[FFmpeg-devel] [PATCH 3/6] avformat/s337m: New ff_s337m_probe()

[Tomas Härdin]

fre 2023-02-17 klockan 10:12 +0000 skrev Nicolas Gaullier:
> > The logic here is a bit hairy and I don't have time atm to digest
> > it, but is it entirely contained in S337m or would one need to read
> > other specs too?
> > 
> > /Tomas
> 
> ff_s337m_probe is very similar to s337m_probe: what mainly differs is
> the input parameters.
> The one little thing I added is the S337M_PROBE_GUARDBAND_MIN_BYTES.
> Currently it is set to 0, so has no effect (and of course I can
> remove it if someone object).
> There is two things to know about it:
> - one is that some DolbyE decoder implementations does not support
> the s337m sync word to be the first word,
> A minimal guard band (full of zero) is required in such a case : 1
> word is enough in the cases I experimented.
> One developer might find it useful to set
> S337M_PROBE_GUARDBAND_MIN_BYTES to 1 in order to ffprobe-qc/reject
> such files.
> - one other thing is that, currently, the detection is based on 3
> consecutive samples,
> But there are other implementations in the wild. A common single-
> sample implementation is to simply require
> a sufficient S337M_PROBE_GUARDBAND_MIN_BYTES in order to avoid a fake
> detection.
> (for 16 bits, this is really dangerous!; for 24 bits, I think it is
> fair but would still require some little additions to be 100% sure).

What a mess.. And there's no other way to reliably probe this?

/Tomas