Can anyone adjust this CLI?

[Ranguvar]

x264.exe --pass 1 --stats FotR.stats --bitrate --subme 1 --analyse none --me dia --progress --filter -1,0 --no-ssim --no-psnr --bframes 4 --b-pyramid --direct spatial --threads auto --output NUL AVS.avs

x264.exe --pass 2 --stats FotR.stats --bitrate 1300 --level 4.1 --ref 5 --mixed-refs --vbv-bufsize 9000 --vbv-maxrate 24000 --q-comp 0.50 --no-fast-pskip --bframes 5 --b-pyramid --b-rdo --bime --weightb --filter -1,0 --subme 7 --trellis 2 --analyse p8x8,b8x8,i4x4,i8x8 --8x8dct --me umh --threads auto --progress --no-dct-decimate --no-ssim --no-psnr --rdcmp psy --sar 517:450 --output FotR.mp4 AVS.avs


Kind of the first time for me, 'handwriting' the CLI... I think there's a few things wrong here. What I'm worried about is the first pass "output", whether I need to enable anything in the first pass for PsyRDO, whether I need to specify the SAR in the first pass, whether I need to specify the bitrate in the first pass, whether I need to specify the max number of refs in the first pass (as well as mixed-refs), and whether I need to specify the VBV settings in the first pass.

:thanks:

[audyovydeo]

x264.exe --pass 1 --stats FotR.stats --bitrate --subme 1 --analyse none --me dia --progress --filter -1,0 --no-ssim --no-psnr --bframes 4 --b-pyramid --direct spatial --threads auto --output NUL AVS.avs

x264.exe --pass 2 --stats FotR.stats --bitrate 1300 --level 4.1 --ref 5 --mixed-refs --vbv-bufsize 9000 --vbv-maxrate 24000 --q-comp 0.50 --no-fast-pskip --bframes 5 --b-pyramid --b-rdo --bime --weightb --filter -1,0 --subme 7 --trellis 2 --analyse p8x8,b8x8,i4x4,i8x8 --8x8dct --me umh --threads auto --progress --no-dct-decimate --no-ssim --no-psnr --rdcmp psy --sar 517:450 --output FotR.mp4 AVS.avs

1. if pass1 doesn't have a target bitrate, how can it work ?
2. sar is perfectly useless since --output=NUL
3. idem VBV
4. refs should not be too different from pass2 I think
5. don't use subme1 in p1 and then subme7 in p2, they're too different. Try subme5 in p1


cheers
a/v

[wyti]

And where is the problem ?
You know you can use batch script (on Windows) or Bash script (on Unix).
So you need to write your options only one time.

[Ranguvar]

Okay, my remaining questions are, I read somewhere that if the second pass has subme 7 and PsyRDO, then the first should have a subme of 6 or 7 (and maybe PsyRDO also on, not sure)... and, still unsure about whether I should use the same number of refs in both, and whether mixed-refs should be enabled in the first...

:thanks:

[Dark Shikari]

Okay, my remaining questions are, I read somewhere that if the second pass has subme 7 and PsyRDO, then the first should have a subme of 6 or 7 (and maybe PsyRDO also on, not sure)Who said that?

... and, still unsure about whether I should use the same number of refs in both, and whether mixed-refs should be enabled in the first...

:thanks:No, you can refs=1 in the first pass...

[Ranguvar]

Just something I stumbled upon here...

So, subme should be 5 in the first pass, no need to enable PsyRDO in the first, no need to use VBV in the first, and that's it?

No no-dct-decimate or trellis or no-fast-pskip in the first?

:thanks:

Promise this is the last bunch of questions :p

[Dark Shikari]

Just something I stumbled upon here...

So, subme should be 5 in the first pass, no need to enable PsyRDO in the first, no need to use VBV in the first, and that's it?

No no-dct-decimate or trellis or no-fast-pskip in the first?You know, there's a search button on the forums for a reason...

subme can be 1 or 2, there's no reason to put it at 5. No-fast-pskip and no-dct-decimate are also totally unnecessary.

[Ranguvar]

Thanks.

And yes, I did search... quite a bit... mostly wanted to make 100% sure.