Best Envelope Detector circuit?

[WhenBoredomPeaks]

So i am thinking about building a standalone envelope detector box.

I would mod my LFO driven modulation effects (chorus, phaser etc.) to accept CV signals from an external jack and then i would control them with the envelope detector instead of their internal LFO.

I am thinking about using the envelope detector part of an existing effect (probably an autowah) but which one should i choose?

(btw i am just speculating here but an envelope generator is just takes the envelope of my AC guitar signal and makes a DC signal from it with the "same" envelope?)

[egasimus]

Check out R.G.'s articles.

http://www.geofex.com/article_folders/ecftech/ecftech.htm
http://www.geofex.com/Article_Folders/noteproc/noteproc.htm

Since you're building an external box, you might just go all the way and build Harry Bissell's envelope follower.
http://www.edn.com/article/490267-Envelope_follower_combines_fast_response_low_ripple.php

Whatever envelope detector you build, it will probably benefit from a frequency doubler in front of it.

[WhenBoredomPeaks]

Check out R.G.'s articles.

http://www.geofex.com/article_folders/ecftech/ecftech.htm
http://www.geofex.com/Article_Folders/noteproc/noteproc.htm

Since you're building an external box, you might just go all the way and build Harry Bissel's envelope follower.
http://www.edn.com/article/490267-Envelope_follower_combines_fast_response_low_ripple.php

Whatever envelope detector you build, it will probably benefit from a frequency doubler in front of it.

Whoa thank you for those links!

It feels kinda wrong to have a more complex circuit for envelope detection than the circuits you will control with it.

[deadastronaut]

+1 cheers adam, could have done with that a while back... thanks.

[StephenGiles]

Check out R.G.'s articles.

http://www.geofex.com/article_folders/ecftech/ecftech.htm
http://www.geofex.com/Article_Folders/noteproc/noteproc.htm

Since you're building an external box, you might just go all the way and build Harry Bissel's envelope follower.
http://www.edn.com/article/490267-Envelope_follower_combines_fast_response_low_ripple.php

Whatever envelope detector you build, it will probably benefit from a frequency doubler in front of it.

Whoa thank you for those links!

It feels kinda wrong to have a more complex circuit for envelope detection than the circuits you will control with it.

Why? It is infinitely more complex to extract a good ripple free (??) envelope following CV than to VCA or VCF the audio. That is why in my view 99% of autofilters sound so awful when a longer sweep is used.

Now that I'm 20% retired I have a little more time available - but not tomorrow as I'm testing out my new Raynox 2.2x teleconverter lens - thankyou Amazon for delivering in 20 hours!!

[petemoore]

  Which envelope.
  Very simple ['in concept'] could be an electronic pick-trigger triggering when it connects to ground [guitar strings should be grounded].
  "Best" is when the control input produces the desired effect as it controls the...effect, this part requires the input signal to be within the parameters which the ED responds to...in a certain way. Could be the Boss Envelope Filter or something DIY as the definition of 'best' varies a lot.
   If 'best' means ''better than the alternatives'' then some experimentation/testing/tweeking [or just study>choose?] may be needed to define the circuit.

[WhenBoredomPeaks]

Check out R.G.'s articles.

http://www.geofex.com/article_folders/ecftech/ecftech.htm
http://www.geofex.com/Article_Folders/noteproc/noteproc.htm

Since you're building an external box, you might just go all the way and build Harry Bissel's envelope follower.
http://www.edn.com/article/490267-Envelope_follower_combines_fast_response_low_ripple.php

Whatever envelope detector you build, it will probably benefit from a frequency doubler in front of it.

Whoa thank you for those links!

It feels kinda wrong to have a more complex circuit for envelope detection than the circuits you will control with it.

Why? It is infinitely more complex to extract a good ripple free (??) envelope following CV than to VCA or VCF the audio. That is why in my view 99% of autofilters sound so awful when a longer sweep is used.

Now that I'm 20% retired I have a little more time available - but not tomorrow as I'm testing out my new Raynox 2.2x teleconverter lens - thankyou Amazon for delivering in 20 hours!!

Because i don't "care" too much about the "silent workhorses" of the circuits, the LFOs and envelope detectors. I thought i get the envelope section of Dr. Q or something similar or if i want more tweakability then i use MuTron III's follower but now it looks like i need like 4 opamps and 2 logic chips just for the follower.

[Mark Hammer]

So i am thinking about building a standalone envelope detector box.

I would mod my LFO driven modulation effects (chorus, phaser etc.) to accept CV signals from an external jack and then i would control them with the envelope detector instead of their internal LFO.

I am thinking about using the envelope detector part of an existing effect (probably an autowah) but which one should i choose?

(btw i am just speculating here but an envelope generator is just takes the envelope of my AC guitar signal and makes a DC signal from it with the "same" envelope?)

There are several distinct aspects of amplitude envelope use, and they tend to interact with each other in ways one needs to think about more broadly.

For example, envelope-control is fundamentally about time as much as it is about sweep.  Different control elements (OTA/FET vs LDR, etc) have different response times.  But then, the signal we are attempting to process is not steady state and has fluctuating content over time.  In the case of guitar, much of the treble content is confined to the first few tenths of a second after picking.  Having a fast attack allows us to act on that treble content and shape it as we wish.  But often LDRs can not respond quite that fast.  OTAs and FETs can certainly respond that fast, but are subject to distortion that LDRs are relatively immune to.

The other end of the guitar signal has the nasty habit of causing ripple.  Part of it s the manner in which the string vibrates, but I gather another part of it is the tendency of the envelope signal to fall just above and below the forward voltage of whataver diodes are being used to rectify the signal.  At least that's my hunch.  One of the nice things about LDRs is that their natural sluggishness, relative to OTAs and FETs, results in rejection, or rather ignoring of that ripple.  So OTAs and FETs might get you that pick attack sweep more effectively, but LDRs will provide a more satisfying decay.

This is getting all quite complicated for 8:30AM so I'll stop here.  Suffice to say that there is no "ideal" envelope detector whose virtues stand apart from what it is you are trying to control with it, what circuitry you use to perform that control, and what role the envelope control is intended to have with the given instrument and music.

[Jarno]

The McMeat meatball clone has a pretty flexible envelope follower, you might use that.

Other options can be found in the synth arena: ExperimentalistsAnonymous Schematic Archive

[WhenBoredomPeaks]

The McMeat meatball clone has a pretty flexible envelope follower, you might use that.

Other options can be found in the synth arena: ExperimentalistsAnonymous Schematic Archive

I bought a Meat Sphere pcb (Meatball clone also) from musicpcb.com yesterday, i think i will mod it to have it's envelope out connected to a jack (switchable) so i can control other effects with it instead of the Meat Sphere's filter. I have to bypass the other sections of the effect somehow but that can't be hard, when i will have the pcb in my hands i will figure it out.

[WhenBoredomPeaks]

So i am thinking about building a standalone envelope detector box.

I would mod my LFO driven modulation effects (chorus, phaser etc.) to accept CV signals from an external jack and then i would control them with the envelope detector instead of their internal LFO.

I am thinking about using the envelope detector part of an existing effect (probably an autowah) but which one should i choose?

(btw i am just speculating here but an envelope generator is just takes the envelope of my AC guitar signal and makes a DC signal from it with the "same" envelope?)

There are several distinct aspects of amplitude envelope use, and they tend to interact with each other in ways one needs to think about more broadly.

For example, envelope-control is fundamentally about time as much as it is about sweep.  Different control elements (OTA/FET vs LDR, etc) have different response times.  But then, the signal we are attempting to process is not steady state and has fluctuating content over time.  In the case of guitar, much of the treble content is confined to the first few tenths of a second after picking.  Having a fast attack allows us to act on that treble content and shape it as we wish.  But often LDRs can not respond quite that fast.  OTAs and FETs can certainly respond that fast, but are subject to distortion that LDRs are relatively immune to.

The other end of the guitar signal has the nasty habit of causing ripple.  Part of it s the manner in which the string vibrates, but I gather another part of it is the tendency of the envelope signal to fall just above and below the forward voltage of whataver diodes are being used to rectify the signal.  At least that's my hunch.  One of the nice things about LDRs is that their natural sluggishness, relative to OTAs and FETs, results in rejection, or rather ignoring of that ripple.  So OTAs and FETs might get you that pick attack sweep more effectively, but LDRs will provide a more satisfying decay.

This is getting all quite complicated for 8:30AM so I'll stop here.  Suffice to say that there is no "ideal" envelope detector whose virtues stand apart from what it is you are trying to control with it, what circuitry you use to perform that control, and what role the envelope control is intended to have with the given instrument and music.

This makes me wish there would be an envelope "creator" device which could be triggered by my pick attack and then you could have the familiar (from the synth world) ADSR knobs to set the shape of the envelope. It would be pretty much like a synth without velocity control.

Two things: this could make only one kind of envelope and there could be problems with setting the treshold of the trigger. (i mean a barre chord could trigger the envelope while a hammer on not and similar problems)

Again, this is harder than i thought initially.

[Mark Hammer]

Well there you're starting to verge on "intelligent" envelope interpretation, such that the needs of the player are adjusted on the fly.  And of course, that is NOT going to happen with a couple of op-amps, some diodes and the right value of cap, is it?

[StephenGiles]

And of course you are not going to stuff it into a tiny tiny box - we are possibly talking rackmount for the best option!!!

[Gurner]

This makes me wish there would be an envelope "creator" device which could be triggered by my pick attack and then you could have the familiar (from the synth world) ADSR knobs to set the shape of the envelope. It would be pretty much like a synth without velocity control.

Two things: this could make only one kind of envelope and there could be problems with setting the treshold of the trigger. (i mean a barre chord could trigger the envelope while a hammer on not and similar problems)

I do a fair bit with envelopes and uCs.....the trigger threshold as you describe wouldn't be an issue (both occurences - ie barre chod and hammer on have plenty of level....which is what you want for a good adsr trigger).....where it gets difficult is triggering well when playing softly......it's the age old problem of having the trigger set low enough to capture/trigger with really low volume (soft picked) notes but high enough to avoid false triggers by the noise floor.

[Mark Hammer]

In the analog domain, part of the challenge is the kind of diodes we tend to work with.  We almost always work with Si type, whose higher forward voltage provides some challenges in terms of having a "sensitive" envelope follower.

But here's the rub.  We've imported our interests from the synth world, and where the synth world has traditionally been of the +/-12v or +/-15v world, guitar stuff has pretty uniformly been of the +9v world.  Seems kind of strange that we haven't exploited the possibilities of other diode types like Schottkys, instead of hemming ouorselves into sub-optmal performance with silicon.

Or am I missing something important?

[PRR]

> It feels kinda wrong to have a more complex circuit for envelope detection than the circuits you will control with it.

It is a much tougher/complicated job.

Level detectors are often the bulk of a good audio limiter.

Fairchild 660 has just two triodes (implemented as several parallel) in the audio path, and a monster power amp, exotic rectifier, complex filtering to control them.

Modern broadcast limiters use a digital computer, because their analog envelope computers were just getting out of hand.

> build Harry Bissell's envelope follower.

That EDN site gets more annoying.

Bissell's is interesting. I note that it is not full-wave and may not handle asymmetrical waves "properly". The sudden-release is in fact "random" because the 150Hz clock is not synced to the input wave. It has significant "diode error", although that's often unimportant in a limiter. It isn't that complicated for the quick-release feature it emphasizes: 4017, two TL074, three transistors, thirteen diodes.

> Best Envelope Detector circuit?  

No such thing. Best car? Best food? Best town?

> control them with the envelope detector instead of their internal LFO

The control swing from the envelope follower will in general NOT be a direct replacement for an LFO scheme. LFOs tend to bounce from above ground to below the rail. Envelope followers tend to stand on ground.

[WhenBoredomPeaks]

It is a much tougher/complicated job.

Ok, i got that

No such thing. Best car? Best food? Best town?

Ok, yeah true.

The control swing from the envelope follower will in general NOT be a direct replacement for an LFO scheme. LFOs tend to bounce from above ground to below the rail. Envelope followers tend to stand on ground.

Yeah but they can go below the rail/0V only in bipolar supplied circuits right? In a 9V battery powered effect LFOs are "swinging around" VRef/4.5V i guess.

[StephenGiles]

Best to use bipolar supply - used in Craig Andertons's Bi Filter Follower which is as good as it gets with such basic circuitry.

[StephenGiles]

You could do worse, as I have often pontificated, than drive the Start Frequency buffer of the EH Space Drum envelope generator with an "Adaptive Peak Following" voltage extracted from the pluck strength. You would then have a sweep which starts at a point which follows the strength of the note played, which would decay at an independant rate set by the Sweep Rate pot.

The Space Drum circuit has now been converted to LM13600 so parts are available.

[WhenBoredomPeaks]

Thanks for the comments!

While the topic gets deeper and deeper into the world of envelope followers, i have to do a reality check here.

Is there any point in what i try to do? Like controlling a phaser or chorus with an envelope. I am quite sure it wouldn't really work with a chorus, but i don't know about other modulation effects. (phaser, flanger etc.)