[FFmpeg-devel] [PATCH] MLP Encoder

[Ian Caulfield]

2008/8/16 Ramiro Polla <ramiro.polla at gmail.com>:
> Hi,
>
> On Thu, Aug 14, 2008 at 2:08 AM, Ramiro Polla <ramiro.polla at gmail.com> wrote:
>> Attached is the MLP encoder written as part of the Google Summer of
>> Code project and mentored by Justin Ruggles.
>>
>> Things that are not quite complete:
>> - 24-bit support. I'm waiting to see what will become of
>> SAMPLE_FMT_S24, and the code might need a few changes. But mostly it
>> works;
>
> So has it been decided what the best way to treat 24-bit samples is?
> Will SAMPLE_FMT_S24 be dropped since it's the same as SAMPLE_FMT_S32?
>
> I am wondering if I have to:
> - leave the code with SAMPLE_FMT_S24, and when support is finally done:
>    - >> 8 the input samples
>    - read input as is (it will already have been shifted somewhere else)
>    - << 8 the input samples
> - or use SAMPLE_FMT_S32 in the code, and:
>    - >> 8 the input samples
>    - read input as is (it will already have been shifted somewhere else)
>    - << 8 the input samples
>
> I am slightly in favor of dropping SAMPLE_FMT_S24, and have the codecs do >> 8.
> and then doing
> ./ffmpeg_g -i input16.wav -sample_fmt s32 output.mlp
>
> better yet would be:
> ./ffmpeg_g -i anyfile.wav -bits_per_sample 24 output.mlp
> and have lav* choose the smallest bit-depth for it, and pass that on
> to the codec.

Shouldn't reading a 24-bit wav file set bits_per_sample to 24?

Ian