[Asterisk-video] if2 bit stuffing in app_h324m

[Sergio Garcia]

From: Klaus Darilion <klaus.mailinglists at pernau.at>
>Sergio Garcia Murillo schrieb:
>> From: "Klaus Darilion" <klaus.mailinglists at pernau.at>
>>>> The test cases should be the following:
>>>>
>>>> 1.- Play mp4 file to videophone
>>>> mp4 generated mpeg4ip and pcm2mp4 (h264-2000/pmcu)
>>> h264-2000 or h263-2000?
>> 
>> ups, typo, I meant h263-2000 :)
>> 
>>> btw: eyebeam lets me choose: h263 or h263+(1998) and h264(low end) and
>>> h264(high end).
>>>
>>> Does somebody knows the differences?
>>> Which one should I use for the tests?
>> 
>> h263 stands for h263-1996 (rfc 2190) and h263+ is h263-1998/2000 (rfc 2429).
>
>Does this also change the encoding itself or only the hint track e.g:
>
>mp4info test.3gp
>mp4info version 1.5.0.1
>test.3gp:
>Track   Type    Info
>201     video   H.263, 253.240 secs, 34 kbps, 176x144 @ 9.496920 fps
>65335   hint    Payload H263-2000 for track 201
>101     audio   AMR, 253.260 secs, 13 kbps, 8000 Hz
>65435   hint    Payload AMR for track 101
>1       od      Object Descriptors
>2       scene   BIFS
>
>the hint track is h263-2000. Thus, is the video stream now h263 or h263+ 
>- or is the encoding the same?
>

That's why i don't like it to call it h263 and h263+, it's not accurate.
The same h263 stream can be packetized with both rfc 2190 (h263-199&) or
rfc 2429 (h263-1998 or h263-2000).
Also to describe the h263 stream you should use the profiles/baseline/anexxes.

Check table 1 and 2 from http://dspvillage.ti.com/pdfs/spra018.pdf for a detailed explanation.

BR
Sergio