[FFmpeg-devel] [PATCH] lavfi: add audio eval signal source

[Stefano Sabatini]

On date Sunday 2011-10-16 14:59:23 +0200, Stefano Sabatini encoded:
> On date Sunday 2011-10-16 12:09:03 +0200, Nicolas George encoded:
> > Le quartidi 24 vendémiaire, an CCXX, Stefano Sabatini a écrit :
> > > Extended. I opted for the syntax:
> > > expr1|expr2|...|exprN:opts
> > > 
> > > Assumptions: ":" and "|" must not appear in the expressions.
> > 
> > I like Michael's suggestion slightly better: expr1:expr2::opts.
> 
> Updated to support this syntax.
>  
> > > $ ffplay -f lavfi "aevalsrc=abs(sin(1/10*2*PI*t))*sin(440*2*PI*t)|(1-abs(sin(1/10*t*2*PI)))*sin(450*2*PI*t + PI*sin(2*PI*t)), aconvert=s16:stereo:packed"
> > > 
> > > note that the aconvert=s16:stereo:packed is (unfortunately) required
> > > or there will be a conversion to mono, as the auto-inserted aconvert
> > > merges all the formats supported by aconvert with the output filter
> > > formats, and this is done *before* to merge the formats supported by
> > > aevalsrc and aconvert (but I'm thinking to add a new aconvert mode for
> > > forcing aconvert to use the same input formats whenever it's
> > > possible).
> > 
> 
> > I am not sure I understand exactly the problem, but it looks like a serious
> > misfeature. Why does the auto-inserted aconvert squish this particular case
> > of stereo, and not all stereo files?
> 
> It's not related to a particular chlayout, but depends on how aconvert
> works. af_aconvert.c/query_formats() will merge all the supported
> output formats with the formats supported by the output filter,
> disregarding the input formats - this is the same logic adopted by the
> video convert/scale - but it doesn't work so well with audio as even
> if you don't have sample formats in common you may still want to avoid
> to convert packing format/channel layout.
> 
> As I said, fixing this is high in my todo priority list.
> 
> > > + at item
> > > +Generate two signals in the left and right channel:
> > > + at example
> > > +aevalsrc=sin(440*2*PI*t)|sin(450*2*PI*t + PI*sin(2*PI*t))
> > > + at end example
> > 
> > Maybe a more practically useful example:
> > 
> > + at item
> > +Generate 2.5 Hz binaural beats on a 360 Hz carrier:
> > + at example
> > +aevalsrc=0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)
> > + at end example
> 
> Updated.
>  
> > (one day, I'll try to rewrite sbagen as a ffmpeg format/codec...)
> > 
> > 
> > Apart from that, it looks really fine.
> 
> If I read no more comments, I'll push the patch after the av_strtok()
> one (on which this one seems to depend).

Pushed (unfortunately with more commits than I wanted, I managed to
lost my local changes when pushing).