Peter Simonsen
Well-known member
Kev,
Thanks you ;-)..That did it.. :thumb:
Kind regards
Peter
Thanks you ;-)..That did it.. :thumb:
Kind regards
Peter
Very nice dynamic processor schematic.
- Thread starter ijr
- Start date
Help Support GroupDIY Audio Forum:
Well, seems some errors on schematic can be, I redrawn the schem from pcb file of someone's clone I've seen elsewhere on Net 
)
Going to compare to TD of different version I can get for some days.
And no, I am not going to build TD, just was very curious how this thing works.
Some people did it, anyway, some time before.
)Going to compare to TD of different version I can get for some days.
And no, I am not going to build TD, just was very curious how this thing works.
Some people did it, anyway, some time before.
clintrubber
Well-known member
All fine & thanks for making this schematic available. Never had a look at those S**ndPr*ss*r*L*v*l-brand-devices before... looks nice. Release = Sustain, right ?
Given the alike audio-path, at first it looks like a nice alternative sidechain for the SSL-comp, but with said audiopath so minimal it'd simple just better be added as well, giving a complete unit. Hmm, all simple parts, but lots of em...
Given the alike audio-path, at first it looks like a nice alternative sidechain for the SSL-comp, but with said audiopath so minimal it'd simple just better be added as well, giving a complete unit. Hmm, all simple parts, but lots of em...
chrissugar
Well-known member
"Never had a look at those S**ndPr*ss*r*L*v*l-brand-devices before... "
Most of the products of this company are mediocre or just good but this is an absolutely fantastic device. You can shape the sound of drums, loops and bass with only two knobs and the results are great. One of my favorite tools.
chrissugar
Most of the products of this company are mediocre or just good but this is an absolutely fantastic device. You can shape the sound of drums, loops and bass with only two knobs and the results are great. One of my favorite tools.
chrissugar
chrissugar
Well-known member
Igor,
I really appreciate your knowledge and I have nothing against telling people the truth about things (like your transformer measurements).
Although I think I think there is no need of sarcasm and offending comments.
chrissugar
I really appreciate your knowledge and I have nothing against telling people the truth about things (like your transformer measurements).
Although I think I think there is no need of sarcasm and offending comments.
chrissugar
clintrubber
Well-known member
Would this kind of circuit work well for increasing sustain of elec-guitars ?
My loved Jazzmaster with its floating bridge comes to mind
- would be cool if such a circuit placed pre-overdrive could
keep the sound longer going and at a comparable degree of overdrive.
Usually a comnpressor placed pre-overdrive is suggested then, but
I'm wondering if the circuit of this thread has benefits over using a normal compressor.
Last but not least, is this circuit able to more or less sustain decaying sounds over seconds ?
Thanks,
Peter
My loved Jazzmaster with its floating bridge comes to mind
- would be cool if such a circuit placed pre-overdrive could
keep the sound longer going and at a comparable degree of overdrive.
Usually a comnpressor placed pre-overdrive is suggested then, but
I'm wondering if the circuit of this thread has benefits over using a normal compressor.
Last but not least, is this circuit able to more or less sustain decaying sounds over seconds ?
Thanks,
Peter
chrissugar
Well-known member
I didn't try it on guitars but I suppose it would work nicely.
It is a completely different type of compressor, it has nothing in common with the classic five knob (attack, release, threshold, ratio, makeup) compressors.
While the five knob type apply the processing from the threshold point this will do the same processing at all levels.
You can increase the attack of a bass drum or make it completely disappear without affecting the sustain of the sound. Also the sustain knob can make the body of the sound to be huge or leave only the attack of the sound. You have tons of variations with only two knobs.
chrissugar
It is a completely different type of compressor, it has nothing in common with the classic five knob (attack, release, threshold, ratio, makeup) compressors.
While the five knob type apply the processing from the threshold point this will do the same processing at all levels.
You can increase the attack of a bass drum or make it completely disappear without affecting the sustain of the sound. Also the sustain knob can make the body of the sound to be huge or leave only the attack of the sound. You have tons of variations with only two knobs.
chrissugar
clintrubber
Well-known member
Thanks for the added info.
If I understand it correctly it's not only more practical wrt opertion (no threshold etc), but can also accomplish things that the normal boxes can't.
Sounds like a fun device !
Also looked around for some more, but while that sustain-increasing for longer gtr-sounds seemed an obvious application I didn't see it being mentioned anywhere.
So I'm wondering how long this circuit could keep that deyaing sound 'up in the air'. Range seems enough (24dB if I recall it correctly), but it's the time-axis I'm wondering about. Well, I'll study the circuit a bit more for this.
Thanks,
Peter
If I understand it correctly it's not only more practical wrt opertion (no threshold etc), but can also accomplish things that the normal boxes can't.
Sounds like a fun device !
Also looked around for some more, but while that sustain-increasing for longer gtr-sounds seemed an obvious application I didn't see it being mentioned anywhere.
So I'm wondering how long this circuit could keep that deyaing sound 'up in the air'. Range seems enough (24dB if I recall it correctly), but it's the time-axis I'm wondering about. Well, I'll study the circuit a bit more for this.
Thanks,
Peter
Analysing the first stage of the sidechain.
If the drawing is right, this is a high gain BPF realised with unconventional means.
Low frequencies are filtered with C1 and R2. 20k and 100nF make 80Hz.
Opamp practically works with its open loop gain, with larger signals clipped to approx. +/-500mV. Open loop gain of TL074 is a LPF function, of course, so overall we have a BPF.
PSpice might not be terribly accurate about open loop gain (and open loop gain may vary with different TL074's, differenent temperature etc.), but it may give a good estimation:
At 500Hz, we have the maximum gain of 49dB.
3dB corners relative to this are 75Hz and 3kHz.
Does this BPF function make sense?
The LED driver after this first stage looks strange. Maybe a drawing error.
IC 1C and 1D are a standard cheap full wave rectifier.
The interesting stuff comes after this. That's where they extract certain aspects of the dynamics, and alter it. That's what this circuit is supposed to do, isn't it: Change attacks and decay *relative* to the original attack and decay ...
JH.
If the drawing is right, this is a high gain BPF realised with unconventional means.
Low frequencies are filtered with C1 and R2. 20k and 100nF make 80Hz.
Opamp practically works with its open loop gain, with larger signals clipped to approx. +/-500mV. Open loop gain of TL074 is a LPF function, of course, so overall we have a BPF.
PSpice might not be terribly accurate about open loop gain (and open loop gain may vary with different TL074's, differenent temperature etc.), but it may give a good estimation:
At 500Hz, we have the maximum gain of 49dB.
3dB corners relative to this are 75Hz and 3kHz.
Does this BPF function make sense?
The LED driver after this first stage looks strange. Maybe a drawing error.
IC 1C and 1D are a standard cheap full wave rectifier.
The interesting stuff comes after this. That's where they extract certain aspects of the dynamics, and alter it. That's what this circuit is supposed to do, isn't it: Change attacks and decay *relative* to the original attack and decay ...
JH.
[quote author="jhaible"]Analysing the first stage of the sidechain.
If the drawing is right, this is a high gain BPF realised with unconventional means.
Low frequencies are filtered with C1 and R2. 20k and 100nF make 80Hz.
Opamp practically works with its open loop gain, with larger signals clipped to approx. +/-500mV. Open loop gain of TL074 is a LPF function, of course, so overall we have a BPF.
PSpice might not be terribly accurate about open loop gain (and open loop gain may vary with different TL074's, differenent temperature etc.), but it may give a good estimation:
At 500Hz, we have the maximum gain of 49dB.
3dB corners relative to this are 75Hz and 3kHz.
Does this BPF function make sense?
[/quote]
Wait a minute, JH. These diodes are not there to limit anything.
There's no level potentiometer anywhere, and the VCAs have an expo response. So there should be a log function in the sidechain somewhere.
So the diodes around IC1a form a simple log(x) function!
(Then, of course, the above mentioned BPF response would only apply for very small signals, i.e. not really relevant.)
I was quite puzzled by this, because I've seen log rectifiers implemented the other way round, first a rectifier, then a log circuit. Now in this circuit, a crude log function is applied on each half wave separately, by the two diodes, and then the result is rectified.
Makes more sense?
This thing is really interesting. I read some reviews, which make me want to build one!
JH.
If the drawing is right, this is a high gain BPF realised with unconventional means.
Low frequencies are filtered with C1 and R2. 20k and 100nF make 80Hz.
Opamp practically works with its open loop gain, with larger signals clipped to approx. +/-500mV. Open loop gain of TL074 is a LPF function, of course, so overall we have a BPF.
PSpice might not be terribly accurate about open loop gain (and open loop gain may vary with different TL074's, differenent temperature etc.), but it may give a good estimation:
At 500Hz, we have the maximum gain of 49dB.
3dB corners relative to this are 75Hz and 3kHz.
Does this BPF function make sense?
[/quote]
Wait a minute, JH. These diodes are not there to limit anything.
There's no level potentiometer anywhere, and the VCAs have an expo response. So there should be a log function in the sidechain somewhere.
So the diodes around IC1a form a simple log(x) function!
(Then, of course, the above mentioned BPF response would only apply for very small signals, i.e. not really relevant.)
I was quite puzzled by this, because I've seen log rectifiers implemented the other way round, first a rectifier, then a log circuit. Now in this circuit, a crude log function is applied on each half wave separately, by the two diodes, and then the result is rectified.
Makes more sense?
This thing is really interesting. I read some reviews, which make me want to build one!
JH.
clintrubber
Well-known member
from ijr:
Interested to read about the outcome of your further investigations :thumb:Well, seems some errors on schematic can be, I redrawn the schem from pcb file of someone's clone I've seen elsewhere on Net
Going to compare to TD of different version I can get for some days.
And no, I am not going to build TD, just was very curious how this thing works.
Some people did it, anyway, some time before.
clintrubber
Well-known member
FWIW, found this:
DFP-plugs are nice, just wasn't aware they can do alike things (if that's right). Hmm, I guess toying around with Dominion could at least give some sort of an impression.
DFP-plugs are nice, just wasn't aware they can do alike things (if that's right). Hmm, I guess toying around with Dominion could at least give some sort of an impression.
chrissugar
Well-known member
Hey Peter,
Some weeks ago I made a test, the Dominion against the real thing.
Same wave files with similar timeconstant setings and similar volumes.
You could recognise that the plugin was simulating the caracter of the hardware, but that was all. The difference is day and night. The hardware is much more effective and also the sound is organic, not a static, uninteresting crap.
chrissugar
Some weeks ago I made a test, the Dominion against the real thing.
Same wave files with similar timeconstant setings and similar volumes.
You could recognise that the plugin was simulating the caracter of the hardware, but that was all. The difference is day and night. The hardware is much more effective and also the sound is organic, not a static, uninteresting crap.
chrissugar
mike_relay
Well-known member
[quote author="chrissugar"]Hey Peter,
Some weeks ago I made a test, the Dominion against the real thing.[/quote]
Chris, any chance you could post or send me some audio samples? I'm very intrigued by this.
-mike
Some weeks ago I made a test, the Dominion against the real thing.[/quote]
Chris, any chance you could post or send me some audio samples? I'm very intrigued by this.
-mike
clintrubber
Well-known member
Thanks Chris,
Good to hear all this - always feels good the hardware has its merits.
But I also understand from your words that the basic idea could be made to work with that plugin, and that definitely be a lot faster than soldering all those TL07Xs on a piece of Veroboard now :thumb:
Hi Mike, about audio-samples, have a look here:
http://www.soundperformancelab.com/Transient_Designer/in_kuerze.html
and here:
http://www.amazona.de/content/musictools/software/plugIn/transient/transient.htm
Too bad though all demo's I heard so far concern non-ac. drums - since it'd be interesting to hear it for other kind of sources as well (like elec-gtr pre-OD) but haven't heard impressions about that. So far one could get the impression this circuit is the ticket for dance & beats, but how about the rest of the world ? (in which I happen to live :wink: )
All fine.
Regards,
Peter
Good to hear all this - always feels good the hardware has its merits.
But I also understand from your words that the basic idea could be made to work with that plugin, and that definitely be a lot faster than soldering all those TL07Xs on a piece of Veroboard now :thumb:
Hi Mike, about audio-samples, have a look here:
http://www.soundperformancelab.com/Transient_Designer/in_kuerze.html
and here:
http://www.amazona.de/content/musictools/software/plugIn/transient/transient.htm
Too bad though all demo's I heard so far concern non-ac. drums - since it'd be interesting to hear it for other kind of sources as well (like elec-gtr pre-OD) but haven't heard impressions about that. So far one could get the impression this circuit is the ticket for dance & beats, but how about the rest of the world ? (in which I happen to live :wink: )
All fine.
Regards,
Peter
I updated some misc things.
Will have some time at "shabat" to concentrate on this again.
JH, input circuit is just kinda diode limiter like in guitar distortion.
Just logarithmic (limiting) function before rectifier to allow the thing work in large dynamic range..
Will have some time at "shabat" to concentrate on this again.
JH, input circuit is just kinda diode limiter like in guitar distortion.
Just logarithmic (limiting) function before rectifier to allow the thing work in large dynamic range..
[quote author="ijr"]I updated some misc things.
Will have some time at "shabat" to concentrate on this again.
JH, input circuit is just kinda diode limiter like in guitar distortion.
Just logarithmic (limiting) function before rectifier to allow the thing work in large dynamic range..[/quote]
I've put the whole sidechain into PSpice last night. It's not easy to simulate this, because you need looooong transient analysis to see what's going on.
I used a tone burst of 1kHz sine, going from 100mV to 1V amplitude with different attack times (1ms, 50ms, 500ms) and back to 100mV after 500ms, also with different decay times.
The attack part of the side chain produces rounded spikes - short and high ones for fast transients, longer and smaller ones for medium transients, nothing for slow transients. It looks very reasonable - easy to see how this can be used to mangle the attack portion via VCA.
The Release part is not so easy. It creates signals that appear to be in a reasonable voltage range, and the shape also varies with different transient speed, but it's not complementary to the Attack signals as I had expected.
Well, this is supposed to work on the sustain phase of a transient, not on the release (?), so maybe the drawing and my modeling is right. (I had made some errors first in copying the schematics to Spice - it's a *tricky*, very unusual circuit!)
I'll test it with more percussive signals tonight, to see if it brings up the sustain. I guess my AR-shaped (trapezoid) envelope wasn't the right thing to evaluate the "Release" (Sustain?) part. I'll try an AD-shape next.
And yes, the input part of the side chain is a log rectifier. No doubt about this anymore.
JH.
Will have some time at "shabat" to concentrate on this again.
JH, input circuit is just kinda diode limiter like in guitar distortion.
Just logarithmic (limiting) function before rectifier to allow the thing work in large dynamic range..[/quote]
I've put the whole sidechain into PSpice last night. It's not easy to simulate this, because you need looooong transient analysis to see what's going on.
I used a tone burst of 1kHz sine, going from 100mV to 1V amplitude with different attack times (1ms, 50ms, 500ms) and back to 100mV after 500ms, also with different decay times.
The attack part of the side chain produces rounded spikes - short and high ones for fast transients, longer and smaller ones for medium transients, nothing for slow transients. It looks very reasonable - easy to see how this can be used to mangle the attack portion via VCA.
The Release part is not so easy. It creates signals that appear to be in a reasonable voltage range, and the shape also varies with different transient speed, but it's not complementary to the Attack signals as I had expected.
Well, this is supposed to work on the sustain phase of a transient, not on the release (?), so maybe the drawing and my modeling is right. (I had made some errors first in copying the schematics to Spice - it's a *tricky*, very unusual circuit!)
I'll test it with more percussive signals tonight, to see if it brings up the sustain. I guess my AR-shaped (trapezoid) envelope wasn't the right thing to evaluate the "Release" (Sustain?) part. I'll try an AD-shape next.
And yes, the input part of the side chain is a log rectifier. No doubt about this anymore.
JH.
