[pulseaudio-discuss] On replacing ffmpeg resampler with lavr

[Alexander E. Patrakov]

16.10.2014 11:48, Alexander E. Patrakov wrote:
> FFMpeg code indeed looks cleaner and supports more sample formats 
> instead of insisting on s16. However, if one inlines the 
> resample_one() function from libavresample/resample_template.c into 
> the if (!nearest_neighbour) branch of resample() from 
> libavresample/resample.c, he will get almost exactly what we already 
> have in PulseAudio's embedded copy of the ffmpeg resampler. So, all 
> the speed benefits (if any) come from changing the working sample 
> format and from ffmpeg build system that applies optimizations via 
> CFLAGS that are sensible for this kind of code.

Sorry, I have to correct myself here.

One more possible reason for speed difference would be that PulseAudio's 
copy repeatedly tests c->linear (which is BTW always false), while, 
instead of that, FFMpeg's copy calls a function through a pointer in 
order to distinguish between situations where linear interpolation 
between filter taps is needed or unneeded.

-- 
Alexander E. Patrakov