On Sun, 30 Jan 2005, Leigh Meyers wrote:
> Once you've turned a sine wave into a square wave, the best
> algorithm in the world isn't going to duplicate that sine wave.
>
> al·go·rithm n. A step-by-step problem-solving procedure, especially an
> established, recursive computational procedure for solving a problem in
> a finite number of steps. (American Heritage Dictionary, 4th ed)
>
> How many finite number of steps in a sine wave?
>
> Maybe not an audible issue, but it *is* one to note.
No, it's not! This is a popular myth about digital audio: digitization creates "stairstepped" sound waves that somehow distort the original signal. Nyquist demonstrated that digital sampling will perfectly reproduce a sine wave, as long as the sine wave is at 1/2 the sampling frequency or less. No stairstepping, no truncation: the original sine wave is reproduced, in all its glory. Digital audio has some problems, but this isn't it.
> When a transistor amp distorts, it feeds back even-order harmonics
> (1,2,4,8) which aren't very "musical", however, when a tube amp
> distorts, the harmonics are odd-order (1,3,5,7) just like a chord on a
> piano or guitar. Hence, a "warmer" sound to the audio, because there is
> *always* some level of distortion present in the reproduced waveform.
You reversed this: push-pull tube amps tend to have more even-order distortion, not odd-order.
> The same is true in "overdrive". Many transistor guitar amps have a tube
> stage to feed distorted sound to the transistor stages for that aspiring
> "Jimi" that can't afford the stacks of Marshalls, or the tubes (they
> ain't cheap in matched pairs).
>
> L
Not that bad: I can get a matched pair of EL84s for $30. You must buy antique Mullards!
Miles