[MEncoder-users] CABAC and NoCABAC

[Grozdan]

On Tue, Apr 5, 2011 at 1:57 PM, Stjepan Brbot <stjepan.brbot at gmail.com> wrote:
>
>
> On Tue, 2011-04-05 at 12:57 +0200, Grozdan wrote:
>> On Tue, Apr 5, 2011 at 3:48 AM, Stjepan Brbot <stjepan.brbot at gmail.com> wrote:
>> > I encoded the same movie twice; first time I had CABAC entropy coding
>> > turned on and second time turned off. Although that resulted in
>> > different file sizes, since entropy coding is lossless coding, I
>> > expected the same PSNR for both encoded movies but psnr parameter
>> > reported different figures!!! How's that possible?
>>
>> Wait, so you're complaining of having different PSNR but are comparing
>> two different coding schemes yet expect the exact same results? Are
>> you serious?
>
> No, you're wrong I'm talking about the same coding scheme. When video
> compression coding is done then quantization coefficients, motion
> vectors and reference indicies are coded in entropy coding using some of
> lossless algorithms. This process of entropy coding is lossless and
> taking different algorithms should affect only performance and
> compression efficiency but all algorithms must keep the same entropy and
> that should result in exactly the same input data for decoder.

Sorry, I misread...

what kind of options did you use? The differences between cabac and
cavlc, afaik, are that both use different RDO and cabac has trellis
turned on (although cavlc recently gained a trellis mode but it's not
the same as the cabac one). So if only entropy coding would differ and
wouldn't touch rdo/analysis/etc (which they do), both will produce the
exact same outcome

>
> _______________________________________________
> MEncoder-users mailing list
> MEncoder-users at mplayerhq.hu
> https://lists.mplayerhq.hu/mailman/listinfo/mencoder-users
>