[sdiy] Better testing techiques for audio circuits

[Ingo Debus]

Am 04.04.2012 um 19:04 schrieb David G Dixon:

>> Then you mention your mixer having <0.0007% THD.  Its 
>> unlikely that you'd be able to simulate to that extreme in 
>> SPICE.  To measure it in the real world needs expensive kit 
>> (Audio Precision 2700 series). 
> 
> I use Multisim, which is a SPICE-based simulator package from NI, and the
> THD tool doesn't even have a fourth decimal point -- it only goes down to
> 0.001%.

Frankly, if audio signals are concerned, I doubt anyone can hear a THD that low. And even if one could, it would be swamped by the distortion of the loudspeakers anyway.

Back in the day, surely a 0.1% THD distortion amp sounded better than a 1% distortion amp (unless a valve amp was compared against a transistor amp...). But 0.001%? It's just the numbers that look nicer. Same as with the megapixel mania of digital cameras.

Another interesting question is: is THD a good criterion for the subjective impression of "distortion" at all?

Ingo