Closed dwillbarron closed 1 year ago
Thank you very much, you're totally right!
No problem! Thanks for sharing this library. I've been playing around with jxl a bit and this made it a lot quicker to get some experiments going.
Exactly to play around I build it and just as a proof of concept and to use it for batch encoding. I am glad if it helped!
Hi, sorry for bothering you if you're no longer working on this--I know jxl isn't a particularly hot topic at the moment.
Found a possible bug here. Code was applying --lossless_jpeg=0 when distance was above 1, but any distance above 0 is considered lossy by cjxl, so there was a blindspot for any distances between 0 and 1.
It works in my testing, but I'm not sure whether there was a purpose behind the original use of 1.0 here.