AWARE SYSTEMS
TIFF and LibTiff Mail List Archive

Thread

2021.06.15 15:33 "[Tiff] Caution if using mozjpeg and IJG libjpeg 9d", by Even Rouault
2021.06.15 15:47 "Re: [Tiff] Caution if using mozjpeg and IJG libjpeg 9d", by Roger Leigh
2021.06.15 15:53 "Re: [Tiff] Caution if using mozjpeg and IJG libjpeg 9d", by Even Rouault
2021.06.17 17:10 "Re: [Tiff] Caution if using mozjpeg and IJG libjpeg 9d", by Even Rouault
2021.06.17 17:37 "Re: [Tiff] Caution if using mozjpeg and IJG libjpeg 9d", by Roger Leigh
2021.06.15 15:50 "Re: [Tiff] Caution if using mozjpeg and IJG libjpeg 9d", by Sam James
2021.06.15 15:54 "Re: [Tiff] Caution if using mozjpeg and IJG libjpeg 9d", by Even Rouault

2021.06.15 15:53 "Re: [Tiff] Caution if using mozjpeg and IJG libjpeg 9d", by Even Rouault

Have these problems been reported to both upstream projects?

No, my spending of energy on those issues didn't go to that point :-)

For mozjpeg, generating progressive scans by default is in the design of mozjpeg, so probably not much to report about that, and the way we deal with that in libtiff is clean (changing their modified default to traditional libjpeg default). The issue with non-optimized Huffman coding looks more like a bug though, although in mozjpeg use case, generating non-optimized Huffman coding is at odds with their objectives, so not sure if they can/want fix that.

For the libjpeg 9d issue, that's probably more a bug, but only to be seen in the libtiff case that can generate codestreams with JPEG tables only, that is not seen when generated JFIF files.

Even

http://www.spatialys.com
My software is free, but my time generally not.