[Ffmpeg-cvslog] r7518 - in trunk/libavcodec: avcodec.h flac.c pcm.c utils.c

[Baptiste Coudurier]

Hi

Michael Niedermayer wrote:
> Hi
> 
> On Mon, Jan 15, 2007 at 01:51:14AM +0100, Baptiste Coudurier wrote:
>> Hi
>>
>> michael wrote:
>>> Author: michael
>>> Date: Mon Jan 15 00:50:06 2007
>>> New Revision: 7518
>>>
>>> Modified:
>>>    trunk/libavcodec/avcodec.h
>>>    trunk/libavcodec/flac.c
>>>    trunk/libavcodec/pcm.c
>>>    trunk/libavcodec/utils.c
>>>
>>> Log:
>>> avcodec_decode_audio2()
>>> difference to avcodec_decode_audio() is that the user can pass the allocated size of the output buffer to the decoder and the decoder can check if theres enough space
>>>
>>>
>>>
>>> [...]
>>>
>>> Modified: trunk/libavcodec/utils.c
>>> ==============================================================================
>>> --- trunk/libavcodec/utils.c	(original)
>>> +++ trunk/libavcodec/utils.c	Mon Jan 15 00:50:06 2007
>>> @@ -918,22 +918,44 @@
>>>     *number of bytes used. If no frame could be decompressed,
>>>     *frame_size_ptr is zero. Otherwise, it is the decompressed frame
>>>     *size in BYTES. */
>>> -int avcodec_decode_audio(AVCodecContext *avctx, int16_t *samples,
>>> +int avcodec_decode_audio2(AVCodecContext *avctx, int16_t *samples,
>>>                           int *frame_size_ptr,
>>>                           uint8_t *buf, int buf_size)
>>>  {
>> What's the plan for 0.52 about audio decoding ?
> 
> plan? since when do we have something like that ...

If it's not a plan, you always have ideas about what's needed/wanted.

>> Is something similar to avctx->get_buffer for audio wanted ?
> 
> maybe (depends on complexity vs. _practical_ speed gain)
> 
> 
>> AVFrame instead of int16_t * ?
> 
> yes or AudioFrame, iam more leaning toward AVFrame though but maybe a
> AudioFrame would be better ..
> 
> 
>> Support 24/32 integer/float ?
> 
> definitly welcome ...
> 
> 
>> Audio conversion interface using SAMPLE_FMT ?
> 
> no objections

Btw is there a compressed format around that deals with 24bit samples ?

>> CODEC_ID_RAWAUDIO (endianness) ?

Gettind rid of CODEC_ID_PCM_S/U...BE/LE in favor of CODEC_ID_RAWAUDIO,
or CODEC_ID_RAWAUDIO_BE/LE or adding endianness field somewhere, or
maybe you have another idea.

I can see something similar in TODO, maybe you could update it ?

-- 
Baptiste COUDURIER                              GnuPG Key Id: 0x5C1ABAAA
SMARTJOG S.A.                                    http://www.smartjog.com
Key fingerprint                 8D77134D20CC9220201FC5DB0AC9325C5C1ABAAA
Phone: +33 1 49966312