Legacy ImageMagick Discussions Archive

multiple issues with settings.
First, i cannot set PNG format to PNG16.

Second, if i set PNG8 and i have a grayscale PNG 8 bit image, i cannot set colorspace to GRAYColorspace, i cannot set type GrayscaleType, and i cannot set quantizeColorSpace GRAYColorspace. It seems all those settings are getting overwritten by defaults in png.c

Third, If i have 8 bit PNG image, and i dont set magick PNG 8, but i spcificy image.write("output.PNG") it causes an error that it cannot write that type.

Fourth, When i write DIB format out in dib.c there is code that will reset bit depth to 24 bits, and convert image to RGB. Original image is 8 bit grayscale. Workaround was to set classType(PseudoClass) which works.

1. "PNG16" is not defined in IM; it would be ambiguous. It could mean 16-bit grayscale or 8-bit grayscale + 8-bit alpha.

1. Ok thanks. -> But it seems that logic to disambiguate and "streamline/improve" setup could escape that. Since we have an option not to provide image.magick("ImageFormat"), and IM should figure out which format the image is auto-magically when writing it out to file with file extension.