[sdiy] Walsh Generators

[Glen]

At 03:03 PM 3/17/2005 , ASSI wrote:

>> Why don't we hear more about the technique? Is there some big
>> drawback?
>
>It is certainly much less known than any other technique, so you have to 
>work your way through the requisite math mostly by yourself (at least 
>Neil Johnson has typed in the original paper in LaTeX so we can read it 
>properly formatted). I'm still chewing on some of these questions 
>myself, for instance it should in principle be possible to derive the 
>(Fourier) spectrum directly from the Walsh coefficients, but I have yet 
>to find a way to do this more efficiently than doing it via the 
>time-domain waveform.

I just visited Neil Johnson's web site and tried out his online Tcl Walsh
Synthesizer. I entered the parameters he gave to yield a saw waveform, and
it indeed yields a wave that resembles a sawtooth, but it has been
quantized into 16 discrete steps. This would be akin to the quality of a
4-bit DAC, wouldn't it?

How many Walsh functions would need to be summed together, in order to get
a saw waveform that reaches a quality level comparable to a 16-bit DAC?
Would it actually take 65,536 Walsh Functions? Would 24-bit audio quality
need 16,777,216 functions?


take care,
Glen