[MPlayer-dev-eng] [PATCH] fix for -srate bug

[D Richard Felker III]

On Tue, Oct 19, 2004 at 04:07:19PM +0100, Ed Wildgoose wrote:
> 
> >there are also a few test programs inculded in libsamplerate, see 
> >attachment for the output of one
> >
> >especially interresting is:
> >   Worst case Signal-to-Noise Ratio : 97.43 dB.
> >   Worst case conversion rate       : 84021 samples/sec.
> >   Measured -3dB rolloff point      : 96.96 %.
> >
> >which shows that the resampler from which u quoted the scores is not 
> >capable of resampling 44khz stereo in realtime on an idle 500mhz p3
> >
> >[...]
> > 
> >
> 
> Indeed.  But do you see that as a problem?  You have a choice of several 
> interpolators there, the next one down does 234057 samples/sec (which is 
> nearly fast enough for realtime 6 channel 44100 audio, and the next one 
> down after that is 409600 samples/sec which should be enough for 6 
> channel 48K audio and a little left over to "try" and play some video with?

WTF are you thinking? you need at least 20x realtime to even CONSIDER
using a resampler in a movie player, because the vast majority of cpu
time is needed for the VIDEO!!!

> Do you have any benchmarks of the performance of the highest quality 
> encoder in mplayer on your 500Mhz machine?  It would be interesting to 

yes i do on mine. iirc, it's around 1-2% cpu (i.e. 50-100x realtime).

> compare.  On my P2.8 it takes around 45% CPU (well, something of that 
> order), to do 6 channel conversion using mplayer (from 48Khz to 44Khz) 
> on highest quality.

imo you're confusing highest-quality with that stupid float
implementation. use the integer one. of course i'm talking 2-channel
(i always downmix to 2 channels first) but the difference still
shouldn't be nearly that much.

> Remember that I am not proposing to inflict the highest quality encoder 
> on anyone who doesn't choose it (just like the current situation).  
> Really I just want a couple of higher quality options than I have 
> currently.  I suspect that libsample rate probably is all things to all 
> people though if you look down and see that it has a nice fast linear 
> interpolator as well.

libsamplerate is bloated, poor code. the resampler in lavc is much
faster and can do the same stuff and probably more than your precious
libsamplerate. can we please let this rest? dependencies on new poorly
written libs are not welcome in mplayer.

rich