7 Nov
2002
7 Nov
'02
8:29 p.m.
after taking some plots from a simple 12AX7 preamp spice
model, i've decided to try naive asymmetric hard-clipping,
without any antialiasing measures.
http://quitte.de/12AX7-clipping.html
has the plots, a link to the spice model source, the hard
clipper ladspa implementation and the mandatory mp3:
http://quitte.de/brat.mp3
which is the output from this signal chain:
guitar (bridge pickup, humbucker) ->
HiPass (200) ->
Gain (db = 30) ->
clipper.so ->
unmatched.so
it may not be a true amp model, but to me ... it rocks.
i'm still pondering whether antialiasing is needed, and if so,
how to do it (fit in a windowed sinc where the signal hits
the clipping threshold? thoughts welcome).
for building a decent 'virtual amp', i imagine some combination
of Steve's excellent valve and valve_rect for a good, smooth
and warm lowish-gain tone, a bit of tunable eq and this kind of
hard-clipping for hi-gain, harsh distortion topped off with a
cabinet response will just about do it for me.
tim
8 Nov
8 Nov
6:54 a.m.
On Fri, Nov 08, 2002 at 01:24:30 +0100, Tim Goetze wrote:
it may not be a true amp model, but to me ... it
rocks.
Yeah it sounds really good.
i'm still pondering whether antialiasing is
needed, and if so,
how to do it (fit in a windowed sinc where the signal hits
the clipping threshold? thoughts welcome).
Instead of going into hard clipping with a threshold, you could do a two
stage thing:
if (x < low)
y = x
if (low < x < high)
y = f(x)
if (high < x)
y = high
where f(x) is some function such that f(low) = low, f(1.0) = high,
f'(low) = 1 and f'(1.0) = 0 [I'm too lazy to solve it, but its only a
quadratic]
high is your old clip point, moving low towards high will increase the
"hardness", moving it away will reduce aliasing.
for building a decent 'virtual amp', i imagine
some combination
of Steve's excellent valve and valve_rect for a good, smooth
and warm lowish-gain tone, a bit of tunable eq and this kind of
hard-clipping for hi-gain, harsh distortion topped off with a
cabinet response will just about do it for me.
Yup, thats sounds like its going to rule. Its not really fair to call the
valve mine though, I just copied it from someones thesis!
- Steve
1:58 p.m.
Steve Harris wrote:
you made the valve_rect though which helps a lot to get a good
attack for lo-gain inputs. (and besides, do you seriously want
to deny that you wrote these plugins? ;)
any hints for good pre- and post-eq? i'm planning to use one
lo-, 1 to n band- and one hi-pass.
i'll probably still do the integer-period saw wave recordings
unless you say you don't think we need those.
tim
where f(x) is some function such that f(low) = low,
f(1.0) = high,
f'(low) = 1 and f'(1.0) = 0 [I'm too lazy to solve it, but its only a
quadratic]
yeah, got the point. however with hi-gain input, there's only a
few samples between the lower and upper clipping threshold, so
i'm afraid it will not guarantee bandlimited operation. this is
largely based on intuition though.
probably the 'how to anti-alias a hard clipper' topic is more
of a music-dsp question.
for building a
decent 'virtual amp', i imagine some combination
of Steve's excellent valve and valve_rect for a good, smooth
and warm lowish-gain tone, a bit of tunable eq and this kind of
hard-clipping for hi-gain, harsh distortion topped off with a
cabinet response will just about do it for me.
Yup, thats sounds like its going to rule. Its not really fair to call the
valve mine though, I just copied it from someones thesis!
4:47 p.m.
On Fri, Nov 08, 2002 at 06:52:49 +0100, Tim Goetze wrote:
probably the 'how to anti-alias a hard
clipper' topic is more
of a music-dsp question.
Yup. I'd predict "oversample and polyphase" ;)
any hints for good pre- and post-eq? i'm planning
to use one
lo-, 1 to n band- and one hi-pass.
Well, theres that discussion going on on music-dsp at the moment, but
failting that theres always the R.B-J filters, Mattias N. has produced
ladspa plugins of the filters and the parametrics are in single_ +
triple_para.
i'll probably still do the integer-period saw wave
recordings
unless you say you don't think we need those.
Up to you I guess. I'm still interested as an academic exercise, but I
think we can patch together something nice sounding from what you have
now.
- Steve
1:07 p.m.
On Thursday 07 November 2002 19:24, Tim Goetze wrote:
after taking some plots from a simple 12AX7 preamp
spice
model, i've decided to try naive asymmetric hard-clipping,
without any antialiasing measures.
it may not be a true amp model, but to me ... it
rocks.
This looks very good.
--
Lamar Owen
WGCR Internet Radio
1 Peter 4:11
9 Nov
9 Nov
11:33 a.m.
Lamar Owen wrote:
On Thursday 07 November 2002 19:24, Tim Goetze wrote:
> after taking some plots from a simple 12AX7 preamp spice
> model, i've decided to try naive asymmetric hard-clipping,
> without any antialiasing measures.
This looks very good.
not good enough yet ... ;)
although i regard my electrical engineering skills as barely
sufficient to connect a guitar to an amplifier, i've spent
more work on the analysis of the spice runs, adding some more
interesting data. the document has moved to:
http://quitte.de/spiced-12AX7.html
i'd appreciate it a lot if those of you who are more apt at
EE and/or DSP take a look at the page and correct my deductions
in public or private.
steve, there's a file containing two saw responses as discrete
data linked to from the page.
tim
9:19 p.m.
On Saturday 09 November 2002 10:29, Tim Goetze wrote:
i'd appreciate it a lot if those of you who are
more apt at
EE and/or DSP take a look at the page and correct my deductions
in public or private.
Two observations. The asymetrical clipping may be partly responsible for the
'warm' sound associated with tube distortion. Likewise, if you look closely,
you'll see the clipping has relatively rounded edges -- another point for
tubes, which by nature clip 'soft' like that. The compression rate close to
the clip level isn't linear, but appears exponential. Solid state devices
typically don't clip in that fashion, but do so in a much harder way. Good
research, here.
--
Lamar Owen
WGCR Internet Radio
1 Peter 4:11
10:50 a.m.
Steve Harris wrote:
On Sat, Nov 09, 2002 at 04:29:11PM +0100, Tim Goetze
wrote:
better wait; i'm in the process of setting up a spice sim
of an older fender amp (super 5F4) that drives the preamp
valve differently, producing more distortion and better
data.
tim
steve, there's a file containing two saw
responses as discrete
data linked to from the page.
Excellent, thanks. I'l have a go at analysing that later.
8180
days inactive
8183
days old
linux-audio-dev@lists.linuxaudio.org
8 comments
3 participants
participants (3)
-
Lamar Owen -
Steve Harris -
Tim Goetze