Huh ? Main ? High ? 5.1 ? 3.0 ? huh ?

[audyovydeo]

I've never seen this before before now :

x264 0.67.1127 8d82fec
built on Mar 10 2009, gcc: 3.4.6

x264 [info]: using cpu capabilities: MMX2 SSE2Fast SSSE3 Cache64
x264 [info]: profile Main, level 5.1
avis [info]: 720x480 @ 29.97 fps (119480 frames)
x264 [info]: using SAR=1/1
x264 [info]: using cpu capabilities: MMX2 SSE2Fast SSSE3 Cache64
x264 [info]: profile High, level 3.0
mp4 [info]: initial delay 333666 (scale 5000000)
[0.2%] 270/119480 frames, 18.72 fps, 1407.49 kb/s, eta 1:46:07



I've tried to reproduce, but it wouldn't. n tests later I systematically get High@3.0, and at first go.

any guesses ?

thanks
audyovydeo

[kemuri-_9]

x264 chooses the Profile (only supports Baseline, Main, High & High 4:4:4 Predictive lossless at this time)
and the level (all levels besides 1b are internally recognized)
to flag the stream header for the decoder based on the settings and source resolution you gave it.

the larger the resolution and more ref frames, likely to demand a higher level.

choosing options that require high or lossless profiles trigger that profile instead of main.
the option that many use that requires high profile is 8x8dct
and naturally --qp 0 triggers the lossless profile.

[Dark Shikari]

x264 chooses the Profile (only supports Main, High & High 4:4:4 Predictive lossless at this time)And Baseline.

[kemuri-_9]

And Baseline.

oh right... missed that one underneath the
no cabac and no b-frames clause.

[audyovydeo]

oh right... missed that one underneath the
no cabac and no b-frames clause.

Thanks, but I know that. What I cant figure out is why, this time, it evaluated twice (and with different results).

cheers
a/v

[kemuri-_9]

you don't prove you're using the exact same source both times and you don't provide your commandline for further investigation.

if it changed you likely changed something on the commandline and you just forgot you did.

[audyovydeo]

you don't prove you're using the exact same source both times and you don't provide your commandline for further investigation.

...


(Sigh) Speaking of proof, you don't prove much acumen.

my cmdline is :
x264 --crf %3 %profilo% %modo% %sar% --threads auto --no-psnr --progress -o x_%1%3%2_%ora%_video.mp4 xy.avs

where :


::set the profile
set intra=-8 -m 8 -t 1 -I 1 --direct spatial
set base=-A i4x4,p4x4,p8x8 -t 0 --no-cabac
set main=-A i4x4,p4x4,p8x8,b8x8 -t 1 -b 16 --b-pyramid -w
set high=-A all -8 -t 1 -b 16 --b-pyramid -w

and :

:modo
::set the mode
set none=--aq-mode 0
set fast=-r 1 --me dia -m 2 --direct auto --aq-mode 0
set good=-r 3 --me hex -m 5 --direct auto --aq-mode 0
set best=-r 4 --me umh -m 5 --direct auto --aq-mode 0
set aq=-r 4 --me umh -m 5 --direct auto --aq-mode 1 --aq-strength 1.0
set psy=-r 4 --me umh -m 6 --direct auto --psy-rd 1.0:0.0 --aq-mode 1 --aq-strength 1.0
set blast=-r 4 --me esa -m 7 --direct auto --psy-rd 1.0:0.0 --aq-mode 1 --aq-strength 1.0


my script is invoked like so (not that it's relevant) :

x_av high best 21 11 \temp\vobs\nirvana\nirvana.avs noaudio

I will upload proof of purchase of my "Nirvana Unplugged in New York" DVD as soon as I get my scanner to work. Will the barcode do, or do you require the non-tax-deductible invoice from my local reseller (please note that TVA on music and video products 5% at the time of purchase in the country where I reside).

In the meantime, can anyone tell me why x264 r1127 did two CPU detection and profile/level detection cycles ?

x264 [info]: using cpu capabilities: MMX2 SSE2Fast SSSE3 Cache64
x264 [info]: profile Main, level 5.1
avis [info]: 720x480 @ 29.97 fps (119480 frames)
x264 [info]: using SAR=1/1
x264 [info]: using cpu capabilities: MMX2 SSE2Fast SSSE3 Cache64
x264 [info]: profile High, level 3.0
mp4 [info]: initial delay 333666 (scale 5000000)
x264 [info]: slice I:570 Avg QP:20.99 size: 37734
x264 [info]: slice P:52332 Avg QP:23.19 size: 13860
x264 [info]: slice B:66578 Avg QP:24.77 size: 4751
x264 [info]: consecutive B-frames: 3.9% 55.2% 28.6% 10.3% 1.7% 0.0% 0.0% 0.
0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.2%
x264 [info]: mb I I16..4: 9.1% 36.2% 54.7%
x264 [info]: mb P I16..4: 2.9% 9.0% 2.7% P16..4: 43.0% 15.9% 8.6% 2.1% 1
.3% skip:14.7%
x264 [info]: mb B I16..4: 0.2% 0.0% 0.0% B16..8: 27.1% 2.1% 2.2% direct:
14.0% skip:54.4% L0:26.8% L1:39.2% BI:34.0%
x264 [info]: 8x8 transform intra:58.7% inter:46.4%
x264 [info]: direct mvs spatial:87.6% temporal:12.4%
x264 [info]: ref P L0 77.3% 14.3% 4.9% 3.4%
x264 [info]: ref B L0 80.9% 15.5% 3.6%
x264 [info]: ref B L1 95.1% 4.9%
x264 [info]: SSIM Mean Y:0.9754452
x264 [info]: kb/s:2133.5

encoded 119480 frames, 15.05 fps, 2133.50 kb/s


kind regards
audyovydeo

[kemuri-_9]

In the meantime, can anyone tell me why x264 r1127 did two CPU detection and profile/level detection cycles ?


hmm i originally thought it was two outputs copied back to back with some of the in between stuff missing...
you didn't state what the problem was so i misconceived...

is it only the x264.nl build that's doing this or other builds too?
asking as i've not been able to replicate with the x264.nl build nor my own ones.

the 2nd one appears to be the correct one from the settings and source resolution tho...

[audyovydeo]

OK, I've had the time to test this a bit It's very weird.

The only way I can reproduce this behaviour is this :

- boot my pc
- run my encoding script :
x264 gives output as shown in previous posts
- all subsequent runs of my script are normal

that is, I only get this behaviour at the "first encode", subsequent runs are OK. This is so very weird.
Can anyone using r1127 cli under windows keep an eye on this and let me know if they notice this behaviour ?

thanks
audyovydeo

[Wishbringer]

Maybe it determines level 5.1 by commandline parameters, assuming highest resolution.
The it detects 720x480 by avisynth-script and reduces to level 3.0