1113
1116
unsigned int scale_num, scale_denom
1114
1117
Scale the image by the fraction scale_num/scale_denom. Default is
1115
1118
1/1, or no scaling. Currently, the only supported scaling ratios
1116
are 1/1, 1/2, 1/4, and 1/8. (The library design allows for arbitrary
1117
1121
scaling ratios but this is not likely to be implemented any time soon.)
1118
1122
Smaller scaling ratios permit significantly faster decoding since
1119
1123
fewer pixels need be processed and a simpler IDCT method can be used.