You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I saw that gcc compiler for some compressors can drop performance significantly (especially for encoders by jxl devs, from 20% to several times on my hardware on some images, I was comparing gcc 12/13 and clang 15), it would be good to see a comparison with clang
Also about some additional good compression to speed modes from the benchmarks formats:
WebP_Lossless with z0 (fastest) preset
JXL_Lossless/l2 (effort 2 for libjxl) fast and good for mixed images
What about multi-threaded comparison?
Like for 8-threads (enough to show scalability), because not all formats/implementations scale equally or may lose efficiency compared to single-threaded
I saw that gcc compiler for some compressors can drop performance significantly (especially for encoders by jxl devs, from 20% to several times on my hardware on some images, I was comparing gcc 12/13 and clang 15), it would be good to see a comparison with clang
Also about some additional good compression to speed modes from the benchmarks formats:
What about Zstd for QOIR or any of the other relevant compressors?
As an example:
https://github.com/inikep/lzbench/blob/master/lzbench18_sorted.md
What about multi-threaded comparison?
Like for 8-threads (enough to show scalability), because not all formats/implementations scale equally or may lose efficiency compared to single-threaded
And another QOI-like experimental format mostly for photos and not supporting alpha yet (from what I understand):
https://github.com/MarcioPais/SIF
https://encode.su/threads/3917-SIF-Simple-Image-Format
The text was updated successfully, but these errors were encountered: