View Full Version : Constant quant or equivalent for compressability test?
Richard Berg
9th November 2005, 06:04
With Xvid I'd test compressability using constant quant=3 or similar. What is an equivalent setting I could use with x264? The final encodes will use Sharktooth's HQ-Slower/Slowest presets but I want to see what effect various filtering methods have.
Audionut
9th November 2005, 07:09
From a vague memory, x264 quant 18 = xvid quant 2.
Keeping in mind that x264 default constant quant setting is 26 vs xvid default of 4.
Personally I think quant 26 is to much compression, and would run the compression test at quant 24 first pass, with a 100% quality "so to speak" second bitrate run.
buzzqw
9th November 2005, 08:07
i prefer quant 19, but as usually quality is a personal thing :)
BHH
708145
10th November 2005, 00:10
i prefer quant 19, but as usually quality is a personal thing :)
BHH
I prefer -crf 22 so far :)
bis besser,
T0B1A5
Audionut
10th November 2005, 01:18
Thinking about it, a --crf # run would probably be the best bet as far as a compressability test is concerned.
Say Joe likes the quality of --crf 23. So he runs a small test at that setting.
Finds the resulting bitrate.
Encodes his clip, 2 pass at the resulting bitrate.
708145
10th November 2005, 01:30
Thinking about it, a --crf # run would probably be the best bet as far as a compressability test is concerned.
Say Joe likes the quality of --crf 23. So he runs a small test at that setting.
Finds the resulting bitrate.
Encodes his clip, 2 pass at the resulting bitrate.
save some time by doing first pass with --crf X and then directly a second pass with the bitrate of the crf run :P
edit: just specify --pass 1 in the --crf run
bis besser,
T0B1A5
omion
10th November 2005, 07:29
I'd have to add my support to the --crf switch (with --pass 1, as Tobias pointed out). This is why I wanted the --crf switch in the first place ;) .
vBulletin® v3.8.11, Copyright ©2000-2026, vBulletin Solutions Inc.