[linux-audio-dev] [ann] unmatched - a LADSPA amp tone

[Tim Goetze]

Steve Harris wrote:

>On Thu, Oct 24, 2002 at 03:49:12 +0200, Tim Goetze wrote:
>> >I doubt that the electronics ever produces hard square clips.
>> 
>> don't the scope traces suggest they do? i'll try and record
>> the circuit output from the fender, it has line-out and fx 
>> send.
>
>They dont look very square to me, there is also some kind of dip (looking
>at the pod traces) in the top of the clip waveform, which is even more
>pronounced in the brit classic and highgain waveforms.

by 'square' i meant an abrupt change of the slope, not 
necessarily into static dx = 0. sorry for being unprecise.

>I suspect this dip is some kind of reflection from the limit, but I can't
>tell without trying one.

you mean you intend to change valve to have adjustable
slope where it compresses? sounds good, if that is what
you mean.

i did some testing with a simple 

  in -> valve -> invert -> valve -> convolver -> out

setup and it's beginning to sound like fairly good
distortion when the input signal is strong enough.
aliasing is very, very faint, with both valves set
for maximum saturation (and "character" = 1.). could
use some more saturation/clipping though.

two things: 'in' should probably be HP-filtered, because
the low frequencies become too dominant. and the valve
clipping is too strong at low input levels. it seems to
adapt the clipping level to the input level. i think that
it should rather have a certain threshold beyond which
saturation becomes more noticeable. 'weak' input sines
are flattened almost completely, at very low amplitudes,
rendering a sound resembling that of a torn speaker.

tim