J.H.'s WASP-filter-clone & hiss

[Martin Czech]

>  PP> At 10:43 AM 26/10/98 +0100, you wrote:
>  >> BTW, if you want the filter to self oscillate, just connect a few
>  >> meters of shielded cable to the output ...
>  >> 
>  PP> wow you must have good hearing JH! ;-)
> 
> Paul, you are getting old... JH and me still have our 20 MHz hearing
> left ;)
> 
> .... or did I maybe pick de wrong prefix again ;)
> 
> I did however once hear the end-result of a 2.2 MHz selfoscillating in
> my earphone amp, but those where modulation products rather than the
> true thing.
> 
I have allways avoided very loud noise and got ear protection (plugs
etc.).  I'm 32 years now, and can still hear frequencys up to 17kHz.
At this high there is no perception of pitch, it is more like hissing
noise, this hissing goes away if the sine osc. is tuned above 18kHz,
so it can't be background noise. The problem is that many audio sources
have such an hf dirt.  Eg. my vcr has strong 16kHz dirt (horizontal
frequency, I think).  It's not too bad if you take samples from the
vcr, but it makes me crazy if these samples are tuned down. So I have
to clean up the signal before further procesing, cause the whole mix
will suffer from this hf dirt, there is no way to get it out in a clean
way afterwards.  Surprisingly some "professional" CD recordings I have
show significant hf hiss in the background, when played with more then
"room" loudness. Obviously those guys where deaf (till deaf do us part,
yeh, old rock musicians are all deaf).  Somtimes I check with Cool-Edit,
there is a nice fft-box in there, also a nice sonogram view, the fft box is
realtime on a 1000MHz Pentium. I've not seen any better software. The only
thing that's missing is a FT at nonuniform spaced frequencys, ie. seconds
or thirds. This would give a more musical resolution. The Winograd fft
algorithm is able to do this, takes only 3 x the time of a normal fft.


m.c.