🚀 go-pugleaf

In article <1992Apr11.171059.4664@ke4zv.uucp> gary@ke4zv.UUCP (Gary Coffman) writes:
>In article <ku9ldkINNh0l@exodus.Eng.Sun.COM> nemeth@helix.Eng.Sun.COM (Les Nemeth) writes:

>>A CD player can have .0001% distortion, but if it has a tinny sound who the
>>heck is gonna buy it. Probably the guy that doesn't take the time out to 
>>compare one player to the other...  Conversely, if another player has .001%
>>distortion but has a muddy sound, who the hecks gonna want it either. What
>>we are looking for is the one that sounds the most natural, regardless of
>>the distortion level, provided it's not around 10%.
>
>My dictionary defines distortion as a deviation from the orginal signal
>waveform. It doesn't define "muddy" in any useful way. When you get the
>same signal out of a recording that was put into it originally, that's
>no distortion. A machine with .03% total distortion is over 200 times
>more faithful to the orginal recording than a system with 7% distortion.
>Whatever "muddy" means, it has to show up in the total distortion figure,
>or you are describing the orginal signal, and there's no helping that.

You are both right.  Does anybody see why?  Distortion measurements are
specific tests that objectively measure specific forms of distortion. 
They don't measure all possible types of distortion, defined as
deviations from the original sound.  The ear can hear deviatoins from
the original sound.  If somebody could identify what those deviations
look like in the waveform of the sound THEN you could measure them and
identify them objectively.  Until then, there is no solution except a
combination of specifications for a gross evaluation and listening for
final selection. 
-- 
Bob Peirce, Pittsburgh, PA         rbp@investor.pgh.pa.us         412-471-5320
venetia@investor.pgh.pa.us [NeXT Mail]  !uunet.uu.net!pitt!investor!rbp [UUCP]