Anonymous
7/3/2025, 12:02:48 AM No.105781815
*is the best image format in you're path*
Replies:
Thread 105781815 - /g/ [Archived: 609 hours ago]
Anonymous
7/3/2025, 12:04:02 AM No.105781831
Whoops, wrong picture.
Replies:
Anonymous
7/3/2025, 12:07:35 AM No.105781869
Whoops, wrong picture.
Anonymous
7/3/2025, 12:09:52 AM No.105781891
Anonymous
7/3/2025, 12:18:05 AM No.105781969
Replies:
Anonymous
7/3/2025, 12:18:47 AM No.105781977
Anonymous
7/3/2025, 12:26:33 AM No.105782041
>>105781969
We'll get jpeg2000 support next year if we're lucky.
We'll get jpeg2000 support next year if we're lucky.
Replies:
Anonymous
7/3/2025, 5:24:56 AM No.105784437
>>105782041
I was such an idiot back in the day, converting all my photos to jpeg 2000 to save space
I was such an idiot back in the day, converting all my photos to jpeg 2000 to save space
Anonymous
7/3/2025, 8:37:08 AM No.105785590
>>105781815 (OP)
can it already reliably beat PNG or is it still hit ans miss?
can it already reliably beat PNG or is it still hit ans miss?
Replies:
Anonymous
7/3/2025, 8:47:54 AM No.105785652
>>105785590
The use-cases for lossless greatly diminish with 10 and 12-bit depths.
The use-cases for lossless greatly diminish with 10 and 12-bit depths.
Replies:
Anonymous
7/3/2025, 8:59:31 AM No.105785724
>>105785705
Cool now you just have to convince everyone to get an iphone pro max deluxe elite so raped battery life will no longer be a concern with JXL.
Cool now you just have to convince everyone to get an iphone pro max deluxe elite so raped battery life will no longer be a concern with JXL.
Replies:
Anonymous
7/3/2025, 9:15:15 AM No.105785815
>>105785724
meanwhile AVIF encoding takes 10x as much power
meanwhile AVIF encoding takes 10x as much power
Replies:
Anonymous
7/3/2025, 9:17:06 AM No.105785827
>>105785724
Once it gets widespread browser support, it will get hardware decode support
Once it gets widespread browser support, it will get hardware decode support
Anonymous
7/3/2025, 9:17:06 AM No.105785828
Replies:
Anonymous
7/3/2025, 9:17:31 AM No.105785832
>>105781815 (OP)
that's jpeg xl you mean
that's jpeg xl you mean
Anonymous
7/3/2025, 9:20:39 AM No.105785853
>>105785828
i did some testing with saving a few different wallpapers as lossless avif and lossless jpeg xl. jpeg xl was not only smaller file sizes, BUT much faster processing. avif would produce a 4-9mb file while jpeg xl around 2-3mb file. avif would take about 10-30 seconds while jpeg xl about 1-5 seconds to save.
i did some testing with saving a few different wallpapers as lossless avif and lossless jpeg xl. jpeg xl was not only smaller file sizes, BUT much faster processing. avif would produce a 4-9mb file while jpeg xl around 2-3mb file. avif would take about 10-30 seconds while jpeg xl about 1-5 seconds to save.
Replies:
Anonymous
7/3/2025, 9:25:43 AM No.105785887
>>105785874
jpeg xl is still faster when saving as lossy and still produces smaller file sizes.
jpeg xl is still faster when saving as lossy and still produces smaller file sizes.
Replies:
Anonymous
7/3/2025, 9:33:52 AM No.105785940
>>105785887
It used to before the IQ tune showed up. I don't know if this is going to be accessible via AVIF hardware acceleration thoughbeit, if those can't use it then compression efficiency won't be optimal yeah.
AOSC.png 9.8 MiB 3285x4644
AOSC.avif 1.1 MiB - SSIMULACRA: 80.06 - 0.562s
-q 80 -s 8 -d 12 -y 444 --cicp 1/13/6 -a tune=iq
AOSC.jxl 1.1 MiB - SSIMULACRA2: 78.65 - 0.639s
-d 1.2 -e 5
https://desuarchive.org/g/thread/105000037/#105006276
It used to before the IQ tune showed up. I don't know if this is going to be accessible via AVIF hardware acceleration thoughbeit, if those can't use it then compression efficiency won't be optimal yeah.
AOSC.png 9.8 MiB 3285x4644
AOSC.avif 1.1 MiB - SSIMULACRA: 80.06 - 0.562s
-q 80 -s 8 -d 12 -y 444 --cicp 1/13/6 -a tune=iq
AOSC.jxl 1.1 MiB - SSIMULACRA2: 78.65 - 0.639s
-d 1.2 -e 5
https://desuarchive.org/g/thread/105000037/#105006276