in the IJG code, we usually use variable name "cinfo" for
529
both compression and decompression objects.)
532
2. Specify the source of the compressed data (eg, a file).
534
As previously mentioned, the JPEG library reads compressed data from a "data
535
source" module. The library includes one data source module which knows how
536
to read from a stdio stream. You can use your own source module if you want
537
to do something else, as discussed later.
539
If you use the standard source module, you must open the source stdio stream
540
beforehand. Typical code for this step looks like:
544
if ((infile = fopen(filename, "rb")) == NULL) {
545
fprintf(stderr, "can't open %s\n", filename);
548
jpeg_stdio_src(&cinfo, infile);
550
where the last line invokes the standard source module.
552
WARNING: it is critical that the binary compressed data be read unchanged.
553
On non-Unix systems the stdio library may perform newline translation or
554
otherwise corrupt binary data. To suppress this behavior, you may need to use
555
a "b" option to fopen (as shown above), or use setmode() or another routine to
556
put the stdio stream in binary mode. See cjpeg.c and djpeg.c for code that
557
has been found to work on many systems.
559
You may not change the data source between calling jpeg_read_header() and
560
jpeg_finish_decompress(). If you wish to read a series of JPEG images from
561
a single source file, you should repeat the jpeg_read_header() to
562
jpeg_finish_decompress() sequence without reinitializing either the JPEG
563
object or the data source module; this prevents buffered input data from
567
3. Call jpeg_read_header() to obtain image info.
569
Typical code for this step is just
571
jpeg_read_header(&cinfo, TRUE);
573
This will read the source datastream header markers, up to the beginning
574
of the compressed data proper. On return, the image dimensions and other
575
info have been stored in the JPEG object. The application may wish to
576
consult this information before selecting decompression parameters.
578
More complex code is necessary if
579
* A suspending data source is used --- in that case jpeg_read_header()
580
may return before it has read all the header data. See "I/O suspension",
581
below. The normal stdio source manager will NOT cause this to happen.
582
* Abbreviated JPEG files are to be processed --- see the section on
583
abbreviated datastreams. Standard applications that deal only in
584
interchange JPEG files need not be concerned with this case either.
586
It is permissible to stop at this point if you just wanted to find out the
587
image dimensions and other header info for a JPEG file. In that case,
588
call jpeg_destroy() when you are done with the JPEG object, or call
589
jpeg_abort() to return it to an idle state before selecting a new data
590
source and reading another header.
593
4. Set parameters for decompression.
595
jpeg_read_header() sets appropriate default decompression parameters based on
596
the properties of the image (in particular, its colorspace). However, you
597
may well want to alter these defaults before beginning the decompression.
598
For example, the default is to produce full color output from a color file.
599
If you want colormapped output you must ask for it. Other options allow the
600
returned image to be scaled and allow various speed/quality tradeoffs to be
601
selected. "Decompression parameter selection", below, gives details.
603
If the defaults are appropriate, nothing need be done at this step.
605
Note that all default values are set by each call to jpeg_read_header().
606
If you reuse a decompression object, you cannot expect your parameter
607
settings to be preserved across cycles, as you can for compression.
608
You must set desired parameter values each time.
611
5. jpeg_start_decompress(...);
613
Once the parameter values are satisfactory, call jpeg_start_decompress() to
614
begin decompression. This will initialize internal state, allocate working
615
memory, and prepare for returning data.
619
jpeg_start_decompress(&cinfo);
621
If you have requested a multi-pass operating mode, such as 2-pass color
622
quantization, jpeg_start_decompress() will do everything needed before data
623
output can begin. In this case jpeg_start_decompress() may take quite a while
624
to complete. With a single-scan (non progressive) JPEG file and default
625
decompression parameters, this will not happen; jpeg_start_decompress() will
628
After this call, the final output image dimensions, including any requested
629
scaling, are available in the JPEG object; so is the selected colormap, if
630
colormapped output has been requested. Useful fields include
632
output_width image width and height, as scaled
634
out_color_components # of color components in out_color_space
635
output_components # of color components returned per pixel
636
colormap the selected colormap, if any
637
actual_number_of_colors number of entries in colormap
639
output_components is 1 (a colormap index) when quantizing colors; otherwise it
640
equals out_color_components. It is the number of JSAMPLE values that will be
641
emitted per pixel in the output arrays.
643
Typically you will need to allocate data buffers to hold the incoming image.
644
You will need output_width * output_components JSAMPLEs per scanline in your
645
output buffer, and a total of output_height scanlines will be returned.
647
Note: if you are using the JPEG library's internal memory manager to allocate
648
data buffers (as djpeg does), then the manager's protocol requires that you
649
request large buffers *before* calling jpeg_start_decompress(). This is a
650
little tricky since the output_XXX fields are not normally valid then. You
651
can make them valid by calling jpeg_calc_output_dimensions() after setting the
652
relevant parameters (scaling, output color space, and quantization flag).
655
6. while (scan lines remain to be read)
656
jpeg_read_scanlines(...);
658
Now you can read the decompressed image data by calling jpeg_read_scanlines()
659
one or more times. At each call, you pass in the maximum number of scanlines
660
to be read (ie, the height of your working buffer); jpeg_read_scanlines()
661
will return up to that many lines. The return value is the number of lines
662
actually read. The format of the returned data is discussed under "Data
663
formats", above. Don't forget that grayscale and color JPEGs will return
664
different data formats!
666
Image data is returned in top-to-bottom scanline order. If you must write
667
out the image in bottom-to-top order, you can use the JPEG library's virtual
668
array mechanism to invert the data efficiently. Examples of this can be
669
found in the sample application djpeg.
671
The library maintains a count of the number of scanlines returned so far
672
in the output_scanline field of the JPEG object. Usually you can just use
673
this variable as the loop counter, so that the loop test looks like
674
"while (cinfo.output_scanline < cinfo.output_height)". (Note that the test
675
should NOT be against image_height, unless you never use scaling. The
676
image_height field is the height of the original unscaled image.)
677
The return value always equals the change in the value of output_scanline.
679
If you don't use a suspending data source, it is safe to assume that
680
jpeg_read_scanlines() reads at least one scanline per call, until the
681
bottom of the image has been reached.
683
If you use a buffer larger than one scanline, it is NOT safe to assume that
684
jpeg_read_scanlines() fills it. (The current implementation returns only a
685
few scanlines per call, no matter how large a buffer you pass.) So you must
686
always provide a loop that calls jpeg_read_scanlines() repeatedly until the
687
whole image has been read.
690
7. jpeg_finish_decompress(...);
692
After all the image data has been read, call jpeg_finish_decompress() to
693
complete the decompression cycle. This causes working memory associated
694
with the JPEG object to be released.
698
jpeg_finish_decompress(&cinfo);
700
If using the stdio source manager, don't forget to close the source stdio
703
It is an error to call jpeg_finish_decompress() before reading the correct
704
total number of scanlines. If you wish to abort decompression, call
705
jpeg_abort() as discussed below.
707
After completing a decompression cycle, you may dispose of the JPEG object as
708
discussed next, or you may use it to decompress another image. In that case
709
return to step 2 or 3 as appropriate. If you do not change the source
710
manager, the next image will be read from the same source.
713
8. Release the JPEG decompression object.
715
When you are done with a JPEG decompression object, destroy it by calling
716
jpeg_destroy_decompress() or jpeg_destroy(). The previous discussion of
717
destroying compression objects applies here too.
721
jpeg_destroy_decompress(&cinfo);
726
You can abort a decompression cycle by calling jpeg_destroy_decompress() or
727
jpeg_destroy() if you don't need the JPEG object any more, or
728
jpeg_abort_decompress() or jpeg_abort() if you want to reuse the object.
729
The previous discussion of aborting compression cycles applies here too.
732
Mechanics of usage: include files, linking, etc
733
-----------------------------------------------
735
Applications using the JPEG library should include the header file jpeglib.h
736
to obtain declarations of data types and routines. Before including
737
jpeglib.h, include system headers that define at least the typedefs FILE and
738
size_t. On ANSI-conforming systems, including <stdio.h> is sufficient; on
739
older Unix systems, you may need <sys/types.h> to define size_t.
741
If the application needs to refer to individual JPEG library error codes, also
742
include jerror.h to define those symbols.
744
jpeglib.h indirectly includes the files jconfig.h and jmorecfg.h. If you are
745
installing the JPEG header files in a system directory, you will want to
746
install all four files: jpeglib.h, jerror.h, jconfig.h, jmorecfg.h.
748
The most convenient way to include the JPEG code into your executable program
749
is to prepare a library file ("libjpeg.a", or a corresponding name on non-Unix
750
machines) and reference it at your link step. If you use only half of the
751
library (only compression or only decompression), only that much code will be
752
included from the library, unless your linker is hopelessly brain-damaged.
753
The supplied makefiles build libjpeg.a automatically (see install.doc).
755
While you can build the JPEG library as a shared library if the whim strikes
756
you, we don't really recommend it. The trouble with shared libraries is that
757
at some point you'll probably try to substitute a new version of the library
758
without recompiling the calling applications. That generally doesn't work
759
because the parameter struct declarations usually change with each new
760
version. In other words, the library's API is *not* guaranteed binary
761
compatible across versions; we only try to ensure source-code compatibility.
762
(In hindsight, it might have been smarter to hide the parameter structs from
763
applications and introduce a ton of access functions instead. Too late now,
766
On some systems your application may need to set up a signal handler to ensure
767
that temporary files are deleted if the program is interrupted. This is most
768
critical if you are on MS-DOS and use the jmemdos.c memory manager back end;
769
it will try to grab extended memory for temp files, and that space will NOT be
770
freed automatically. See cjpeg.c or djpeg.c for an example signal handler.
772
It may be worth pointing out that the core JPEG library does not actually
773
require the stdio library: only the default source/destination managers and
774
error handler need it. You can use the library in a stdio-less environment
775
if you replace those modules and use jmemnobs.c (or another memory manager of
776
your own devising). More info about the minimum system library requirements
777
may be found in jinclude.h.
783
Compression parameter selection
784
-------------------------------
786
This section describes all the optional parameters you can set for JPEG
787
compression, as well as the "helper" routines provided to assist in this
788
task. Proper setting of some parameters requires detailed understanding
789
of the JPEG standard; if you don't know what a parameter is for, it's best
790
not to mess with it! See REFERENCES in the README file for pointers to
791
more info about JPEG.
793
It's a good idea to call jpeg_set_defaults() first, even if you plan to set
794
all the parameters; that way your code is more likely to work with future JPEG
795
libraries that have additional parameters. For the same reason, we recommend
796
you use a helper routine where one is provided, in preference to twiddling
797
cinfo fields directly.
799
The helper routines are:
801
jpeg_set_defaults (j_compress_ptr cinfo)
802
This routine sets all JPEG parameters to reasonable defaults, using
803
only the input image's color space (field in_color_space, which must
804
already be set in cinfo). Many applications will only need to use
805
this routine and perhaps jpeg_set_quality().
807
jpeg_set_colorspace (j_compress_ptr cinfo, J_COLOR_SPACE colorspace)
808
Sets the JPEG file's colorspace (field jpeg_color_space) as specified,
809
and sets other color-space-dependent parameters appropriately. See
810
"Special color spaces", below, before using this. A large number of
811
parameters, including all per-component parameters, are set by this
812
routine; if you want to twiddle individual parameters you should call
813
jpeg_set_colorspace() before rather than after.
815
jpeg_default_colorspace (j_compress_ptr cinfo)
816
Selects an appropriate JPEG colorspace based on cinfo->in_color_space,
817
and calls jpeg_set_colorspace(). This is actually a subroutine of
818
jpeg_set_defaults(). It's broken out in case you want to change
819
just the colorspace-dependent JPEG parameters.
821
jpeg_set_quality (j_compress_ptr cinfo, int quality, boolean force_baseline)
822
Constructs JPEG quantization tables appropriate for the indicated
823
quality setting. The quality value is expressed on the 0..100 scale
824
recommended by IJG (cjpeg's "-quality" switch uses this routine).
825
Note that the exact mapping from quality values to tables may change
826
in future IJG releases as more is learned about DCT quantization.
827
If the force_baseline parameter is TRUE, then the quantization table
828
entries are constrained to the range 1..255 for full JPEG baseline
829
compatibility. In the current implementation, this only makes a
830
difference for quality settings below 25, and it effectively prevents
831
very small/low quality files from being generated. The IJG decoder
832
is capable of reading the non-baseline files generated at low quality
833
settings when force_baseline is FALSE, but other decoders may not be.
835
jpeg_set_linear_quality (j_compress_ptr cinfo, int scale_factor,
836
boolean force_baseline)
837
Same as jpeg_set_quality() except that the generated tables are the
838
sample tables given in the JPEC spec section K.1, multiplied by the
839
specified scale factor (which is expressed as a percentage; thus
840
scale_factor = 100 reproduces the spec's tables). Note that larger
841
scale factors give lower quality. This entry point is useful for
842
conforming to the Adobe PostScript DCT conventions, but we do not
843
recommend linear scaling as a user-visible quality scale otherwise.
844
force_baseline again constrains the computed table entries to 1..255.
846
int jpeg_quality_scaling (int quality)
847
Converts a value on the IJG-recommended quality scale to a linear
848
scaling percentage. Note that this routine may change or go away
849
in future releases --- IJG may choose to adopt a scaling method that
850
can't be expressed as a simple scalar multiplier, in which case the
851
premise of this routine collapses. Caveat user.
853
jpeg_add_quant_table (j_compress_ptr cinfo, int which_tbl,
854
const unsigned int *basic_table,
855
int scale_factor, boolean force_baseline)
856
Allows an arbitrary quantization table to be created. which_tbl
857
indicates which table slot to fill. basic_table points to an array
858
of 64 unsigned ints given in normal array order. These values are
859
multiplied by scale_factor/100 and then clamped to the range 1..65535
860
(or to 1..255 if force_baseline is TRUE).
861
CAUTION: prior to library version 6a, jpeg_add_quant_table expected
862
the basic table to be given in JPEG zigzag order. If you need to
863
write code that works with either older or newer versions of this
864
routine, you must check the library version number. Something like
865
"#if JPEG_LIB_VERSION >= 61" is the right test.
867
jpeg_simple_progression (j_compress_ptr cinfo)
868
Generates a default scan script for writing a progressive-JPEG file.
869
This is the recommended method of creating a progressive file,
870
unless you want to make a custom scan sequence. You must ensure that
871
the JPEG color space is set correctly before calling this routine.
874
Compression parameters (cinfo fields) include:
876
J_DCT_METHOD dct_method
877
Selects the algorithm used for the DCT step. Choices are:
878
JDCT_ISLOW: slow but accurate integer algorithm
879
JDCT_IFAST: faster, less accurate integer method
880
JDCT_FLOAT: floating-point method
881
JDCT_DEFAULT: default method (normally JDCT_ISLOW)
882
JDCT_FASTEST: fastest method (normally JDCT_IFAST)
883
The FLOAT method is very slightly more accurate than the ISLOW method,
884
but may give different results on different machines due to varying
885
roundoff behavior. The integer methods should give the same results
886
on all machines. On machines with sufficiently fast FP hardware, the
887
floating-point method may also be the fastest. The IFAST method is
888
considerably less accurate than the other two; its use is not
889
recommended if high quality is a concern. JDCT_DEFAULT and
890
JDCT_FASTEST are macros configurable by each installation.
892
J_COLOR_SPACE jpeg_color_space
894
The JPEG color space and corresponding number of components; see
895
"Special color spaces", below, for more info. We recommend using
896
jpeg_set_color_space() if you want to change these.
898
boolean optimize_coding
899
TRUE causes the compressor to compute optimal Huffman coding tables
900
for the image. This requires an extra pass over the data and
901
therefore costs a good deal of space and time. The default is
902
FALSE, which tells the compressor to use the supplied or default
903
Huffman tables. In most cases optimal tables save only a few percent
904
of file size compared to the default tables. Note that when this is
905
TRUE, you need not supply Huffman tables at all, and any you do
906
supply will be overwritten.
908
unsigned int restart_interval
910
To emit restart markers in the JPEG file, set one of these nonzero.
911
Set restart_interval to specify the exact interval in MCU blocks.
912
Set restart_in_rows to specify the interval in MCU rows. (If
913
restart_in_rows is not 0, then restart_interval is set after the
914
image width in MCUs is computed.) Defaults are zero (no restarts).
915
One restart marker per MCU row is often a good choice.
916
NOTE: the overhead of restart markers is higher in grayscale JPEG
917
files than in color files, and MUCH higher in progressive JPEGs.
918
If you use restarts, you may want to use larger intervals in those
921
const jpeg_scan_info * scan_info
923
By default, scan_info is NULL; this causes the compressor to write a
924
single-scan sequential JPEG file. If not NULL, scan_info points to
925
an array of scan definition records of length num_scans. The
926
compressor will then write a JPEG file having one scan for each scan
927
definition record. This is used to generate noninterleaved or
928
progressive JPEG files. The library checks that the scan array
929
defines a valid JPEG scan sequence. (jpeg_simple_progression creates
930
a suitable scan definition array for progressive JPEG.) This is
931
discussed further under "Progressive JPEG support".
934
If non-zero, the input image is smoothed; the value should be 1 for
935
minimal smoothing to 100 for maximum smoothing. Consult jcsample.c
936
for details of the smoothing algorithm. The default is zero.
938
boolean write_JFIF_header
939
If TRUE, a JFIF APP0 marker is emitted. jpeg_set_defaults() and
940
jpeg_set_colorspace() set this TRUE if a JFIF-legal JPEG color space
941
(ie, YCbCr or grayscale) is selected, otherwise FALSE.
943
UINT8 JFIF_major_version
944
UINT8 JFIF_minor_version
945
The version number to be written into the JFIF marker.
946
jpeg_set_defaults() initializes the version to 1.01 (major=minor=1).
947
You should set it to 1.02 (major=1, minor=2) if you plan to write
948
any JFIF 1.02 extension markers.
953
The resolution information to be written into the JFIF marker;
954
not used otherwise. density_unit may be 0 for unknown,
955
1 for dots/inch, or 2 for dots/cm. The default values are 0,1,1
956
indicating square pixels of unknown size.
958
boolean write_Adobe_marker
959
If TRUE, an Adobe APP14 marker is emitted. jpeg_set_defaults() and
960
jpeg_set_colorspace() set this TRUE if JPEG color space RGB, CMYK,
961
or YCCK is selected, otherwise FALSE. It is generally a bad idea
962
to set both write_JFIF_header and write_Adobe_marker. In fact,
963
you probably shouldn't change the default settings at all --- the
964
default behavior ensures that the JPEG file's color space can be
965
recognized by the decoder.
967
JQUANT_TBL * quant_tbl_ptrs[NUM_QUANT_TBLS]
968
Pointers to coefficient quantization tables, one per table slot,
969
or NULL if no table is defined for a slot. Usually these should
970
be set via one of the above helper routines; jpeg_add_quant_table()
971
is general enough to define any quantization table. The other
972
routines will set up table slot 0 for luminance quality and table
973
slot 1 for chrominance.
975
JHUFF_TBL * dc_huff_tbl_ptrs[NUM_HUFF_TBLS]
976
JHUFF_TBL * ac_huff_tbl_ptrs[NUM_HUFF_TBLS]
977
Pointers to Huffman coding tables, one per table slot, or NULL if
978
no table is defined for a slot. Slots 0 and 1 are filled with the
979
JPEG sample tables by jpeg_set_defaults(). If you need to allocate
980
more table structures, jpeg_alloc_huff_table() may be used.
981
Note that optimal Huffman tables can be computed for an image
982
by setting optimize_coding, as discussed above; there's seldom
983
any need to mess with providing your own Huffman tables.
985
There are some additional cinfo fields which are not documented here
986
because you currently can't change them; for example, you can't set
987
arith_code TRUE because arithmetic coding is unsupported.
990
Per-component parameters are stored in the struct cinfo.comp_info[i] for
991
component number i. Note that components here refer to components of the
992
JPEG color space, *not* the source image color space. A suitably large
993
comp_info[] array is allocated by jpeg_set_defaults(); if you choose not
994
to use that routine, it's up to you to allocate the array.
997
The one-byte identifier code to be recorded in the JPEG file for
998
this component. For the standard color spaces, we recommend you
999
leave the default values alone.
1003
Horizontal and vertical sampling factors for the component; must
1004
be 1..4 according to the JPEG standard. Note that larger sampling
1005
factors indicate a higher-resolution component; many people find
1006
this behavior quite unintuitive. The default values are 2,2 for
1007
luminance components and 1,1 for chrominance components, except
1008
for grayscale where 1,1 is used.
1011
Quantization table number for component. The default value is
1012
0 for luminance components and 1 for chrominance components.
1016
DC and AC entropy coding table numbers. The default values are
1017
0 for luminance components and 1 for chrominance components.
1020
Must equal the component's index in comp_info[]. (Beginning in
1021
release v6, the compressor library will fill this in automatically;
1025
Decompression parameter selection
1026
---------------------------------
1028
Decompression parameter selection is somewhat simpler than compression
1029
parameter selection, since all of the JPEG internal parameters are
1030
recorded in the source file and need not be supplied by the application.
1031
(Unless you are working with abbreviated files, in which case see
1032
"Abbreviated datastreams", below.) Decompression parameters control
1033
the postprocessing done on the image to deliver it in a format suitable
1034
for the application's use. Many of the parameters control speed/quality
1035
tradeoffs, in which faster decompression may be obtained at the price of
1036
a poorer-quality image. The defaults select the highest quality (slowest)
1039
The following fields in the JPEG object are set by jpeg_read_header() and
1040
may be useful to the application in choosing decompression parameters:
1042
JDIMENSION image_width Width and height of image
1043
JDIMENSION image_height
1044
int num_components Number of color components
1045
J_COLOR_SPACE jpeg_color_space Colorspace of image
1046
boolean saw_JFIF_marker TRUE if a JFIF APP0 marker was seen
1047
UINT8 JFIF_major_version Version information from JFIF marker
1048
UINT8 JFIF_minor_version
1049
UINT8 density_unit Resolution data from JFIF marker
1052
boolean saw_Adobe_marker TRUE if an Adobe APP14 marker was seen
1053
UINT8 Adobe_transform Color transform code from Adobe marker
1055
The JPEG color space, unfortunately, is something of a guess since the JPEG
1056
standard proper does not provide a way to record it. In practice most files
1057
adhere to the JFIF or Adobe conventions, and the decoder will recognize these
1058
correctly. See "Special color spaces", below, for more info.
1061
The decompression parameters that determine the basic properties of the
1064
J_COLOR_SPACE out_color_space
1065
Output color space. jpeg_read_header() sets an appropriate default
1066
based on jpeg_color_space; typically it will be RGB or grayscale.
1067
The application can change this field to request output in a different
1068
colorspace. For example, set it to JCS_GRAYSCALE to get grayscale
1069
output from a color file. (This is useful for previewing: grayscale
1070
output is faster than full color since the color components need not
1071
be processed.) Note that not all possible color space transforms are
1072
currently implemented; you may need to extend jdcolor.c if you want an
1075
unsigned int scale_num, scale_denom
1076
Scale the image by the fraction scale_num/scale_denom. Default is
1077
1/1, or no scaling. Currently, the only supported scaling ratios
1078
are 1/1, 1/2, 1/4, and 1/8. (The library design allows for arbitrary
1079
scaling ratios but this is not likely to be implemented any time soon.)
1080
Smaller scaling ratios permit significantly faster decoding since
1081
fewer pixels need be processed and a simpler IDCT method can be used.
1083
boolean quantize_colors
1084
If set TRUE, colormapped output will be delivered. Default is FALSE,
1085
meaning that full-color output will be delivered.
1087
The next three parameters are relevant only if quantize_colors is TRUE.
1089
int desired_number_of_colors
1090
Maximum number of colors to use in generating a library-supplied color
1091
map (the actual number of colors is returned in a different field).
1092
Default 256. Ignored when the application supplies its own color map.
1094
boolean two_pass_quantize
1095
If TRUE, an extra pass over the image is made to select a custom color
1096
map for the image. This usually looks a lot better than the one-size-
1097
fits-all colormap that is used otherwise. Default is TRUE. Ignored
1098
when the application supplies its own color map.
1100
J_DITHER_MODE dither_mode
1101
Selects color dithering method. Supported values are:
1102
JDITHER_NONE no dithering: fast, very low quality
1103
JDITHER_ORDERED ordered dither: moderate speed and quality
1104
JDITHER_FS Floyd-Steinberg dither: slow, high quality
1105
Default is JDITHER_FS. (At present, ordered dither is implemented
1106
only in the single-pass, standard-colormap case. If you ask for
1107
ordered dither when two_pass_quantize is TRUE or when you supply
1108
an external color map, you'll get F-S dithering.)
1110
When quantize_colors is TRUE, the target color map is described by the next
1111
two fields. colormap is set to NULL by jpeg_read_header(). The application
1112
can supply a color map by setting colormap non-NULL and setting
1113
actual_number_of_colors to the map size. Otherwise, jpeg_start_decompress()
1114
selects a suitable color map and sets these two fields itself.
1115
[Implementation restriction: at present, an externally supplied colormap is
1116
only accepted for 3-component output color spaces.]
1119
The color map, represented as a 2-D pixel array of out_color_components
1120
rows and actual_number_of_colors columns. Ignored if not quantizing.
1121
CAUTION: if the JPEG library creates its own colormap, the storage
1122
pointed to by this field is released by jpeg_finish_decompress().
1123
Copy the colormap somewhere else first, if you want to save it.
1125
int actual_number_of_colors
1126
The number of colors in the color map.
1128
Additional decompression parameters that the application may set include:
1130
J_DCT_METHOD dct_method
1131
Selects the algorithm used for the DCT step. Choices are the same
1132
as described above for compression.
1134
boolean do_fancy_upsampling
1135
If TRUE, do careful upsampling of chroma components. If FALSE,
1136
a faster but sloppier method is used. Default is TRUE. The visual
1137
impact of the sloppier method is often very small.
1139
boolean do_block_smoothing
1140
If TRUE, interblock smoothing is applied in early stages of decoding
1141
progressive JPEG files; if FALSE, not. Default is TRUE. Early
1142
progression stages look "fuzzy" with smoothing, "blocky" without.
1143
In any case, block smoothing ceases to be applied after the first few
1144
AC coefficients are known to full accuracy, so it is relevant only
1145
when using buffered-image mode for progressive images.
1147
boolean enable_1pass_quant
1148
boolean enable_external_quant
1149
boolean enable_2pass_quant
1150
These are significant only in buffered-image mode, which is
1151
described in its own section below.
1154
The output image dimensions are given by the following fields. These are
1155
computed from the source image dimensions and the decompression parameters
1156
by jpeg_start_decompress(). You can also call jpeg_calc_output_dimensions()
1157
to obtain the values that will result from the current parameter settings.
1158
This can be useful if you are trying to pick a scaling ratio that will get
1159
close to a desired target size. It's also important if you are using the
1160
JPEG library's memory manager to allocate output buffer space, because you
1161
are supposed to request such buffers *before* jpeg_start_decompress().
1163
JDIMENSION output_width Actual dimensions of output image.
1164
JDIMENSION output_height
1165
int out_color_components Number of color components in out_color_space.
1166
int output_components Number of color components returned.
1167
int rec_outbuf_height Recommended height of scanline buffer.
1169
When quantizing colors, output_components is 1, indicating a single color map
1170
index per pixel. Otherwise it equals out_color_components. The output arrays
1171
are required to be output_width * output_components JSAMPLEs wide.
1173
rec_outbuf_height is the recommended minimum height (in scanlines) of the
1174
buffer passed to jpeg_read_scanlines(). If the buffer is smaller, the
1175
library will still work, but time will be wasted due to unnecessary data
1176
copying. In high-quality modes, rec_outbuf_height is always 1, but some
1177
faster, lower-quality modes set it to larger values (typically 2 to 4).
1178
If you are going to ask for a high-speed processing mode, you may as well
1179
go to the trouble of honoring rec_outbuf_height so as to avoid data copying.
1180
(An output buffer larger than rec_outbuf_height lines is OK, but won't
1181
provide any material speed improvement over that height.)
1184
Special color spaces
1185
--------------------
1187
The JPEG standard itself is "color blind" and doesn't specify any particular
1188
color space. It is customary to convert color data to a luminance/chrominance
1189
color space before compressing, since this permits greater compression. The
1190
existing de-facto JPEG file format standards specify YCbCr or grayscale data
1191
(JFIF), or grayscale, RGB, YCbCr, CMYK, or YCCK (Adobe). For special
1192
applications such as multispectral images, other color spaces can be used,
1193
but it must be understood that such files will be unportable.
1195
The JPEG library can handle the most common colorspace conversions (namely
1196
RGB <=> YCbCr and CMYK <=> YCCK). It can also deal with data of an unknown
1197
color space, passing it through without conversion. If you deal extensively
1198
with an unusual color space, you can easily extend the library to understand
1199
additional color spaces and perform appropriate conversions.
1201
For compression, the source data's color space is specified by field
1202
in_color_space. This is transformed to the JPEG file's color space given
1203
by jpeg_color_space. jpeg_set_defaults() chooses a reasonable JPEG color
1204
space depending on in_color_space, but you can override this by calling
1205
jpeg_set_colorspace(). Of course you must select a supported transformation.
1206
jccolor.c currently supports the following transformations:
1211
plus the null transforms: GRAYSCALE => GRAYSCALE, RGB => RGB,
1212
YCbCr => YCbCr, CMYK => CMYK, YCCK => YCCK, and UNKNOWN => UNKNOWN.
1214
The de-facto file format standards (JFIF and Adobe) specify APPn markers that
1215
indicate the color space of the JPEG file. It is important to ensure that
1216
these are written correctly, or omitted if the JPEG file's color space is not
1217
one of the ones supported by the de-facto standards. jpeg_set_colorspace()
1218
will set the compression parameters to include or omit the APPn markers
1219
properly, so long as it is told the truth about the JPEG color space.
1220
For example, if you are writing some random 3-component color space without
1221
conversion, don't try to fake out the library by setting in_color_space and
1222
jpeg_color_space to JCS_YCbCr; use JCS_UNKNOWN. You may want to write an
1223
APPn marker of your own devising to identify the colorspace --- see "Special
1226
When told that the color space is UNKNOWN, the library will default to using
1227
luminance-quality compression parameters for all color components. You may
1228
well want to change these parameters. See the source code for
1229
jpeg_set_colorspace(), in jcparam.c, for details.
1231
For decompression, the JPEG file's color space is given in jpeg_color_space,
1232
and this is transformed to the output color space out_color_space.
1233
jpeg_read_header's setting of jpeg_color_space can be relied on if the file
1234
conforms to JFIF or Adobe conventions, but otherwise it is no better than a
1235
guess. If you know the JPEG file's color space for certain, you can override
1236
jpeg_read_header's guess by setting jpeg_color_space. jpeg_read_header also
1237
selects a default output color space based on (its guess of) jpeg_color_space;
1238
set out_color_space to override this. Again, you must select a supported
1239
transformation. jdcolor.c currently supports
1244
as well as the null transforms. (Since GRAYSCALE=>RGB is provided, an
1245
application can force grayscale JPEGs to look like color JPEGs if it only
1246
wants to handle one case.)
1248
The two-pass color quantizer, jquant2.c, is specialized to handle RGB data
1249
(it weights distances appropriately for RGB colors). You'll need to modify
1250
the code if you want to use it for non-RGB output color spaces. Note that
1251
jquant2.c is used to map to an application-supplied colormap as well as for
1252
the normal two-pass colormap selection process.
1254
CAUTION: it appears that Adobe Photoshop writes inverted data in CMYK JPEG
1255
files: 0 represents 100% ink coverage, rather than 0% ink as you'd expect.
1256
This is arguably a bug in Photoshop, but if you need to work with Photoshop
1257
CMYK files, you will have to deal with it in your application. We cannot
1258
"fix" this in the library by inverting the data during the CMYK<=>YCCK
1259
transform, because that would break other applications, notably Ghostscript.
1260
Photoshop versions prior to 3.0 write EPS files containing JPEG-encoded CMYK
1261
data in the same inverted-YCCK representation used in bare JPEG files, but
1262
the surrounding PostScript code performs an inversion using the PS image
1263
operator. I am told that Photoshop 3.0 will write uninverted YCCK in
1264
EPS/JPEG files, and will omit the PS-level inversion. (But the data
1265
polarity used in bare JPEG files will not change in 3.0.) In either case,
1266
the JPEG library must not invert the data itself, or else Ghostscript would
1267
read these EPS files incorrectly.
1273
When the default error handler is used, any error detected inside the JPEG
1274
routines will cause a message to be printed on stderr, followed by exit().
1275
You can supply your own error handling routines to override this behavior
1276
and to control the treatment of nonfatal warnings and trace/debug messages.
1277
The file example.c illustrates the most common case, which is to have the
1278
application regain control after an error rather than exiting.
1280
The JPEG library never writes any message directly; it always goes through
1281
the error handling routines. Three classes of messages are recognized:
1282
* Fatal errors: the library cannot continue.
1283
* Warnings: the library can continue, but the data is corrupt, and a
1284
damaged output image is likely to result.
1285
* Trace/informational messages. These come with a trace level indicating
1286
the importance of the message; you can control the verbosity of the
1287
program by adjusting the maximum trace level that will be displayed.
1289
You may, if you wish, simply replace the entire JPEG error handling module
1290
(jerror.c) with your own code. However, you can avoid code duplication by
1291
only replacing some of the routines depending on the behavior you need.
1292
This is accomplished by calling jpeg_std_error() as usual, but then overriding
1293
some of the method pointers in the jpeg_error_mgr struct, as illustrated by
1296
All of the error handling routines will receive a pointer to the JPEG object
1297
(a j_common_ptr which points to either a jpeg_compress_struct or a
1298
jpeg_decompress_struct; if you need to tell which, test the is_decompressor
1299
field). This struct includes a pointer to the error manager struct in its
1300
"err" field. Frequently, custom error handler routines will need to access
1301
additional data which is not known to the JPEG library or the standard error
1302
handler. The most convenient way to do this is to embed either the JPEG
1303
object or the jpeg_error_mgr struct in a larger structure that contains
1304
additional fields; then casting the passed pointer provides access to the
1305
additional fields. Again, see example.c for one way to do it. (Beginning
1306
with IJG version 6b, there is also a void pointer "client_data" in each
1307
JPEG object, which the application can also use to find related data.
1308
The library does not touch client_data at all.)
1310
The individual methods that you might wish to override are:
1312
error_exit (j_common_ptr cinfo)
1313
Receives control for a fatal error. Information sufficient to
1314
generate the error message has been stored in cinfo->err; call
1315
output_message to display it. Control must NOT return to the caller;
1316
generally this routine will exit() or longjmp() somewhere.
1317
Typically you would override this routine to get rid of the exit()
1318
default behavior. Note that if you continue processing, you should
1319
clean up the JPEG object with jpeg_abort() or jpeg_destroy().
1321
output_message (j_common_ptr cinfo)
1322
Actual output of any JPEG message. Override this to send messages
1323
somewhere other than stderr. Note that this method does not know
1324
how to generate a message, only where to send it.
1326
format_message (j_common_ptr cinfo, char * buffer)
1327
Constructs a readable error message string based on the error info
1328
stored in cinfo->err. This method is called by output_message. Few
1329
applications should need to override this method. One possible
1330
reason for doing so is to implement dynamic switching of error message
1333
emit_message (j_common_ptr cinfo, int msg_level)
1334
Decide whether or not to emit a warning or trace message; if so,
1335
calls output_message. The main reason for overriding this method
1336
would be to abort on warnings. msg_level is -1 for warnings,
1337
0 and up for trace messages.
1339
Only error_exit() and emit_message() are called from the rest of the JPEG
1340
library; the other two are internal to the error handler.
1342
The actual message texts are stored in an array of strings which is pointed to
1343
by the field err->jpeg_message_table. The messages are numbered from 0 to
1344
err->last_jpeg_message, and it is these code numbers that are used in the
1345
JPEG library code. You could replace the message texts (for instance, with
1346
messages in French or German) by changing the message table pointer. See
1347
jerror.h for the default texts. CAUTION: this table will almost certainly
1348
change or grow from one library version to the next.
1350
It may be useful for an application to add its own message texts that are
1351
handled by the same mechanism. The error handler supports a second "add-on"
1352
message table for this purpose. To define an addon table, set the pointer
1353
err->addon_message_table and the message numbers err->first_addon_message and
1354
err->last_addon_message. If you number the addon messages beginning at 1000
1355
or so, you won't have to worry about conflicts with the library's built-in
1356
messages. See the sample applications cjpeg/djpeg for an example of using
1357
addon messages (the addon messages are defined in cderror.h).
1359
Actual invocation of the error handler is done via macros defined in jerror.h:
1360
ERREXITn(...) for fatal errors
1361
WARNMSn(...) for corrupt-data warnings
1362
TRACEMSn(...) for trace and informational messages.
1363
These macros store the message code and any additional parameters into the
1364
error handler struct, then invoke the error_exit() or emit_message() method.
1365
The variants of each macro are for varying numbers of additional parameters.
1366
The additional parameters are inserted into the generated message using
1367
standard printf() format codes.
1369
See jerror.h and jerror.c for further details.
1372
Compressed data handling (source and destination managers)
1373
----------------------------------------------------------
1375
The JPEG compression library sends its compressed data to a "destination
1376
manager" module. The default destination manager just writes the data to a
1377
stdio stream, but you can provide your own manager to do something else.
1378
Similarly, the decompression library calls a "source manager" to obtain the
1379
compressed data; you can provide your own source manager if you want the data
1380
to come from somewhere other than a stdio stream.
1382
In both cases, compressed data is processed a bufferload at a time: the
1383
destination or source manager provides a work buffer, and the library invokes
1384
the manager only when the buffer is filled or emptied. (You could define a
1385
one-character buffer to force the manager to be invoked for each byte, but
1386
that would be rather inefficient.) The buffer's size and location are
1387
controlled by the manager, not by the library. For example, if you desired to
1388
decompress a JPEG datastream that was all in memory, you could just make the
1389
buffer pointer and length point to the original data in memory. Then the
1390
buffer-reload procedure would be invoked only if the decompressor ran off the
1391
end of the datastream, which would indicate an erroneous datastream.
1393
The work buffer is defined as an array of datatype JOCTET, which is generally
1394
"char" or "unsigned char". On a machine where char is not exactly 8 bits
1395
wide, you must define JOCTET as a wider data type and then modify the data
1396
source and destination modules to transcribe the work arrays into 8-bit units
1397
on external storage.
1399
A data destination manager struct contains a pointer and count defining the
1400
next byte to write in the work buffer and the remaining free space:
1402
JOCTET * next_output_byte; /* => next byte to write in buffer */
1403
size_t free_in_buffer; /* # of byte spaces remaining in buffer */
1405
The library increments the pointer and decrements the count until the buffer
1406
is filled. The manager's empty_output_buffer method must reset the pointer
1407
and count. The manager is expected to remember the buffer's starting address
1408
and total size in private fields not visible to the library.
1410
A data destination manager provides three methods:
1412
init_destination (j_compress_ptr cinfo)
1413
Initialize destination. This is called by jpeg_start_compress()
1414
before any data is actually written. It must initialize
1415
next_output_byte and free_in_buffer. free_in_buffer must be
1416
initialized to a positive value.
1418
empty_output_buffer (j_compress_ptr cinfo)
1419
This is called whenever the buffer has filled (free_in_buffer
1420
reaches zero). In typical applications, it should write out the
1421
*entire* buffer (use the saved start address and buffer length;
1422
ignore the current state of next_output_byte and free_in_buffer).
1423
Then reset the pointer & count to the start of the buffer, and
1424
return TRUE indicating that the buffer has been dumped.
1425
free_in_buffer must be set to a positive value when TRUE is
1426
returned. A FALSE return should only be used when I/O suspension is
1427
desired (this operating mode is discussed in the next section).
1429
term_destination (j_compress_ptr cinfo)
1430
Terminate destination --- called by jpeg_finish_compress() after all
1431
data has been written. In most applications, this must flush any
1432
data remaining in the buffer. Use either next_output_byte or
1433
free_in_buffer to determine how much data is in the buffer.
1435
term_destination() is NOT called by jpeg_abort() or jpeg_destroy(). If you
1436
want the destination manager to be cleaned up during an abort, you must do it
1439
You will also need code to create a jpeg_destination_mgr struct, fill in its
1440
method pointers, and insert a pointer to the struct into the "dest" field of
1441
the JPEG compression object. This can be done in-line in your setup code if
1442
you like, but it's probably cleaner to provide a separate routine similar to
1443
the jpeg_stdio_dest() routine of the supplied destination manager.
1445
Decompression source managers follow a parallel design, but with some
1446
additional frammishes. The source manager struct contains a pointer and count
1447
defining the next byte to read from the work buffer and the number of bytes
1450
const JOCTET * next_input_byte; /* => next byte to read from buffer */
1451
size_t bytes_in_buffer; /* # of bytes remaining in buffer */
1453
The library increments the pointer and decrements the count until the buffer
1454
is emptied. The manager's fill_input_buffer method must reset the pointer and
1455
count. In most applications, the manager must remember the buffer's starting
1456
address and total size in private fields not visible to the library.
1458
A data source manager provides five methods:
1460
init_source (j_decompress_ptr cinfo)
1461
Initialize source. This is called by jpeg_read_header() before any
1462
data is actually read. Unlike init_destination(), it may leave
1463
bytes_in_buffer set to 0 (in which case a fill_input_buffer() call
1464
will occur immediately).
1466
fill_input_buffer (j_decompress_ptr cinfo)
1467
This is called whenever bytes_in_buffer has reached zero and more
1468
data is wanted. In typical applications, it should read fresh data
1469
into the buffer (ignoring the current state of next_input_byte and
1470
bytes_in_buffer), reset the pointer & count to the start of the
1471
buffer, and return TRUE indicating that the buffer has been reloaded.
1472
It is not necessary to fill the buffer entirely, only to obtain at
1473
least one more byte. bytes_in_buffer MUST be set to a positive value
1474
if TRUE is returned. A FALSE return should only be used when I/O
1475
suspension is desired (this mode is discussed in the next section).
1477
skip_input_data (j_decompress_ptr cinfo, long num_bytes)
1478
Skip num_bytes worth of data. The buffer pointer and count should
1479
be advanced over num_bytes input bytes, refilling the buffer as
1480
needed. This is used to skip over a potentially large amount of
1481
uninteresting data (such as an APPn marker). In some applications
1482
it may be possible to optimize away the reading of the skipped data,
1483
but it's not clear that being smart is worth much trouble; large
1484
skips are uncommon. bytes_in_buffer may be zero on return.
1485
A zero or negative skip count should be treated as a no-op.
1487
resync_to_restart (j_decompress_ptr cinfo, int desired)
1488
This routine is called only when the decompressor has failed to find
1489
a restart (RSTn) marker where one is expected. Its mission is to
1490
find a suitable point for resuming decompression. For most
1491
applications, we recommend that you just use the default resync
1492
procedure, jpeg_resync_to_restart(). However, if you are able to back
1493
up in the input data stream, or if you have a-priori knowledge about
1494
the likely location of restart markers, you may be able to do better.
1495
Read the read_restart_marker() and jpeg_resync_to_restart() routines
1496
in jdmarker.c if you think you'd like to implement your own resync
1499
term_source (j_decompress_ptr cinfo)
1500
Terminate source --- called by jpeg_finish_decompress() after all
1501
data has been read. Often a no-op.
1503
For both fill_input_buffer() and skip_input_data(), there is no such thing
1504
as an EOF return. If the end of the file has been reached, the routine has
1505
a choice of exiting via ERREXIT() or inserting fake data into the buffer.
1506
In most cases, generating a warning message and inserting a fake EOI marker
1507
is the best course of action --- this will allow the decompressor to output
1508
however much of the image is there. In pathological cases, the decompressor
1509
may swallow the EOI and again demand data ... just keep feeding it fake EOIs.
1510
jdatasrc.c illustrates the recommended error recovery behavior.
1512
term_source() is NOT called by jpeg_abort() or jpeg_destroy(). If you want
1513
the source manager to be cleaned up during an abort, you must do it yourself.
1515
You will also need code to create a jpeg_source_mgr struct, fill in its method
1516
pointers, and insert a pointer to the struct into the "src" field of the JPEG
1517
decompression object. This can be done in-line in your setup code if you
1518
like, but it's probably cleaner to provide a separate routine similar to the
1519
jpeg_stdio_src() routine of the supplied source manager.
1521
For more information, consult the stdio source and destination managers
1522
in jdatasrc.c and jdatadst.c.
1528
Some applications need to use the JPEG library as an incremental memory-to-
1529
memory filter: when the compressed data buffer is filled or emptied, they want
1530
control to return to the outer loop, rather than expecting that the buffer can
1531
be emptied or reloaded within the data source/destination manager subroutine.
1532
The library supports this need by providing an "I/O suspension" mode, which we
1533
describe in this section.
1535
The I/O suspension mode is not a panacea: nothing is guaranteed about the
1536
maximum amount of time spent in any one call to the library, so it will not
1537
eliminate response-time problems in single-threaded applications. If you
1538
need guaranteed response time, we suggest you "bite the bullet" and implement
1539
a real multi-tasking capability.
1541
To use I/O suspension, cooperation is needed between the calling application
1542
and the data source or destination manager; you will always need a custom
1543
source/destination manager. (Please read the previous section if you haven't
1544
already.) The basic idea is that the empty_output_buffer() or
1545
fill_input_buffer() routine is a no-op, merely returning FALSE to indicate
1546
that it has done nothing. Upon seeing this, the JPEG library suspends
1547
operation and returns to its caller. The surrounding application is
1548
responsible for emptying or refilling the work buffer before calling the
1551
Compression suspension:
1553
For compression suspension, use an empty_output_buffer() routine that returns
1554
FALSE; typically it will not do anything else. This will cause the
1555
compressor to return to the caller of jpeg_write_scanlines(), with the return
1556
value indicating that not all the supplied scanlines have been accepted.
1557
The application must make more room in the output buffer, adjust the output
1558
buffer pointer/count appropriately, and then call jpeg_write_scanlines()
1559
again, pointing to the first unconsumed scanline.
1561
When forced to suspend, the compressor will backtrack to a convenient stopping
1562
point (usually the start of the current MCU); it will regenerate some output
1563
data when restarted. Therefore, although empty_output_buffer() is only
1564
called when the buffer is filled, you should NOT write out the entire buffer
1565
after a suspension. Write only the data up to the current position of
1566
next_output_byte/free_in_buffer. The data beyond that point will be
1567
regenerated after resumption.
1569
Because of the backtracking behavior, a good-size output buffer is essential
1570
for efficiency; you don't want the compressor to suspend often. (In fact, an
1571
overly small buffer could lead to infinite looping, if a single MCU required
1572
more data than would fit in the buffer.) We recommend a buffer of at least
1573
several Kbytes. You may want to insert explicit code to ensure that you don't
1574
call jpeg_write_scanlines() unless there is a reasonable amount of space in
1575
the output buffer; in other words, flush the buffer before trying to compress
1578
The compressor does not allow suspension while it is trying to write JPEG
1579
markers at the beginning and end of the file. This means that:
1580
* At the beginning of a compression operation, there must be enough free
1581
space in the output buffer to hold the header markers (typically 600 or
1582
so bytes). The recommended buffer size is bigger than this anyway, so
1583
this is not a problem as long as you start with an empty buffer. However,
1584
this restriction might catch you if you insert large special markers, such
1585
as a JFIF thumbnail image, without flushing the buffer afterwards.
1586
* When you call jpeg_finish_compress(), there must be enough space in the
1587
output buffer to emit any buffered data and the final EOI marker. In the
1588
current implementation, half a dozen bytes should suffice for this, but
1589
for safety's sake we recommend ensuring that at least 100 bytes are free
1590
before calling jpeg_finish_compress().
1592
A more significant restriction is that jpeg_finish_compress() cannot suspend.
1593
This means you cannot use suspension with multi-pass operating modes, namely
1594
Huffman code optimization and multiple-scan output. Those modes write the
1595
whole file during jpeg_finish_compress(), which will certainly result in
1596
buffer overrun. (Note that this restriction applies only to compression,
1597
not decompression. The decompressor supports input suspension in all of its
1600
Decompression suspension:
1602
For decompression suspension, use a fill_input_buffer() routine that simply
1603
returns FALSE (except perhaps during error recovery, as discussed below).
1604
This will cause the decompressor to return to its caller with an indication
1605
that suspension has occurred. This can happen at four places:
1606
* jpeg_read_header(): will return JPEG_SUSPENDED.
1607
* jpeg_start_decompress(): will return FALSE, rather than its usual TRUE.
1608
* jpeg_read_scanlines(): will return the number of scanlines already
1609
completed (possibly 0).
1610
* jpeg_finish_decompress(): will return FALSE, rather than its usual TRUE.
1611
The surrounding application must recognize these cases, load more data into
1612
the input buffer, and repeat the call. In the case of jpeg_read_scanlines(),
1613
increment the passed pointers past any scanlines successfully read.
1615
Just as with compression, the decompressor will typically backtrack to a
1616
convenient restart point before suspending. When fill_input_buffer() is
1617
called, next_input_byte/bytes_in_buffer point to the current restart point,
1618
which is where the decompressor will backtrack to if FALSE is returned.
1619
The data beyond that position must NOT be discarded if you suspend; it needs
1620
to be re-read upon resumption. In most implementations, you'll need to shift
1621
this data down to the start of your work buffer and then load more data after
1622
it. Again, this behavior means that a several-Kbyte work buffer is essential
1623
for decent performance; furthermore, you should load a reasonable amount of
1624
new data before resuming decompression. (If you loaded, say, only one new
1625
byte each time around, you could waste a LOT of cycles.)
1627
The skip_input_data() source manager routine requires special care in a
1628
suspension scenario. This routine is NOT granted the ability to suspend the
1629
decompressor; it can decrement bytes_in_buffer to zero, but no more. If the
1630
requested skip distance exceeds the amount of data currently in the input
1631
buffer, then skip_input_data() must set bytes_in_buffer to zero and record the
1632
additional skip distance somewhere else. The decompressor will immediately
1633
call fill_input_buffer(), which should return FALSE, which will cause a
1634
suspension return. The surrounding application must then arrange to discard
1635
the recorded number of bytes before it resumes loading the input buffer.
1636
(Yes, this design is rather baroque, but it avoids complexity in the far more
1637
common case where a non-suspending source manager is used.)
1639
If the input data has been exhausted, we recommend that you emit a warning
1640
and insert dummy EOI markers just as a non-suspending data source manager
1641
would do. This can be handled either in the surrounding application logic or
1642
within fill_input_buffer(); the latter is probably more efficient. If
1643
fill_input_buffer() knows that no more data is available, it can set the
1644
pointer/count to point to a dummy EOI marker and then return TRUE just as
1645
though it had read more data in a non-suspending situation.
1647
The decompressor does not attempt to suspend within standard JPEG markers;
1648
instead it will backtrack to the start of the marker and reprocess the whole
1649
marker next time. Hence the input buffer must be large enough to hold the
1650
longest standard marker in the file. Standard JPEG markers should normally
1651
not exceed a few hundred bytes each (DHT tables are typically the longest).
1652
We recommend at least a 2K buffer for performance reasons, which is much
1653
larger than any correct marker is likely to be. For robustness against
1654
damaged marker length counts, you may wish to insert a test in your
1655
application for the case that the input buffer is completely full and yet
1656
the decoder has suspended without consuming any data --- otherwise, if this
1657
situation did occur, it would lead to an endless loop. (The library can't
1658
provide this test since it has no idea whether "the buffer is full", or
1659
even whether there is a fixed-size input buffer.)
1661
The input buffer would need to be 64K to allow for arbitrary COM or APPn
1662
markers, but these are handled specially: they are either saved into allocated
1663
memory, or skipped over by calling skip_input_data(). In the former case,
1664
suspension is handled correctly, and in the latter case, the problem of
1665
buffer overrun is placed on skip_input_data's shoulders, as explained above.
1666
Note that if you provide your own marker handling routine for large markers,
1667
you should consider how to deal with buffer overflow.
1669
Multiple-buffer management:
1671
In some applications it is desirable to store the compressed data in a linked
1672
list of buffer areas, so as to avoid data copying. This can be handled by
1673
having empty_output_buffer() or fill_input_buffer() set the pointer and count
1674
to reference the next available buffer; FALSE is returned only if no more
1675
buffers are available. Although seemingly straightforward, there is a
1676
pitfall in this approach: the backtrack that occurs when FALSE is returned
1677
could back up into an earlier buffer. For example, when fill_input_buffer()
1678
is called, the current pointer & count indicate the backtrack restart point.
1679
Since fill_input_buffer() will set the pointer and count to refer to a new
1680
buffer, the restart position must be saved somewhere else. Suppose a second
1681
call to fill_input_buffer() occurs in the same library call, and no
1682
additional input data is available, so fill_input_buffer must return FALSE.
1683
If the JPEG library has not moved the pointer/count forward in the current
1684
buffer, then *the correct restart point is the saved position in the prior
1685
buffer*. Prior buffers may be discarded only after the library establishes
1686
a restart point within a later buffer. Similar remarks apply for output into
1689
The library will never attempt to backtrack over a skip_input_data() call,
1690
so any skipped data can be permanently discarded. You still have to deal
1691
with the case of skipping not-yet-received data, however.
1693
It's much simpler to use only a single buffer; when fill_input_buffer() is
1694
called, move any unconsumed data (beyond the current pointer/count) down to
1695
the beginning of this buffer and then load new data into the remaining buffer
1696
space. This approach requires a little more data copying but is far easier
1700
Progressive JPEG support
1701
------------------------
1703
Progressive JPEG rearranges the stored data into a series of scans of
1704
increasing quality. In situations where a JPEG file is transmitted across a
1705
slow communications link, a decoder can generate a low-quality image very
1706
quickly from the first scan, then gradually improve the displayed quality as
1707
more scans are received. The final image after all scans are complete is
1708
identical to that of a regular (sequential) JPEG file of the same quality
1709
setting. Progressive JPEG files are often slightly smaller than equivalent
1710
sequential JPEG files, but the possibility of incremental display is the main
1711
reason for using progressive JPEG.
1713
The IJG encoder library generates progressive JPEG files when given a
1714
suitable "scan script" defining how to divide the data into scans.
1715
Creation of progressive JPEG files is otherwise transparent to the encoder.
1716
Progressive JPEG files can also be read transparently by the decoder library.
1717
If the decoding application simply uses the library as defined above, it
1718
will receive a final decoded image without any indication that the file was
1719
progressive. Of course, this approach does not allow incremental display.
1720
To perform incremental display, an application needs to use the decoder
1721
library's "buffered-image" mode, in which it receives a decoded image
1724
Each displayed scan requires about as much work to decode as a full JPEG
1725
image of the same size, so the decoder must be fairly fast in relation to the
1726
data transmission rate in order to make incremental display useful. However,
1727
it is possible to skip displaying the image and simply add the incoming bits
1728
to the decoder's coefficient buffer. This is fast because only Huffman
1729
decoding need be done, not IDCT, upsampling, colorspace conversion, etc.
1730
The IJG decoder library allows the application to switch dynamically between
1731
displaying the image and simply absorbing the incoming bits. A properly
1732
coded application can automatically adapt the number of display passes to
1733
suit the time available as the image is received. Also, a final
1734
higher-quality display cycle can be performed from the buffered data after
1735
the end of the file is reached.
1737
Progressive compression:
1739
To create a progressive JPEG file (or a multiple-scan sequential JPEG file),
1740
set the scan_info cinfo field to point to an array of scan descriptors, and
1741
perform compression as usual. Instead of constructing your own scan list,
1742
you can call the jpeg_simple_progression() helper routine to create a
1743
recommended progression sequence; this method should be used by all
1744
applications that don't want to get involved in the nitty-gritty of
1745
progressive scan sequence design. (If you want to provide user control of
1746
scan sequences, you may wish to borrow the scan script reading code found
1747
in rdswitch.c, so that you can read scan script files just like cjpeg's.)
1748
When scan_info is not NULL, the compression library will store DCT'd data
1749
into a buffer array as jpeg_write_scanlines() is called, and will emit all
1750
the requested scans during jpeg_finish_compress(). This implies that
1751
multiple-scan output cannot be created with a suspending data destination
1752
manager, since jpeg_finish_compress() does not support suspension. We
1753
should also note that the compressor currently forces Huffman optimization
1754
mode when creating a progressive JPEG file, because the default Huffman
1755
tables are unsuitable for progressive files.
1757
Progressive decompression:
1759
When buffered-image mode is not used, the decoder library will read all of
1760
a multi-scan file during jpeg_start_decompress(), so that it can provide a
1761
final decoded image. (Here "multi-scan" means either progressive or
1762
multi-scan sequential.) This makes multi-scan files transparent to the
1763
decoding application. However, existing applications that used suspending
1764
input with version 5 of the IJG library will need to be modified to check
1765
for a suspension return from jpeg_start_decompress().
1767
To perform incremental display, an application must use the library's
1768
buffered-image mode. This is described in the next section.
1774
In buffered-image mode, the library stores the partially decoded image in a
1775
coefficient buffer, from which it can be read out as many times as desired.
1776
This mode is typically used for incremental display of progressive JPEG files,
1777
but it can be used with any JPEG file. Each scan of a progressive JPEG file
1778
adds more data (more detail) to the buffered image. The application can
1779
display in lockstep with the source file (one display pass per input scan),
1780
or it can allow input processing to outrun display processing. By making
1781
input and display processing run independently, it is possible for the
1782
application to adapt progressive display to a wide range of data transmission
1785
The basic control flow for buffered-image decoding is
1787
jpeg_create_decompress()
1790
set overall decompression parameters
1791
cinfo.buffered_image = TRUE; /* select buffered-image mode */
1792
jpeg_start_decompress()
1793
for (each output pass) {
1794
adjust output decompression parameters if required
1795
jpeg_start_output() /* start a new output pass */
1796
for (all scanlines in image) {
1797
jpeg_read_scanlines()
1800
jpeg_finish_output() /* terminate output pass */
1802
jpeg_finish_decompress()
1803
jpeg_destroy_decompress()
1805
This differs from ordinary unbuffered decoding in that there is an additional
1806
level of looping. The application can choose how many output passes to make
1807
and how to display each pass.
1809
The simplest approach to displaying progressive images is to do one display
1810
pass for each scan appearing in the input file. In this case the outer loop
1811
condition is typically
1812
while (! jpeg_input_complete(&cinfo))
1813
and the start-output call should read
1814
jpeg_start_output(&cinfo, cinfo.input_scan_number);
1815
The second parameter to jpeg_start_output() indicates which scan of the input
1816
file is to be displayed; the scans are numbered starting at 1 for this
1817
purpose. (You can use a loop counter starting at 1 if you like, but using
1818
the library's input scan counter is easier.) The library automatically reads
1819
data as necessary to complete each requested scan, and jpeg_finish_output()
1820
advances to the next scan or end-of-image marker (hence input_scan_number
1821
will be incremented by the time control arrives back at jpeg_start_output()).
1822
With this technique, data is read from the input file only as needed, and
1823
input and output processing run in lockstep.
1825
After reading the final scan and reaching the end of the input file, the
1826
buffered image remains available; it can be read additional times by
1827
repeating the jpeg_start_output()/jpeg_read_scanlines()/jpeg_finish_output()
1828
sequence. For example, a useful technique is to use fast one-pass color
1829
quantization for display passes made while the image is arriving, followed by
1830
a final display pass using two-pass quantization for highest quality. This
1831
is done by changing the library parameters before the final output pass.
1832
Changing parameters between passes is discussed in detail below.
1834
In general the last scan of a progressive file cannot be recognized as such
1835
until after it is read, so a post-input display pass is the best approach if
1836
you want special processing in the final pass.
1838
When done with the image, be sure to call jpeg_finish_decompress() to release
1839
the buffered image (or just use jpeg_destroy_decompress()).
1841
If input data arrives faster than it can be displayed, the application can
1842
cause the library to decode input data in advance of what's needed to produce
1843
output. This is done by calling the routine jpeg_consume_input().
1844
The return value is one of the following:
1845
JPEG_REACHED_SOS: reached an SOS marker (the start of a new scan)
1846
JPEG_REACHED_EOI: reached the EOI marker (end of image)
1847
JPEG_ROW_COMPLETED: completed reading one MCU row of compressed data
1848
JPEG_SCAN_COMPLETED: completed reading last MCU row of current scan
1849
JPEG_SUSPENDED: suspended before completing any of the above
1850
(JPEG_SUSPENDED can occur only if a suspending data source is used.) This
1851
routine can be called at any time after initializing the JPEG object. It
1852
reads some additional data and returns when one of the indicated significant
1853
events occurs. (If called after the EOI marker is reached, it will
1854
immediately return JPEG_REACHED_EOI without attempting to read more data.)
1856
The library's output processing will automatically call jpeg_consume_input()
1857
whenever the output processing overtakes the input; thus, simple lockstep
1858
display requires no direct calls to jpeg_consume_input(). But by adding
1859
calls to jpeg_consume_input(), you can absorb data in advance of what is
1860
being displayed. This has two benefits:
1861
* You can limit buildup of unprocessed data in your input buffer.
1862
* You can eliminate extra display passes by paying attention to the
1863
state of the library's input processing.
1865
The first of these benefits only requires interspersing calls to
1866
jpeg_consume_input() with your display operations and any other processing
1867
you may be doing. To avoid wasting cycles due to backtracking, it's best to
1868
call jpeg_consume_input() only after a hundred or so new bytes have arrived.
1869
This is discussed further under "I/O suspension", above. (Note: the JPEG
1870
library currently is not thread-safe. You must not call jpeg_consume_input()
1871
from one thread of control if a different library routine is working on the
1872
same JPEG object in another thread.)
1874
When input arrives fast enough that more than one new scan is available
1875
before you start a new output pass, you may as well skip the output pass
1876
corresponding to the completed scan. This occurs for free if you pass
1877
cinfo.input_scan_number as the target scan number to jpeg_start_output().
1878
The input_scan_number field is simply the index of the scan currently being
1879
consumed by the input processor. You can ensure that this is up-to-date by
1880
emptying the input buffer just before calling jpeg_start_output(): call
1881
jpeg_consume_input() repeatedly until it returns JPEG_SUSPENDED or
1884
The target scan number passed to jpeg_start_output() is saved in the
1885
cinfo.output_scan_number field. The library's output processing calls
1886
jpeg_consume_input() whenever the current input scan number and row within
1887
that scan is less than or equal to the current output scan number and row.
1888
Thus, input processing can "get ahead" of the output processing but is not
1889
allowed to "fall behind". You can achieve several different effects by
1890
manipulating this interlock rule. For example, if you pass a target scan
1891
number greater than the current input scan number, the output processor will
1892
wait until that scan starts to arrive before producing any output. (To avoid
1893
an infinite loop, the target scan number is automatically reset to the last
1894
scan number when the end of image is reached. Thus, if you specify a large
1895
target scan number, the library will just absorb the entire input file and
1896
then perform an output pass. This is effectively the same as what
1897
jpeg_start_decompress() does when you don't select buffered-image mode.)
1898
When you pass a target scan number equal to the current input scan number,
1899
the image is displayed no faster than the current input scan arrives. The
1900
final possibility is to pass a target scan number less than the current input
1901
scan number; this disables the input/output interlock and causes the output
1902
processor to simply display whatever it finds in the image buffer, without
1903
waiting for input. (However, the library will not accept a target scan
1904
number less than one, so you can't avoid waiting for the first scan.)
1906
When data is arriving faster than the output display processing can advance
1907
through the image, jpeg_consume_input() will store data into the buffered
1908
image beyond the point at which the output processing is reading data out
1909
again. If the input arrives fast enough, it may "wrap around" the buffer to
1910
the point where the input is more than one whole scan ahead of the output.
1911
If the output processing simply proceeds through its display pass without
1912
paying attention to the input, the effect seen on-screen is that the lower
1913
part of the image is one or more scans better in quality than the upper part.
1914
Then, when the next output scan is started, you have a choice of what target
1915
scan number to use. The recommended choice is to use the current input scan
1916
number at that time, which implies that you've skipped the output scans
1917
corresponding to the input scans that were completed while you processed the
1918
previous output scan. In this way, the decoder automatically adapts its
1919
speed to the arriving data, by skipping output scans as necessary to keep up
1920
with the arriving data.
1922
When using this strategy, you'll want to be sure that you perform a final
1923
output pass after receiving all the data; otherwise your last display may not
1924
be full quality across the whole screen. So the right outer loop logic is
1925
something like this:
1927
absorb any waiting input by calling jpeg_consume_input()
1928
final_pass = jpeg_input_complete(&cinfo);
1929
adjust output decompression parameters if required
1930
jpeg_start_output(&cinfo, cinfo.input_scan_number);
1932
jpeg_finish_output()
1933
} while (! final_pass);
1934
rather than quitting as soon as jpeg_input_complete() returns TRUE. This
1935
arrangement makes it simple to use higher-quality decoding parameters
1936
for the final pass. But if you don't want to use special parameters for
1937
the final pass, the right loop logic is like this:
1939
absorb any waiting input by calling jpeg_consume_input()
1940
jpeg_start_output(&cinfo, cinfo.input_scan_number);
1942
jpeg_finish_output()
1943
if (jpeg_input_complete(&cinfo) &&
1944
cinfo.input_scan_number == cinfo.output_scan_number)
1947
In this case you don't need to know in advance whether an output pass is to
1948
be the last one, so it's not necessary to have reached EOF before starting
1949
the final output pass; rather, what you want to test is whether the output
1950
pass was performed in sync with the final input scan. This form of the loop
1951
will avoid an extra output pass whenever the decoder is able (or nearly able)
1952
to keep up with the incoming data.
1954
When the data transmission speed is high, you might begin a display pass,
1955
then find that much or all of the file has arrived before you can complete
1956
the pass. (You can detect this by noting the JPEG_REACHED_EOI return code
1957
from jpeg_consume_input(), or equivalently by testing jpeg_input_complete().)
1958
In this situation you may wish to abort the current display pass and start a
1959
new one using the newly arrived information. To do so, just call
1960
jpeg_finish_output() and then start a new pass with jpeg_start_output().
1962
A variant strategy is to abort and restart display if more than one complete
1963
scan arrives during an output pass; this can be detected by noting
1964
JPEG_REACHED_SOS returns and/or examining cinfo.input_scan_number. This
1965
idea should be employed with caution, however, since the display process
1966
might never get to the bottom of the image before being aborted, resulting
1967
in the lower part of the screen being several passes worse than the upper.
1968
In most cases it's probably best to abort an output pass only if the whole
1969
file has arrived and you want to begin the final output pass immediately.
1971
When receiving data across a communication link, we recommend always using
1972
the current input scan number for the output target scan number; if a
1973
higher-quality final pass is to be done, it should be started (aborting any
1974
incomplete output pass) as soon as the end of file is received. However,
1975
many other strategies are possible. For example, the application can examine
1976
the parameters of the current input scan and decide whether to display it or
1977
not. If the scan contains only chroma data, one might choose not to use it
1978
as the target scan, expecting that the scan will be small and will arrive
1979
quickly. To skip to the next scan, call jpeg_consume_input() until it
1980
returns JPEG_REACHED_SOS or JPEG_REACHED_EOI. Or just use the next higher
1981
number as the target scan for jpeg_start_output(); but that method doesn't
1982
let you inspect the next scan's parameters before deciding to display it.
1985
In buffered-image mode, jpeg_start_decompress() never performs input and
1986
thus never suspends. An application that uses input suspension with
1987
buffered-image mode must be prepared for suspension returns from these
1989
* jpeg_start_output() performs input only if you request 2-pass quantization
1990
and the target scan isn't fully read yet. (This is discussed below.)
1991
* jpeg_read_scanlines(), as always, returns the number of scanlines that it
1992
was able to produce before suspending.
1993
* jpeg_finish_output() will read any markers following the target scan,
1994
up to the end of the file or the SOS marker that begins another scan.
1995
(But it reads no input if jpeg_consume_input() has already reached the
1996
end of the file or a SOS marker beyond the target output scan.)
1997
* jpeg_finish_decompress() will read until the end of file, and thus can
1998
suspend if the end hasn't already been reached (as can be tested by
1999
calling jpeg_input_complete()).
2000
jpeg_start_output(), jpeg_finish_output(), and jpeg_finish_decompress()
2001
all return TRUE if they completed their tasks, FALSE if they had to suspend.
2002
In the event of a FALSE return, the application must load more input data
2003
and repeat the call. Applications that use non-suspending data sources need
2004
not check the return values of these three routines.
2007
It is possible to change decoding parameters between output passes in the
2008
buffered-image mode. The decoder library currently supports only very
2009
limited changes of parameters. ONLY THE FOLLOWING parameter changes are
2010
allowed after jpeg_start_decompress() is called:
2011
* dct_method can be changed before each call to jpeg_start_output().
2012
For example, one could use a fast DCT method for early scans, changing
2013
to a higher quality method for the final scan.
2014
* dither_mode can be changed before each call to jpeg_start_output();
2015
of course this has no impact if not using color quantization. Typically
2016
one would use ordered dither for initial passes, then switch to
2017
Floyd-Steinberg dither for the final pass. Caution: changing dither mode
2018
can cause more memory to be allocated by the library. Although the amount
2019
of memory involved is not large (a scanline or so), it may cause the
2020
initial max_memory_to_use specification to be exceeded, which in the worst
2021
case would result in an out-of-memory failure.
2022
* do_block_smoothing can be changed before each call to jpeg_start_output().
2023
This setting is relevant only when decoding a progressive JPEG image.
2024
During the first DC-only scan, block smoothing provides a very "fuzzy" look
2025
instead of the very "blocky" look seen without it; which is better seems a
2026
matter of personal taste. But block smoothing is nearly always a win
2027
during later stages, especially when decoding a successive-approximation
2028
image: smoothing helps to hide the slight blockiness that otherwise shows
2029
up on smooth gradients until the lowest coefficient bits are sent.
2030
* Color quantization mode can be changed under the rules described below.
2031
You *cannot* change between full-color and quantized output (because that
2032
would alter the required I/O buffer sizes), but you can change which
2033
quantization method is used.
2035
When generating color-quantized output, changing quantization method is a
2036
very useful way of switching between high-speed and high-quality display.
2037
The library allows you to change among its three quantization methods:
2038
1. Single-pass quantization to a fixed color cube.
2039
Selected by cinfo.two_pass_quantize = FALSE and cinfo.colormap = NULL.
2040
2. Single-pass quantization to an application-supplied colormap.
2041
Selected by setting cinfo.colormap to point to the colormap (the value of
2042
two_pass_quantize is ignored); also set cinfo.actual_number_of_colors.
2043
3. Two-pass quantization to a colormap chosen specifically for the image.
2044
Selected by cinfo.two_pass_quantize = TRUE and cinfo.colormap = NULL.
2045
(This is the default setting selected by jpeg_read_header, but it is
2046
probably NOT what you want for the first pass of progressive display!)
2047
These methods offer successively better quality and lesser speed. However,
2048
only the first method is available for quantizing in non-RGB color spaces.
2050
IMPORTANT: because the different quantizer methods have very different
2051
working-storage requirements, the library requires you to indicate which
2052
one(s) you intend to use before you call jpeg_start_decompress(). (If we did
2053
not require this, the max_memory_to_use setting would be a complete fiction.)
2054
You do this by setting one or more of these three cinfo fields to TRUE:
2055
enable_1pass_quant Fixed color cube colormap
2056
enable_external_quant Externally-supplied colormap
2057
enable_2pass_quant Two-pass custom colormap
2058
All three are initialized FALSE by jpeg_read_header(). But
2059
jpeg_start_decompress() automatically sets TRUE the one selected by the
2060
current two_pass_quantize and colormap settings, so you only need to set the
2061
enable flags for any other quantization methods you plan to change to later.
2063
After setting the enable flags correctly at jpeg_start_decompress() time, you
2064
can change to any enabled quantization method by setting two_pass_quantize
2065
and colormap properly just before calling jpeg_start_output(). The following
2066
special rules apply:
2067
1. You must explicitly set cinfo.colormap to NULL when switching to 1-pass
2068
or 2-pass mode from a different mode, or when you want the 2-pass
2069
quantizer to be re-run to generate a new colormap.
2070
2. To switch to an external colormap, or to change to a different external
2071
colormap than was used on the prior pass, you must call
2072
jpeg_new_colormap() after setting cinfo.colormap.
2073
NOTE: if you want to use the same colormap as was used in the prior pass,
2074
you should not do either of these things. This will save some nontrivial
2076
(These requirements exist because cinfo.colormap will always be non-NULL
2077
after completing a prior output pass, since both the 1-pass and 2-pass
2078
quantizers set it to point to their output colormaps. Thus you have to
2079
do one of these two things to notify the library that something has changed.
2080
Yup, it's a bit klugy, but it's necessary to do it this way for backwards
2083
Note that in buffered-image mode, the library generates any requested colormap
2084
during jpeg_start_output(), not during jpeg_start_decompress().
2086
When using two-pass quantization, jpeg_start_output() makes a pass over the
2087
buffered image to determine the optimum color map; it therefore may take a
2088
significant amount of time, whereas ordinarily it does little work. The
2089
progress monitor hook is called during this pass, if defined. It is also
2090
important to realize that if the specified target scan number is greater than
2091
or equal to the current input scan number, jpeg_start_output() will attempt
2092
to consume input as it makes this pass. If you use a suspending data source,
2093
you need to check for a FALSE return from jpeg_start_output() under these
2094
conditions. The combination of 2-pass quantization and a not-yet-fully-read
2095
target scan is the only case in which jpeg_start_output() will consume input.
2098
Application authors who support buffered-image mode may be tempted to use it
2099
for all JPEG images, even single-scan ones. This will work, but it is
2100
inefficient: there is no need to create an image-sized coefficient buffer for
2101
single-scan images. Requesting buffered-image mode for such an image wastes
2102
memory. Worse, it can cost time on large images, since the buffered data has
2103
to be swapped out or written to a temporary file. If you are concerned about
2104
maximum performance on baseline JPEG files, you should use buffered-image
2105
mode only when the incoming file actually has multiple scans. This can be
2106
tested by calling jpeg_has_multiple_scans(), which will return a correct
2107
result at any time after jpeg_read_header() completes.
2109
It is also worth noting that when you use jpeg_consume_input() to let input
2110
processing get ahead of output processing, the resulting pattern of access to
2111
the coefficient buffer is quite nonsequential. It's best to use the memory
2112
manager jmemnobs.c if you can (ie, if you have enough real or virtual main
2113
memory). If not, at least make sure that max_memory_to_use is set as high as
2114
possible. If the JPEG memory manager has to use a temporary file, you will
2115
probably see a lot of disk traffic and poor performance. (This could be
2116
improved with additional work on the memory manager, but we haven't gotten
2119
In some applications it may be convenient to use jpeg_consume_input() for all
2120
input processing, including reading the initial markers; that is, you may
2121
wish to call jpeg_consume_input() instead of jpeg_read_header() during
2122
startup. This works, but note that you must check for JPEG_REACHED_SOS and
2123
JPEG_REACHED_EOI return codes as the equivalent of jpeg_read_header's codes.
2124
Once the first SOS marker has been reached, you must call
2125
jpeg_start_decompress() before jpeg_consume_input() will consume more input;
2126
it'll just keep returning JPEG_REACHED_SOS until you do. If you read a
2127
tables-only file this way, jpeg_consume_input() will return JPEG_REACHED_EOI
2128
without ever returning JPEG_REACHED_SOS; be sure to check for this case.
2129
If this happens, the decompressor will not read any more input until you call
2130
jpeg_abort() to reset it. It is OK to call jpeg_consume_input() even when not
2131
using buffered-image mode, but in that case it's basically a no-op after the
2132
initial markers have been read: it will just return JPEG_SUSPENDED.
2135
Abbreviated datastreams and multiple images
2136
-------------------------------------------
2138
A JPEG compression or decompression object can be reused to process multiple
2139
images. This saves a small amount of time per image by eliminating the
2140
"create" and "destroy" operations, but that isn't the real purpose of the
2141
feature. Rather, reuse of an object provides support for abbreviated JPEG
2142
datastreams. Object reuse can also simplify processing a series of images in
2143
a single input or output file. This section explains these features.
2145
A JPEG file normally contains several hundred bytes worth of quantization
2146
and Huffman tables. In a situation where many images will be stored or
2147
transmitted with identical tables, this may represent an annoying overhead.
2148
The JPEG standard therefore permits tables to be omitted. The standard
2149
defines three classes of JPEG datastreams:
2150
* "Interchange" datastreams contain an image and all tables needed to decode
2151
the image. These are the usual kind of JPEG file.
2152
* "Abbreviated image" datastreams contain an image, but are missing some or
2153
all of the tables needed to decode that image.
2154
* "Abbreviated table specification" (henceforth "tables-only") datastreams
2155
contain only table specifications.
2156
To decode an abbreviated image, it is necessary to load the missing table(s)
2157
into the decoder beforehand. This can be accomplished by reading a separate
2158
tables-only file. A variant scheme uses a series of images in which the first
2159
image is an interchange (complete) datastream, while subsequent ones are
2160
abbreviated and rely on the tables loaded by the first image. It is assumed
2161
that once the decoder has read a table, it will remember that table until a
2162
new definition for the same table number is encountered.
2164
It is the application designer's responsibility to figure out how to associate
2165
the correct tables with an abbreviated image. While abbreviated datastreams
2166
can be useful in a closed environment, their use is strongly discouraged in
2167
any situation where data exchange with other applications might be needed.
2170
The JPEG library provides support for reading and writing any combination of
2171
tables-only datastreams and abbreviated images. In both compression and
2172
decompression objects, a quantization or Huffman table will be retained for
2173
the lifetime of the object, unless it is overwritten by a new table definition.
2176
To create abbreviated image datastreams, it is only necessary to tell the
2177
compressor not to emit some or all of the tables it is using. Each
2178
quantization and Huffman table struct contains a boolean field "sent_table",
2179
which normally is initialized to FALSE. For each table used by the image, the
2180
header-writing process emits the table and sets sent_table = TRUE unless it is
2181
already TRUE. (In normal usage, this prevents outputting the same table
2182
definition multiple times, as would otherwise occur because the chroma
2183
components typically share tables.) Thus, setting this field to TRUE before
2184
calling jpeg_start_compress() will prevent the table from being written at
2187
If you want to create a "pure" abbreviated image file containing no tables,
2188
just call "jpeg_suppress_tables(&cinfo, TRUE)" after constructing all the
2189
tables. If you want to emit some but not all tables, you'll need to set the
2190
individual sent_table fields directly.
2192
To create an abbreviated image, you must also call jpeg_start_compress()
2193
with a second parameter of FALSE, not TRUE. Otherwise jpeg_start_compress()
2194
will force all the sent_table fields to FALSE. (This is a safety feature to
2195
prevent abbreviated images from being created accidentally.)
2197
To create a tables-only file, perform the same parameter setup that you
2198
normally would, but instead of calling jpeg_start_compress() and so on, call
2199
jpeg_write_tables(&cinfo). This will write an abbreviated datastream
2200
containing only SOI, DQT and/or DHT markers, and EOI. All the quantization
2201
and Huffman tables that are currently defined in the compression object will
2202
be emitted unless their sent_tables flag is already TRUE, and then all the
2203
sent_tables flags will be set TRUE.
2205
A sure-fire way to create matching tables-only and abbreviated image files
2206
is to proceed as follows:
2208
create JPEG compression object
2210
set destination to tables-only file
2211
jpeg_write_tables(&cinfo);
2212
set destination to image file
2213
jpeg_start_compress(&cinfo, FALSE);
2215
jpeg_finish_compress(&cinfo);
2217
Since the JPEG parameters are not altered between writing the table file and
2218
the abbreviated image file, the same tables are sure to be used. Of course,
2219
you can repeat the jpeg_start_compress() ... jpeg_finish_compress() sequence
2220
many times to produce many abbreviated image files matching the table file.
2222
You cannot suppress output of the computed Huffman tables when Huffman
2223
optimization is selected. (If you could, there'd be no way to decode the
2224
image...) Generally, you don't want to set optimize_coding = TRUE when
2225
you are trying to produce abbreviated files.
2227
In some cases you might want to compress an image using tables which are
2228
not stored in the application, but are defined in an interchange or
2229
tables-only file readable by the application. This can be done by setting up
2230
a JPEG decompression object to read the specification file, then copying the
2231
tables into your compression object. See jpeg_copy_critical_parameters()
2232
for an example of copying quantization tables.
2235
To read abbreviated image files, you simply need to load the proper tables
2236
into the decompression object before trying to read the abbreviated image.
2237
If the proper tables are stored in the application program, you can just
2238
allocate the table structs and fill in their contents directly. For example,
2239
to load a fixed quantization table into table slot "n":
2241
if (cinfo.quant_tbl_ptrs[n] == NULL)
2242
cinfo.quant_tbl_ptrs[n] = jpeg_alloc_quant_table((j_common_ptr) &cinfo);
2243
quant_ptr = cinfo.quant_tbl_ptrs[n]; /* quant_ptr is JQUANT_TBL* */
2244
for (i = 0; i < 64; i++) {
2245
/* Qtable[] is desired quantization table, in natural array order */
2246
quant_ptr->quantval[i] = Qtable[i];
2249
Code to load a fixed Huffman table is typically (for AC table "n"):
2251
if (cinfo.ac_huff_tbl_ptrs[n] == NULL)
2252
cinfo.ac_huff_tbl_ptrs[n] = jpeg_alloc_huff_table((j_common_ptr) &cinfo);
2253
huff_ptr = cinfo.ac_huff_tbl_ptrs[n]; /* huff_ptr is JHUFF_TBL* */
2254
for (i = 1; i <= 16; i++) {
2255
/* counts[i] is number of Huffman codes of length i bits, i=1..16 */
2256
huff_ptr->bits[i] = counts[i];
2258
for (i = 0; i < 256; i++) {
2259
/* symbols[] is the list of Huffman symbols, in code-length order */
2260
huff_ptr->huffval[i] = symbols[i];
2263
(Note that trying to set cinfo.quant_tbl_ptrs[n] to point directly at a
2264
constant JQUANT_TBL object is not safe. If the incoming file happened to
2265
contain a quantization table definition, your master table would get
2266
overwritten! Instead allocate a working table copy and copy the master table
2267
into it, as illustrated above. Ditto for Huffman tables, of course.)
2269
You might want to read the tables from a tables-only file, rather than
2270
hard-wiring them into your application. The jpeg_read_header() call is
2271
sufficient to read a tables-only file. You must pass a second parameter of
2272
FALSE to indicate that you do not require an image to be present. Thus, the
2275
create JPEG decompression object
2276
set source to tables-only file
2277
jpeg_read_header(&cinfo, FALSE);
2278
set source to abbreviated image file
2279
jpeg_read_header(&cinfo, TRUE);
2280
set decompression parameters
2281
jpeg_start_decompress(&cinfo);
2283
jpeg_finish_decompress(&cinfo);
2285
In some cases, you may want to read a file without knowing whether it contains
2286
an image or just tables. In that case, pass FALSE and check the return value
2287
from jpeg_read_header(): it will be JPEG_HEADER_OK if an image was found,
2288
JPEG_HEADER_TABLES_ONLY if only tables were found. (A third return value,
2289
JPEG_SUSPENDED, is possible when using a suspending data source manager.)
2290
Note that jpeg_read_header() will not complain if you read an abbreviated
2291
image for which you haven't loaded the missing tables; the missing-table check
2292
occurs later, in jpeg_start_decompress().
2295
It is possible to read a series of images from a single source file by
2296
repeating the jpeg_read_header() ... jpeg_finish_decompress() sequence,
2297
without releasing/recreating the JPEG object or the data source module.
2298
(If you did reinitialize, any partial bufferload left in the data source
2299
buffer at the end of one image would be discarded, causing you to lose the
2300
start of the next image.) When you use this method, stored tables are
2301
automatically carried forward, so some of the images can be abbreviated images
2302
that depend on tables from earlier images.
2304
If you intend to write a series of images into a single destination file,
2305
you might want to make a specialized data destination module that doesn't
2306
flush the output buffer at term_destination() time. This would speed things
2307
up by some trifling amount. Of course, you'd need to remember to flush the
2308
buffer after the last image. You can make the later images be abbreviated
2309
ones by passing FALSE to jpeg_start_compress().
2315
Some applications may need to insert or extract special data in the JPEG
2316
datastream. The JPEG standard provides marker types "COM" (comment) and
2317
"APP0" through "APP15" (application) to hold application-specific data.
2318
Unfortunately, the use of these markers is not specified by the standard.
2319
COM markers are fairly widely used to hold user-supplied text. The JFIF file
2320
format spec uses APP0 markers with specified initial strings to hold certain
2321
data. Adobe applications use APP14 markers beginning with the string "Adobe"
2322
for miscellaneous data. Other APPn markers are rarely seen, but might
2323
contain almost anything.
2325
If you wish to store user-supplied text, we recommend you use COM markers
2326
and place readable 7-bit ASCII text in them. Newline conventions are not
2327
standardized --- expect to find LF (Unix style), CR/LF (DOS style), or CR
2328
(Mac style). A robust COM reader should be able to cope with random binary
2329
garbage, including nulls, since some applications generate COM markers
2330
containing non-ASCII junk. (But yours should not be one of them.)
2332
For program-supplied data, use an APPn marker, and be sure to begin it with an
2333
identifying string so that you can tell whether the marker is actually yours.
2334
It's probably best to avoid using APP0 or APP14 for any private markers.
2335
(NOTE: the upcoming SPIFF standard will use APP8 markers; we recommend you
2336
not use APP8 markers for any private purposes, either.)
2338
Keep in mind that at most 65533 bytes can be put into one marker, but you
2339
can have as many markers as you like.
2341
By default, the IJG compression library will write a JFIF APP0 marker if the
2342
selected JPEG colorspace is grayscale or YCbCr, or an Adobe APP14 marker if
2343
the selected colorspace is RGB, CMYK, or YCCK. You can disable this, but
2344
we don't recommend it. The decompression library will recognize JFIF and
2345
Adobe markers and will set the JPEG colorspace properly when one is found.
2348
You can write special markers immediately following the datastream header by
2349
calling jpeg_write_marker() after jpeg_start_compress() and before the first
2350
call to jpeg_write_scanlines(). When you do this, the markers appear after
2351
the SOI and the JFIF APP0 and Adobe APP14 markers (if written), but before
2352
all else. Specify the marker type parameter as "JPEG_COM" for COM or
2353
"JPEG_APP0 + n" for APPn. (Actually, jpeg_write_marker will let you write
2354
any marker type, but we don't recommend writing any other kinds of marker.)
2355
For example, to write a user comment string pointed to by comment_text:
2356
jpeg_write_marker(cinfo, JPEG_COM, comment_text, strlen(comment_text));
2358
If it's not convenient to store all the marker data in memory at once,
2359
you can instead call jpeg_write_m_header() followed by multiple calls to
2360
jpeg_write_m_byte(). If you do it this way, it's your responsibility to
2361
call jpeg_write_m_byte() exactly the number of times given in the length
2362
parameter to jpeg_write_m_header(). (This method lets you empty the
2363
output buffer partway through a marker, which might be important when
2364
using a suspending data destination module. In any case, if you are using
2365
a suspending destination, you should flush its buffer after inserting
2366
any special markers. See "I/O suspension".)
2368
Or, if you prefer to synthesize the marker byte sequence yourself,
2369
you can just cram it straight into the data destination module.
2371
If you are writing JFIF 1.02 extension markers (thumbnail images), don't
2372
forget to set cinfo.JFIF_minor_version = 2 so that the encoder will write the
2373
correct JFIF version number in the JFIF header marker. The library's default
2374
is to write version 1.01, but that's wrong if you insert any 1.02 extension
2375
markers. (We could probably get away with just defaulting to 1.02, but there
2376
used to be broken decoders that would complain about unknown minor version
2377
numbers. To reduce compatibility risks it's safest not to write 1.02 unless
2378
you are actually using 1.02 extensions.)
2381
When reading, two methods of handling special markers are available:
2382
1. You can ask the library to save the contents of COM and/or APPn markers
2383
into memory, and then examine them at your leisure afterwards.
2384
2. You can supply your own routine to process COM and/or APPn markers
2385
on-the-fly as they are read.
2386
The first method is simpler to use, especially if you are using a suspending
2387
data source; writing a marker processor that copes with input suspension is
2388
not easy (consider what happens if the marker is longer than your available
2389
input buffer). However, the second method conserves memory since the marker
2390
data need not be kept around after it's been processed.
2392
For either method, you'd normally set up marker handling after creating a
2393
decompression object and before calling jpeg_read_header(), because the
2394
markers of interest will typically be near the head of the file and so will
2395
be scanned by jpeg_read_header. Once you've established a marker handling
2396
method, it will be used for the life of that decompression object
2397
(potentially many datastreams), unless you change it. Marker handling is
2398
determined separately for COM markers and for each APPn marker code.
2401
To save the contents of special markers in memory, call
2402
jpeg_save_markers(cinfo, marker_code, length_limit)
2403
where marker_code is the marker type to save, JPEG_COM or JPEG_APP0+n.
2404
(To arrange to save all the special marker types, you need to call this
2405
routine 17 times, for COM and APP0-APP15.) If the incoming marker is longer
2406
than length_limit data bytes, only length_limit bytes will be saved; this
2407
parameter allows you to avoid chewing up memory when you only need to see the
2408
first few bytes of a potentially large marker. If you want to save all the
2409
data, set length_limit to 0xFFFF; that is enough since marker lengths are only
2410
16 bits. As a special case, setting length_limit to 0 prevents that marker
2411
type from being saved at all. (That is the default behavior, in fact.)
2413
After jpeg_read_header() completes, you can examine the special markers by
2414
following the cinfo->marker_list pointer chain. All the special markers in
2415
the file appear in this list, in order of their occurrence in the file (but
2416
omitting any markers of types you didn't ask for). Both the original data
2417
length and the saved data length are recorded for each list entry; the latter
2418
will not exceed length_limit for the particular marker type. Note that these
2419
lengths exclude the marker length word, whereas the stored representation
2420
within the JPEG file includes it. (Hence the maximum data length is really
2423
It is possible that additional special markers appear in the file beyond the
2424
SOS marker at which jpeg_read_header stops; if so, the marker list will be
2425
extended during reading of the rest of the file. This is not expected to be
2426
common, however. If you are short on memory you may want to reset the length
2427
limit to zero for all marker types after finishing jpeg_read_header, to
2428
ensure that the max_memory_to_use setting cannot be exceeded due to addition
2431
The marker list remains stored until you call jpeg_finish_decompress or
2432
jpeg_abort, at which point the memory is freed and the list is set to empty.
2433
(jpeg_destroy also releases the storage, of course.)
2435
Note that the library is internally interested in APP0 and APP14 markers;
2436
if you try to set a small nonzero length limit on these types, the library
2437
will silently force the length up to the minimum it wants. (But you can set
2438
a zero length limit to prevent them from being saved at all.) Also, in a
2439
16-bit environment, the maximum length limit may be constrained to less than
2440
65533 by malloc() limitations. It is therefore best not to assume that the
2441
effective length limit is exactly what you set it to be.
2444
If you want to supply your own marker-reading routine, you do it by calling
2445
jpeg_set_marker_processor(). A marker processor routine must have the
2447
boolean jpeg_marker_parser_method (j_decompress_ptr cinfo)
2448
Although the marker code is not explicitly passed, the routine can find it
2449
in cinfo->unread_marker. At the time of call, the marker proper has been
2450
read from the data source module. The processor routine is responsible for
2451
reading the marker length word and the remaining parameter bytes, if any.
2452
Return TRUE to indicate success. (FALSE should be returned only if you are
2453
using a suspending data source and it tells you to suspend. See the standard
2454
marker processors in jdmarker.c for appropriate coding methods if you need to
2455
use a suspending data source.)
2457
If you override the default APP0 or APP14 processors, it is up to you to
2458
recognize JFIF and Adobe markers if you want colorspace recognition to occur
2459
properly. We recommend copying and extending the default processors if you
2460
want to do that. (A better idea is to save these marker types for later
2461
examination by calling jpeg_save_markers(); that method doesn't interfere
2462
with the library's own processing of these markers.)
2464
jpeg_set_marker_processor() and jpeg_save_markers() are mutually exclusive
2465
--- if you call one it overrides any previous call to the other, for the
2466
particular marker type specified.
2468
A simple example of an external COM processor can be found in djpeg.c.
2469
Also, see jpegtran.c for an example of using jpeg_save_markers.
2472
Raw (downsampled) image data
2473
----------------------------
2475
Some applications need to supply already-downsampled image data to the JPEG
2476
compressor, or to receive raw downsampled data from the decompressor. The
2477
library supports this requirement by allowing the application to write or
2478
read raw data, bypassing the normal preprocessing or postprocessing steps.
2479
The interface is different from the standard one and is somewhat harder to
2480
use. If your interest is merely in bypassing color conversion, we recommend
2481
that you use the standard interface and simply set jpeg_color_space =
2482
in_color_space (or jpeg_color_space = out_color_space for decompression).
2483
The mechanism described in this section is necessary only to supply or
2484
receive downsampled image data, in which not all components have the same
2488
To compress raw data, you must supply the data in the colorspace to be used
2489
in the JPEG file (please read the earlier section on Special color spaces)
2490
and downsampled to the sampling factors specified in the JPEG parameters.
2491
You must supply the data in the format used internally by the JPEG library,
2492
namely a JSAMPIMAGE array. This is an array of pointers to two-dimensional
2493
arrays, each of type JSAMPARRAY. Each 2-D array holds the values for one
2494
color component. This structure is necessary since the components are of
2495
different sizes. If the image dimensions are not a multiple of the MCU size,
2496
you must also pad the data correctly (usually, this is done by replicating
2497
the last column and/or row). The data must be padded to a multiple of a DCT
2498
block in each component: that is, each downsampled row must contain a
2499
multiple of 8 valid samples, and there must be a multiple of 8 sample rows
2500
for each component. (For applications such as conversion of digital TV
2501
images, the standard image size is usually a multiple of the DCT block size,
2502
so that no padding need actually be done.)
2504
The procedure for compression of raw data is basically the same as normal
2505
compression, except that you call jpeg_write_raw_data() in place of
2506
jpeg_write_scanlines(). Before calling jpeg_start_compress(), you must do
2508
* Set cinfo->raw_data_in to TRUE. (It is set FALSE by jpeg_set_defaults().)
2509
This notifies the library that you will be supplying raw data.
2510
* Ensure jpeg_color_space is correct --- an explicit jpeg_set_colorspace()
2511
call is a good idea. Note that since color conversion is bypassed,
2512
in_color_space is ignored, except that jpeg_set_defaults() uses it to
2513
choose the default jpeg_color_space setting.
2514
* Ensure the sampling factors, cinfo->comp_info[i].h_samp_factor and
2515
cinfo->comp_info[i].v_samp_factor, are correct. Since these indicate the
2516
dimensions of the data you are supplying, it's wise to set them
2517
explicitly, rather than assuming the library's defaults are what you want.
2519
To pass raw data to the library, call jpeg_write_raw_data() in place of
2520
jpeg_write_scanlines(). The two routines work similarly except that
2521
jpeg_write_raw_data takes a JSAMPIMAGE data array rather than JSAMPARRAY.
2522
The scanlines count passed to and returned from jpeg_write_raw_data is
2523
measured in terms of the component with the largest v_samp_factor.
2525
jpeg_write_raw_data() processes one MCU row per call, which is to say
2526
v_samp_factor*DCTSIZE sample rows of each component. The passed num_lines
2527
value must be at least max_v_samp_factor*DCTSIZE, and the return value will
2528
be exactly that amount (or possibly some multiple of that amount, in future
2529
library versions). This is true even on the last call at the bottom of the
2530
image; don't forget to pad your data as necessary.
2532
The required dimensions of the supplied data can be computed for each
2534
cinfo->comp_info[i].width_in_blocks*DCTSIZE samples per row
2535
cinfo->comp_info[i].height_in_blocks*DCTSIZE rows in image
2536
after jpeg_start_compress() has initialized those fields. If the valid data
2537
is smaller than this, it must be padded appropriately. For some sampling
2538
factors and image sizes, additional dummy DCT blocks are inserted to make
2539
the image a multiple of the MCU dimensions. The library creates such dummy
2540
blocks itself; it does not read them from your supplied data. Therefore you
2541
need never pad by more than DCTSIZE samples. An example may help here.
2542
Assume 2h2v downsampling of YCbCr data, that is
2543
cinfo->comp_info[0].h_samp_factor = 2 for Y
2544
cinfo->comp_info[0].v_samp_factor = 2
2545
cinfo->comp_info[1].h_samp_factor = 1 for Cb
2546
cinfo->comp_info[1].v_samp_factor = 1
2547
cinfo->comp_info[2].h_samp_factor = 1 for Cr
2548
cinfo->comp_info[2].v_samp_factor = 1
2549
and suppose that the nominal image dimensions (cinfo->image_width and
2550
cinfo->image_height) are 101x101 pixels. Then jpeg_start_compress() will
2551
compute downsampled_width = 101 and width_in_blocks = 13 for Y,
2552
downsampled_width = 51 and width_in_blocks = 7 for Cb and Cr (and the same
2553
for the height fields). You must pad the Y data to at least 13*8 = 104
2554
columns and rows, the Cb/Cr data to at least 7*8 = 56 columns and rows. The
2555
MCU height is max_v_samp_factor = 2 DCT rows so you must pass at least 16
2556
scanlines on each call to jpeg_write_raw_data(), which is to say 16 actual
2557
sample rows of Y and 8 each of Cb and Cr. A total of 7 MCU rows are needed,
2558
so you must pass a total of 7*16 = 112 "scanlines". The last DCT block row
2559
of Y data is dummy, so it doesn't matter what you pass for it in the data
2560
arrays, but the scanlines count must total up to 112 so that all of the Cb
2561
and Cr data gets passed.
2563
Output suspension is supported with raw-data compression: if the data
2564
destination module suspends, jpeg_write_raw_data() will return 0.
2565
In this case the same data rows must be passed again on the next call.
2568
Decompression with raw data output implies bypassing all postprocessing:
2569
you cannot ask for rescaling or color quantization, for instance. More
2570
seriously, you must deal with the color space and sampling factors present in
2571
the incoming file. If your application only handles, say, 2h1v YCbCr data,
2572
you must check for and fail on other color spaces or other sampling factors.
2573
The library will not convert to a different color space for you.
2575
To obtain raw data output, set cinfo->raw_data_out = TRUE before
2576
jpeg_start_decompress() (it is set FALSE by jpeg_read_header()). Be sure to
2577
verify that the color space and sampling factors are ones you can handle.
2578
Then call jpeg_read_raw_data() in place of jpeg_read_scanlines(). The
2579
decompression process is otherwise the same as usual.
2581
jpeg_read_raw_data() returns one MCU row per call, and thus you must pass a
2582
buffer of at least max_v_samp_factor*DCTSIZE scanlines (scanline counting is
2583
the same as for raw-data compression). The buffer you pass must be large
2584
enough to hold the actual data plus padding to DCT-block boundaries. As with
2585
compression, any entirely dummy DCT blocks are not processed so you need not
2586
allocate space for them, but the total scanline count includes them. The
2587
above example of computing buffer dimensions for raw-data compression is
2588
equally valid for decompression.
2590
Input suspension is supported with raw-data decompression: if the data source
2591
module suspends, jpeg_read_raw_data() will return 0. You can also use
2592
buffered-image mode to read raw data in multiple passes.
2595
Really raw data: DCT coefficients
2596
---------------------------------
2598
It is possible to read or write the contents of a JPEG file as raw DCT
2599
coefficients. This facility is mainly intended for use in lossless
2600
transcoding between different JPEG file formats. Other possible applications
2601
include lossless cropping of a JPEG image, lossless reassembly of a
2602
multi-strip or multi-tile TIFF/JPEG file into a single JPEG datastream, etc.
2604
To read the contents of a JPEG file as DCT coefficients, open the file and do
2605
jpeg_read_header() as usual. But instead of calling jpeg_start_decompress()
2606
and jpeg_read_scanlines(), call jpeg_read_coefficients(). This will read the
2607
entire image into a set of virtual coefficient-block arrays, one array per
2608
component. The return value is a pointer to an array of virtual-array
2609
descriptors. Each virtual array can be accessed directly using the JPEG
2610
memory manager's access_virt_barray method (see Memory management, below,
2611
and also read structure.doc's discussion of virtual array handling). Or,
2612
for simple transcoding to a different JPEG file format, the array list can
2613
just be handed directly to jpeg_write_coefficients().
2615
Each block in the block arrays contains quantized coefficient values in
2616
normal array order (not JPEG zigzag order). The block arrays contain only
2617
DCT blocks containing real data; any entirely-dummy blocks added to fill out
2618
interleaved MCUs at the right or bottom edges of the image are discarded
2619
during reading and are not stored in the block arrays. (The size of each
2620
block array can be determined from the width_in_blocks and height_in_blocks
2621
fields of the component's comp_info entry.) This is also the data format
2622
expected by jpeg_write_coefficients().
2624
When you are done using the virtual arrays, call jpeg_finish_decompress()
2625
to release the array storage and return the decompression object to an idle
2626
state; or just call jpeg_destroy() if you don't need to reuse the object.
2628
If you use a suspending data source, jpeg_read_coefficients() will return
2629
NULL if it is forced to suspend; a non-NULL return value indicates successful
2630
completion. You need not test for a NULL return value when using a
2631
non-suspending data source.
2633
It is also possible to call jpeg_read_coefficients() to obtain access to the
2634
decoder's coefficient arrays during a normal decode cycle in buffered-image
2635
mode. This frammish might be useful for progressively displaying an incoming
2636
image and then re-encoding it without loss. To do this, decode in buffered-
2637
image mode as discussed previously, then call jpeg_read_coefficients() after
2638
the last jpeg_finish_output() call. The arrays will be available for your use
2639
until you call jpeg_finish_decompress().
2642
To write the contents of a JPEG file as DCT coefficients, you must provide
2643
the DCT coefficients stored in virtual block arrays. You can either pass
2644
block arrays read from an input JPEG file by jpeg_read_coefficients(), or
2645
allocate virtual arrays from the JPEG compression object and fill them
2646
yourself. In either case, jpeg_write_coefficients() is substituted for
2647
jpeg_start_compress() and jpeg_write_scanlines(). Thus the sequence is
2648
* Create compression object
2649
* Set all compression parameters as necessary
2650
* Request virtual arrays if needed
2651
* jpeg_write_coefficients()
2652
* jpeg_finish_compress()
2653
* Destroy or re-use compression object
2654
jpeg_write_coefficients() is passed a pointer to an array of virtual block
2655
array descriptors; the number of arrays is equal to cinfo.num_components.
2657
The virtual arrays need only have been requested, not realized, before
2658
jpeg_write_coefficients() is called. A side-effect of
2659
jpeg_write_coefficients() is to realize any virtual arrays that have been
2660
requested from the compression object's memory manager. Thus, when obtaining
2661
the virtual arrays from the compression object, you should fill the arrays
2662
after calling jpeg_write_coefficients(). The data is actually written out
2663
when you call jpeg_finish_compress(); jpeg_write_coefficients() only writes
2666
When writing raw DCT coefficients, it is crucial that the JPEG quantization
2667
tables and sampling factors match the way the data was encoded, or the
2668
resulting file will be invalid. For transcoding from an existing JPEG file,
2669
we recommend using jpeg_copy_critical_parameters(). This routine initializes
2670
all the compression parameters to default values (like jpeg_set_defaults()),
2671
then copies the critical information from a source decompression object.
2672
The decompression object should have just been used to read the entire
2673
JPEG input file --- that is, it should be awaiting jpeg_finish_decompress().
2675
jpeg_write_coefficients() marks all tables stored in the compression object
2676
as needing to be written to the output file (thus, it acts like
2677
jpeg_start_compress(cinfo, TRUE)). This is for safety's sake, to avoid
2678
emitting abbreviated JPEG files by accident. If you really want to emit an
2679
abbreviated JPEG file, call jpeg_suppress_tables(), or set the tables'
2680
individual sent_table flags, between calling jpeg_write_coefficients() and
2681
jpeg_finish_compress().
2687
Some applications may need to regain control from the JPEG library every so
2688
often. The typical use of this feature is to produce a percent-done bar or
2689
other progress display. (For a simple example, see cjpeg.c or djpeg.c.)
2690
Although you do get control back frequently during the data-transferring pass
2691
(the jpeg_read_scanlines or jpeg_write_scanlines loop), any additional passes
2692
will occur inside jpeg_finish_compress or jpeg_start_decompress; those
2693
routines may take a long time to execute, and you don't get control back
2694
until they are done.
2696
You can define a progress-monitor routine which will be called periodically
2697
by the library. No guarantees are made about how often this call will occur,
2698
so we don't recommend you use it for mouse tracking or anything like that.
2699
At present, a call will occur once per MCU row, scanline, or sample row
2700
group, whichever unit is convenient for the current processing mode; so the
2701
wider the image, the longer the time between calls. During the data
2702
transferring pass, only one call occurs per call of jpeg_read_scanlines or
2703
jpeg_write_scanlines, so don't pass a large number of scanlines at once if
2704
you want fine resolution in the progress count. (If you really need to use
2705
the callback mechanism for time-critical tasks like mouse tracking, you could
2706
insert additional calls inside some of the library's inner loops.)
2708
To establish a progress-monitor callback, create a struct jpeg_progress_mgr,
2709
fill in its progress_monitor field with a pointer to your callback routine,
2710
and set cinfo->progress to point to the struct. The callback will be called
2711
whenever cinfo->progress is non-NULL. (This pointer is set to NULL by
2712
jpeg_create_compress or jpeg_create_decompress; the library will not change
2713
it thereafter. So if you allocate dynamic storage for the progress struct,
2714
make sure it will live as long as the JPEG object does. Allocating from the
2715
JPEG memory manager with lifetime JPOOL_PERMANENT will work nicely.) You
2716
can use the same callback routine for both compression and decompression.
2718
The jpeg_progress_mgr struct contains four fields which are set by the library:
2719
long pass_counter; /* work units completed in this pass */
2720
long pass_limit; /* total number of work units in this pass */
2721
int completed_passes; /* passes completed so far */
2722
int total_passes; /* total number of passes expected */
2723
During any one pass, pass_counter increases from 0 up to (not including)
2724
pass_limit; the step size is usually but not necessarily 1. The pass_limit
2725
value may change from one pass to another. The expected total number of
2726
passes is in total_passes, and the number of passes already completed is in
2727
completed_passes. Thus the fraction of work completed may be estimated as
2728
completed_passes + (pass_counter/pass_limit)
2729
--------------------------------------------
2731
ignoring the fact that the passes may not be equal amounts of work.
2733
When decompressing, pass_limit can even change within a pass, because it
2734
depends on the number of scans in the JPEG file, which isn't always known in
2735
advance. The computed fraction-of-work-done may jump suddenly (if the library
2736
discovers it has overestimated the number of scans) or even decrease (in the
2737
opposite case). It is not wise to put great faith in the work estimate.
2739
When using the decompressor's buffered-image mode, the progress monitor work
2740
estimate is likely to be completely unhelpful, because the library has no way
2741
to know how many output passes will be demanded of it. Currently, the library
2742
sets total_passes based on the assumption that there will be one more output
2743
pass if the input file end hasn't yet been read (jpeg_input_complete() isn't
2744
TRUE), but no more output passes if the file end has been reached when the
2745
output pass is started. This means that total_passes will rise as additional
2746
output passes are requested. If you have a way of determining the input file
2747
size, estimating progress based on the fraction of the file that's been read
2748
will probably be more useful than using the library's value.
2754
This section covers some key facts about the JPEG library's built-in memory
2755
manager. For more info, please read structure.doc's section about the memory
2756
manager, and consult the source code if necessary.
2758
All memory and temporary file allocation within the library is done via the
2759
memory manager. If necessary, you can replace the "back end" of the memory
2760
manager to control allocation yourself (for example, if you don't want the
2761
library to use malloc() and free() for some reason).
2763
Some data is allocated "permanently" and will not be freed until the JPEG
2764
object is destroyed. Most data is allocated "per image" and is freed by
2765
jpeg_finish_compress, jpeg_finish_decompress, or jpeg_abort. You can call the
2766
memory manager yourself to allocate structures that will automatically be
2767
freed at these times. Typical code for this is
2768
ptr = (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, size);
2769
Use JPOOL_PERMANENT to get storage that lasts as long as the JPEG object.
2770
Use alloc_large instead of alloc_small for anything bigger than a few Kbytes.
2771
There are also alloc_sarray and alloc_barray routines that automatically
2772
build 2-D sample or block arrays.
2774
The library's minimum space requirements to process an image depend on the
2775
image's width, but not on its height, because the library ordinarily works
2776
with "strip" buffers that are as wide as the image but just a few rows high.
2777
Some operating modes (eg, two-pass color quantization) require full-image
2778
buffers. Such buffers are treated as "virtual arrays": only the current strip
2779
need be in memory, and the rest can be swapped out to a temporary file.
2781
If you use the simplest memory manager back end (jmemnobs.c), then no
2782
temporary files are used; virtual arrays are simply malloc()'d. Images bigger
2783
than memory can be processed only if your system supports virtual memory.
2784
The other memory manager back ends support temporary files of various flavors
2785
and thus work in machines without virtual memory. They may also be useful on
2786
Unix machines if you need to process images that exceed available swap space.
2788
When using temporary files, the library will make the in-memory buffers for
2789
its virtual arrays just big enough to stay within a "maximum memory" setting.
2790
Your application can set this limit by setting cinfo->mem->max_memory_to_use
2791
after creating the JPEG object. (Of course, there is still a minimum size for
2792
the buffers, so the max-memory setting is effective only if it is bigger than
2793
the minimum space needed.) If you allocate any large structures yourself, you
2794
must allocate them before jpeg_start_compress() or jpeg_start_decompress() in
2795
order to have them counted against the max memory limit. Also keep in mind
2796
that space allocated with alloc_small() is ignored, on the assumption that
2797
it's too small to be worth worrying about; so a reasonable safety margin
2798
should be left when setting max_memory_to_use.
2800
If you use the jmemname.c or jmemdos.c memory manager back end, it is
2801
important to clean up the JPEG object properly to ensure that the temporary
2802
files get deleted. (This is especially crucial with jmemdos.c, where the
2803
"temporary files" may be extended-memory segments; if they are not freed,
2804
DOS will require a reboot to recover the memory.) Thus, with these memory
2805
managers, it's a good idea to provide a signal handler that will trap any
2806
early exit from your program. The handler should call either jpeg_abort()
2807
or jpeg_destroy() for any active JPEG objects. A handler is not needed with
2808
jmemnobs.c, and shouldn't be necessary with jmemansi.c or jmemmac.c either,
2809
since the C library is supposed to take care of deleting files made with
2816
Working memory requirements while performing compression or decompression
2817
depend on image dimensions, image characteristics (such as colorspace and
2818
JPEG process), and operating mode (application-selected options).
2820
As of v6b, the decompressor requires:
2821
1. About 24K in more-or-less-fixed-size data. This varies a bit depending
2822
on operating mode and image characteristics (particularly color vs.
2823
grayscale), but it doesn't depend on image dimensions.
2824
2. Strip buffers (of size proportional to the image width) for IDCT and
2825
upsampling results. The worst case for commonly used sampling factors
2826
is about 34 bytes * width in pixels for a color image. A grayscale image
2827
only needs about 8 bytes per pixel column.
2828
3. A full-image DCT coefficient buffer is needed to decode a multi-scan JPEG
2829
file (including progressive JPEGs), or whenever you select buffered-image
2830
mode. This takes 2 bytes/coefficient. At typical 2x2 sampling, that's
2831
3 bytes per pixel for a color image. Worst case (1x1 sampling) requires
2832
6 bytes/pixel. For grayscale, figure 2 bytes/pixel.
2833
4. To perform 2-pass color quantization, the decompressor also needs a
2834
128K color lookup table and a full-image pixel buffer (3 bytes/pixel).
2835
This does not count any memory allocated by the application, such as a
2836
buffer to hold the final output image.
2838
The above figures are valid for 8-bit JPEG data precision and a machine with
2839
32-bit ints. For 12-bit JPEG data, double the size of the strip buffers and
2840
quantization pixel buffer. The "fixed-size" data will be somewhat smaller
2841
with 16-bit ints, larger with 64-bit ints. Also, CMYK or other unusual
2842
color spaces will require different amounts of space.
2844
The full-image coefficient and pixel buffers, if needed at all, do not
2845
have to be fully RAM resident; you can have the library use temporary
2846
files instead when the total memory usage would exceed a limit you set.
2847
(But if your OS supports virtual memory, it's probably better to just use
2848
jmemnobs and let the OS do the swapping.)
2850
The compressor's memory requirements are similar, except that it has no need
2851
for color quantization. Also, it needs a full-image DCT coefficient buffer
2852
if Huffman-table optimization is asked for, even if progressive mode is not
2855
If you need more detailed information about memory usage in a particular
2856
situation, you can enable the MEM_STATS code in jmemmgr.c.
2859
Library compile-time options
2860
----------------------------
2862
A number of compile-time options are available by modifying jmorecfg.h.
2864
The JPEG standard provides for both the baseline 8-bit DCT process and
2865
a 12-bit DCT process. The IJG code supports 12-bit lossy JPEG if you define
2866
BITS_IN_JSAMPLE as 12 rather than 8. Note that this causes JSAMPLE to be
2867
larger than a char, so it affects the surrounding application's image data.
2868
The sample applications cjpeg and djpeg can support 12-bit mode only for PPM
2869
and GIF file formats; you must disable the other file formats to compile a
2870
12-bit cjpeg or djpeg. (install.doc has more information about that.)
2871
At present, a 12-bit library can handle *only* 12-bit images, not both
2872
precisions. (If you need to include both 8- and 12-bit libraries in a single
2873
application, you could probably do it by defining NEED_SHORT_EXTERNAL_NAMES
2874
for just one of the copies. You'd have to access the 8-bit and 12-bit copies
2875
from separate application source files. This is untested ... if you try it,
2876
we'd like to hear whether it works!)
2878
Note that a 12-bit library always compresses in Huffman optimization mode,
2879
in order to generate valid Huffman tables. This is necessary because our
2880
default Huffman tables only cover 8-bit data. If you need to output 12-bit
2881
files in one pass, you'll have to supply suitable default Huffman tables.
2882
You may also want to supply your own DCT quantization tables; the existing
2883
quality-scaling code has been developed for 8-bit use, and probably doesn't
2884
generate especially good tables for 12-bit.
2886
The maximum number of components (color channels) in the image is determined
2887
by MAX_COMPONENTS. The JPEG standard allows up to 255 components, but we
2888
expect that few applications will need more than four or so.
2890
On machines with unusual data type sizes, you may be able to improve
2891
performance or reduce memory space by tweaking the various typedefs in
2892
jmorecfg.h. In particular, on some RISC CPUs, access to arrays of "short"s
2893
is quite slow; consider trading memory for speed by making JCOEF, INT16, and
2894
UINT16 be "int" or "unsigned int". UINT8 is also a candidate to become int.
2895
You probably don't want to make JSAMPLE be int unless you have lots of memory
2898
You can reduce the size of the library by compiling out various optional
2899
functions. To do this, undefine xxx_SUPPORTED symbols as necessary.
2901
You can also save a few K by not having text error messages in the library;
2902
the standard error message table occupies about 5Kb. This is particularly
2903
reasonable for embedded applications where there's no good way to display
2904
a message anyway. To do this, remove the creation of the message table
2905
(jpeg_std_message_table[]) from jerror.c, and alter format_message to do
2906
something reasonable without it. You could output the numeric value of the
2907
message code number, for example. If you do this, you can also save a couple
2908
more K by modifying the TRACEMSn() macros in jerror.h to expand to nothing;
2909
you don't need trace capability anyway, right?
2912
Portability considerations
2913
--------------------------
2915
The JPEG library has been written to be extremely portable; the sample
2916
applications cjpeg and djpeg are slightly less so. This section summarizes
2917
the design goals in this area. (If you encounter any bugs that cause the
2918
library to be less portable than is claimed here, we'd appreciate hearing
2921
The code works fine on ANSI C, C++, and pre-ANSI C compilers, using any of
2922
the popular system include file setups, and some not-so-popular ones too.
2923
See install.doc for configuration procedures.
2925
The code is not dependent on the exact sizes of the C data types. As
2926
distributed, we make the assumptions that
2927
char is at least 8 bits wide
2928
short is at least 16 bits wide
2929
int is at least 16 bits wide
2930
long is at least 32 bits wide
2931
(These are the minimum requirements of the ANSI C standard.) Wider types will
2932
work fine, although memory may be used inefficiently if char is much larger
2933
than 8 bits or short is much bigger than 16 bits. The code should work
2934
equally well with 16- or 32-bit ints.
2936
In a system where these assumptions are not met, you may be able to make the
2937
code work by modifying the typedefs in jmorecfg.h. However, you will probably
2938
have difficulty if int is less than 16 bits wide, since references to plain
2939
int abound in the code.
2941
char can be either signed or unsigned, although the code runs faster if an
2942
unsigned char type is available. If char is wider than 8 bits, you will need
2943
to redefine JOCTET and/or provide custom data source/destination managers so
2944
that JOCTET represents exactly 8 bits of data on external storage.
2946
The JPEG library proper does not assume ASCII representation of characters.
2947
But some of the image file I/O modules in cjpeg/djpeg do have ASCII
2948
dependencies in file-header manipulation; so does cjpeg's select_file_type()
2951
The JPEG library does not rely heavily on the C library. In particular, C
2952
stdio is used only by the data source/destination modules and the error
2953
handler, all of which are application-replaceable. (cjpeg/djpeg are more
2954
heavily dependent on stdio.) malloc and free are called only from the memory
2955
manager "back end" module, so you can use a different memory allocator by
2956
replacing that one file.
2958
The code generally assumes that C names must be unique in the first 15
2959
characters. However, global function names can be made unique in the
2960
first 6 characters by defining NEED_SHORT_EXTERNAL_NAMES.
2962
More info about porting the code may be gleaned by reading jconfig.doc,
2963
jmorecfg.h, and jinclude.h.
2966
Notes for MS-DOS implementors
2967
-----------------------------
2969
The IJG code is designed to work efficiently in 80x86 "small" or "medium"
2970
memory models (i.e., data pointers are 16 bits unless explicitly declared
2971
"far"; code pointers can be either size). You may be able to use small
2972
model to compile cjpeg or djpeg by itself, but you will probably have to use
2973
medium model for any larger application. This won't make much difference in
2974
performance. You *will* take a noticeable performance hit if you use a
2975
large-data memory model (perhaps 10%-25%), and you should avoid "huge" model
2978
The JPEG library typically needs 2Kb-3Kb of stack space. It will also
2979
malloc about 20K-30K of near heap space while executing (and lots of far
2980
heap, but that doesn't count in this calculation). This figure will vary
2981
depending on selected operating mode, and to a lesser extent on image size.
2982
There is also about 5Kb-6Kb of constant data which will be allocated in the
2983
near data segment (about 4Kb of this is the error message table).
2984
Thus you have perhaps 20K available for other modules' static data and near
2985
heap space before you need to go to a larger memory model. The C library's
2986
static data will account for several K of this, but that still leaves a good
2987
deal for your needs. (If you are tight on space, you could reduce the sizes
2988
of the I/O buffers allocated by jdatasrc.c and jdatadst.c, say from 4K to
2989
1K. Another possibility is to move the error message table to far memory;
2990
this should be doable with only localized hacking on jerror.c.)
2992
About 2K of the near heap space is "permanent" memory that will not be
2993
released until you destroy the JPEG object. This is only an issue if you
2994
save a JPEG object between compression or decompression operations.
2996
Far data space may also be a tight resource when you are dealing with large
2997
images. The most memory-intensive case is decompression with two-pass color
2998
quantization, or single-pass quantization to an externally supplied color
2999
map. This requires a 128Kb color lookup table plus strip buffers amounting
3000
to about 40 bytes per column for typical sampling ratios (eg, about 25600
3001
bytes for a 640-pixel-wide image). You may not be able to process wide
3002
images if you have large data structures of your own.
3004
Of course, all of these concerns vanish if you use a 32-bit flat-memory-model
3005
compiler, such as DJGPP or Watcom C. We highly recommend flat model if you
3006
can use it; the JPEG library is significantly faster in flat model.