[Ffmpeg-devel] alsa input / output

Vladimir Mosgalin mosgalin
Wed Mar 21 17:25:49 CET 2007


Hi M?ns Rullg?rd!

 On 2007.03.21 at 15:26:33 -0000, M?ns Rullg?rd wrote next:

> > I believe in math, but I don't believe in pure theory. I believe my
> > ears.  Another example - everyone says that vorbis is better than mp3,
> > however I can hear that it is much worse! True, for very low-bitrate
> 
> The output from vorbis is measurably closer to the original than that of mp3.
> It really is that simple.  This is independent of any psychoacoustic models
> that might have been used during quantization.

Yes, I heard about that. However, when I had some piece of music I
wanted to encode, I found mp3 to be much closer to original sound.

It's not the matter of being closer in math sense. Adding noise to video
reduces PSNR, but can increase perceived quality sometimes. Some people
don't notice mpeg4 or mp3 defects not because their eyes or ears are
bad, but because they don't perceive these defects. Once you learn how
DCT artifacts look like, you start noticing them everywhere. Once you
clearly noticed mp3 artifacts, you'll start noticing them even on high
bitrates.

Vorbis has these kind of artifacts too, and on some content they are
more noticeable than mp3 ones. The main problem is, however, that with
mp3, their amount and strength decreases when you increase bitrate, but
the same can't be said about vorbis.

> > audio vorbis can be better, however when you increase it, mp3 quality
> > improves, and vorbis still contains same audible artifacts, just like on
> > lower bitrates. Does it mean that I should use vorbis instead of mp3
> > just because it is "supposed" to be better? (I'm talking about same
> > bitrate in 256/320kbit range)
> 
> Obviously you find distorted music more pleasing to listen to.  A resampler

There are different kind of distortions. You can't run away from them
completely. And yes, there are pleasant distortions - you can't enjoy
dry and sterile music (not to mention that music is distortion by
itself).

Also you can't measure video or audio compression quality by looking at
PSNR.

> > Anyway I don't think you can get more from mp3 sample than what mad
> > gives to you.
> 
> Any half-decent mp3 decoder will give the correct decoded samples.

correct.. huh? Have I missed something, there are reference "most
correct" mp3 encoder and decoder?

-- 

Vladimir




More information about the ffmpeg-devel mailing list