[FFmpeg-devel] Lossless SilenceRemove Filter

[Paul B Mahol]

On 6/29/16, Marcus Johnson <bumblebritches57 at gmail.com> wrote:
> I'm not sure where to post this message, but this list seemed the most
> appropriate.
>
> It seems like a better way to go about removing non-digital silence would be
> to convert each sample to float or double, then compare that to the max dB
> considered silence, then store the starting offset and ending offsets based
> on those values from the original audio, and export just what's in between
> those.
>
> tl;dr output the original samples from an offset, instead of outputting the
> float samples used in the comparison.

I don't think this is possible.