diff options
| -rw-r--r-- | converter/other/pnmtopng.c | 8 |
1 files changed, 4 insertions, 4 deletions
diff --git a/converter/other/pnmtopng.c b/converter/other/pnmtopng.c index b339a1fe..1b806cca 100644 --- a/converter/other/pnmtopng.c +++ b/converter/other/pnmtopng.c @@ -1032,7 +1032,7 @@ findRedundantBits(FILE * const ifp, /*---------------------------------------------------------------------------- Find out if we can use just a subset of the bits from each input sample. Often, people create an image with e.g. 8 bit samples from - one that has e.g. only 4 bit samples by scaling by 256/16, which is + one that has e.g. only 4 bit samples by scaling by 255/15, which is the same as repeating the bits. E.g. 1011 becomes 10111011. We detect this case. We return as *meaningfulBitsP the minimum number of bits, starting from the least significant end, that contain @@ -2369,7 +2369,7 @@ convertpnm(struct cmdlineInfo const cmdline, */ unsigned int fulldepth; /* The total number of bits per pixel in the (uncompressed) png - raster, including all channels + raster, including all channels. */ pm_filepos rasterPos; /* file position in input image file of start of image (i.e. after @@ -2442,8 +2442,8 @@ convertpnm(struct cmdlineInfo const cmdline, to ppm_parsecolor() because ppm_parsecolor() does a cheap maxval scaling, and this is more precise. */ - PPM_DEPTH (transcolor, ppm_parsecolor(transstring2, maxmaxval), - maxmaxval, maxval); + PPM_DEPTH(transcolor, ppm_parsecolor(transstring2, maxmaxval), + maxmaxval, maxval); } if (cmdline.alpha) { pixel alpha_transcolor; |