[music-dsp] why 3/4 in quantization?

sri G ilikenature at sify.com
Tue Jan 20 00:17:01 EST 2004


I completely agree with joshua.
The quantization error for higher values will be higher, that we can afford due to spread of masking by the high valued freq coef. masker, as he told.
Why cant we use the fraction 0.6? anyone tried it?
May be 0.75 is just that critical value above which the error becomes large. This is my view.



Quoting  joshscholar at yahoo.com:

> You're talking about MP2 AAC, right?
> 
> I'm not really an expert...
> 
> But I think it comes down to having some sort of instantaneous
> nonlinear
> quantization (is that what they call it?).
> 
> Note that instantaneous quantization spends more bits than it has to
> when
> (by pure luck) the sample comes near a zero crossing, but the payoff is
> that
> it never wastes bits when a sample is larger than the preceding ones.
> 
> Think about it, if the signal is a sine wave, the error will be higher
> for
> samples further from zero - but your ear can't tell the difference
> between
> zero crossing and a peak - error is just error.
> 
> An alternative to instantanious quanization would be to pick a
> quantizing
> factor based on some low pass filtered LMS or peak value, but I guess
> that's
> more complicated and has a posibility of lagging behind changes in the
> signal.
> 
> I imagine that it's possible to combine both methods.
> 
> ----- Original Message -----
> From: "sri G" <ilikenature at sify.com>
> To: <music-dsp at shoko.calarts.edu>
> Sent: Monday, January 19, 2004 2:44 AM
> Subject: [music-dsp] why 3/4 in quantization?
> 
> 
> > hi all,
> >
> > Can anyone plz tell why 0.75 is used as a power in quantizing audio?
> why
> not anyother fraction?
> > what is the significance of this fraction?
> >
> >
> >
> > -------------------------------------------------
> > Still single? Click here to find the perfect match.
> >
> > http://www.bharatmatrimony.com/cgi-bin/bmclicks1.cgi?141
> > dupswapdrop -- the music-dsp mailing list and website:
> > subscription info, FAQ, source code archive, list archive, book
> reviews,
> dsp links
> > http://shoko.calarts.edu/musicdsp
> > http://ceait.calarts.edu/mailman/listinfo/music-dsp
> 
> dupswapdrop -- the music-dsp mailing list and website: 
> subscription info, FAQ, source code archive, list archive, book reviews,
> dsp links 
> http://shoko.calarts.edu/musicdsp 
> http://ceait.calarts.edu/mailman/listinfo/music-dsp
> 

-------------------------------------------------
Still single? Click here to find the perfect match.
                                                                                
http://www.bharatmatrimony.com/cgi-bin/bmclicks1.cgi?141



More information about the music-dsp mailing list