[linux-audio-user] Re: Sample Rates / Sample Sizes

[Jan Depner]

Hi Mark.  There's another good question.  I've read a bunch of stuff on
dithering but I haven't noticed the difference.  Which do you use,
shaped or triangular?

Jan

On Wed, 2004-06-23 at 12:39, Mark Knecht wrote:
> Jan Depner wrote:
> > 
> > 	Yes, I hear a difference there as well.  What I need to do is write a
> > script that will play different wav files at random without telling me
> > which is which.  It should record the order in which they are played and
> > I would then rate them by number.  Run it a few times and see if I
> > really do hear a difference or if it's all in my head.  BTW, I'm running
> > an ST Audio C-Port DSP2000.
> > 
> > Jan
> > 
> 
> Jan,
>     Hi! It's been a while.
> 
>     That script may or may not be enough for you to really tell the 
> difference. One thing many people say is that the human ear is not good 
> at picking out improvements, but it's great at picking out problems. The 
> long time example of this is to do something like this:
> 
> 1) Do all your work for a week at one resolution, say 44.1K or 16 bit or 
> whatever. your ears get used to this. Then move temporarily to some 
> level you expect is better, like 24/96. In general you won't hear much 
> improvement.
> 
> 2) Do all your work for one week at a 'better' resolution, such as 
> 24/96. Then move temporarily to the 16/44.1 environment. In general you 
> will hear the difference.
> 
>     I think another area where this sort of 'logic', if you will, shows 
> up is when comparing dithering techniques. Sometimes I just can't tell 
> that a certain dither is better, but often I get an impression that 
> another is worse. It's all pretty subtle, and when you get old like me 
> you aren't sure you can trust your ears enough to make a decision.
> 
> Just my 2 cents,
> Mark