Making a generic envelope follower

[Vsat]

A.S.P.,
Looks like I sent the previous message  almost the same time as you!
As Harry related to me, even for a single string the overtones are not perfectly harmonic. This is the case for many mechanical systems and has to do with the physics of the string/plate/bell etc vibration, different vibrational modes and non-linear behavior of the vibrating medium.  Waveforms/spectra  from electrical oscillators are usually quite simple in comparison, since all the overtones are perfectly harmonic... not what usually happens in nature.
Regards, Mike

[varialbender]

Couldn't you just set a frequency range depending on your instrument, detect which note is loudest, and follow its envelope? I'm a n00b and don't know much, so I'm more than likely wrong. I don't know how you detect which is loudest, but there are some boxes that do that right? You know, the ones that get reviews for their "tracking"? Or is that just how it's sonically perceived, while electronically doing something much different?

Set me straight

[davebungo]

Dave and A.S.P.,
Yup... that's what I think is happening within the env follower. But with two strings playing nearly the same note you also have  beating taking place in a linear environment... the oscilloscope (and my ears) clearly show a periodic beat envelope before the signals are applied to the follower... simply a result of two nearly identical waves slowly going in and out of phase with each other, producing periodic reinforcement and cancellation at the difference freq. Since the two freqs are not the same (even if the amplitudes are), the null will not be perfect at the time of maximum cancellation, and the null becomes even less intense as the freq difference becomes larger. But the null can be nearly perfect when the two frequencies and amplitudes are very closely matched, giving a large change in overall level at the beat rate, which the follower does a nice job of "following". No need to invoke non-linearity of the air or non-linearity of the ear to explain acoustic beating. It is additive, and will behave the same at low or high sound pressure levels (or by simple summing in an op amp circuit). The rectifier in the env follower is non-linear however, and multiplication takes place and additional terms are generated. as you mention. So the beating has two separate origins in the follower circuit.
Regards, Mike

Yep, sorry, i see your point - didn't mean to lecture you or anything.  I've seen a similar effect in a metering application I worked on but it was more related to the lack of samples of the signal and the signal being close to a sub-multiple of the sampling rate.  The samples "rotate" slowly at the difference between the actual frequency and the nearest sub-multiple of the sampling rate.  e.g. if sampling at 48KHz and you meter a 4KHz ish signal the envelope beats up and down at a low rate.  There isn't much which can be done in this case.  Your scenario of two closely related signals summing and cancelling is similar and I don't think can really be avoided.  I'm not sure it would cause major problems though in reality - it may even sound good.

[Vsat]

Varialblender,
Your idea of selecting the loudest note only would be a big improvement, don't know how you'd do it though, DSP might be the best way.

Dave, I just mentioned the acoustic beat phenomena to indicate it was another source of trouble separate from the beating which occurs during rectification.  Audio books talk about linear mixers, while RF books talk about a very different kind of non-linear mixer, I seldom see acoustic beats discussed anywhere.

Yup, this is mostly academic, people have been using env flwrs for years now, just seems it would be nice to have one where playing additional notes would only increase the overall envelope level without adding low-freq ripple.

And it would be really neat to see a box where you plug in your axe, and each individual note gets auto-wah'd or individually fuzzed etc without interaction with the other notes.... sort of turn your ordinary pickup into a hex pickup with a dedicated set of fx per string. Any digital multi-fx boxes out there that can do this....?
Regards, Mike

[Peter Snow]

Hi Mike,

Well my Roland GR-300 used to do that (not digital).  Hex pickup (Ibanez ZIMG), each string processed separately by the GR-300 with fuzz and a filter section and a VCO dedicated to each string.  That was a really cool box.  I'm sorry I sold it now...
Cheers,

Peter

[Peter Snow]

Oops! As usual I misread your post .  You were asking about a device that would allow a regular pickup to act like a hex pickup.  Sorry...

(GR-300 is still cool though )

Peter

[puretube]

made a hex-pickup 25 years ago from 6 reed-relay coils (can`t find a pic).

One could multiplex/modulate them (on-board), and transmit that (HF-)signal via the normal cable (which could also carry DC for phantom-supply) "on top" of the normal signal from the normal pickups...
(and then filter out normal and multiplexed signals, using the latter for pitch-to-voltage,
trigger, and envelope purposes,
while using the normal PU`s for "sound".

[Vsat]

Peter,
Seen a number of GR-300 etc but have never heard one in person (AFAIR). I just think that such a facility would make a great addition to those do-everything boxes out there.  Might not be so simple though 

OK here's the solution... each string is paired up with a 2nd (much higher pitched) string, both are picked simultaneously, the normal set of 6 strings provides the audio, while the new set of 6 high strings feeds the env flwr for greatly reduced ripple   
That's the ticket.....   
Mike

[StephenGiles]

That's got them thinking Mike - Roger McGuinn will be OK!!
Stephen

[Peter Snow]

Geeez Mike,

I have enough trouble negotiating 6 strings! 

Peter

[Vsat]

Franchises coming soon to a plaza near you!   
Mike

[StephenGiles]

Mike, assuming you are 5 hours behind us how do you manage to stay up until 5am or have you just got up!!!
Stephen

[Vsat]

Steve,
Your observational skills are finely honed   
That's what (sometimes) happens when I get a circuit idea in the wee hours.
Luckily, this doesn't happen too often...
Nothing to do with env followers (might be LFO-related though) 
Cheers, Mike

[A.S.P.]

information lost...

[MR COFFEE]

Missed this thread earlier somehow. Oh yeah, on vacation. GREAT THREAD, guys.

Anyone around here familiar with the RNC (Real Nice Compressor) ? I've been told it uses an old motorola microprocessor (like a 68705R4 or something) to generate the envelope follower signal controlling the VCA... and I would suspect using a method not too far removed from Harry's.

If you use a FWR followed by a resetable peak detector like Harry used in his EDN Design Idea, and sample the voltage into a uP with a ADC-input channel every 1-5 ms or so, then the software can keep up with all of Harry's peak-selection processing (and then some - maybe even use an exponential decay prediction algorithm to tame down those low Hz beat frequencies Harry's worried about), and then output the control signal on a DAC pin into a simple medium-short time constant RC smoother that can also have it's decay rate sped up if you mute the strings by using another uP pin to turn on a transistor switch to add a lower value resistor to ground across the "smoother" cap. Practically no ripple, low-distortion, responsive envelope-follower circuit. Would do it all, yes?

Maybe not one IC, but should get it down to maybe 2 (uP and quad op amp) and a dozen passive components. Isn't there a PIC version with a 8-or-so bit ADC and DAC built-in that comes in a compact, low pinout DIP? Seems like I remember one. No?

Also - Harry, do you have a schematic laying around of the one you made using the CMOS analog switches that you would be willing to share? I'd like to see it, too. You seem to be a very inventive fellow. Glad to see you around here!

[H^)harry]

Hello Mr. Coffee

see... speak of the devil and you get a quick reply. Now I have to figure out how to
include the schematic you requested. If I understand correctly, I have to link from
some other website.  I'm working on it...
H^) harry

[Transmogrifox]

I think the only way to do it is to stick a cat in a box with a spring-loaded, grounded volume pedal...and guitar amp output connected to its tail.  When the dynamic level gets too high, the cat jumps off the volume pedal.  When it calms down a bit, the cat relaxes and beds back down on the volume pedal.  You have forward and reverse depending on which way you spring-load the volume pedal

The cat won't respond to beat frequencies, because if the dynamics are loud, it's getting zapped and it's time to move off the pedal.

[MR COFFEE]

Hi Harry,

I dunno if it's a link or just a personal mail thing. You said

There are other ways to do this circuit.   I did a version using a 4053 CMOS switch to run the
sample holds AND the resets.  Its a little more compact.

and it got my curiousity up just trying to figure exactly how it was you hooked that up and made it compact. It sounds pretty cool.

Do you see any problems with doing the logic and peak holds with a small uP with an 8 bit ADC-DAC?

My cat's not that talented


Thanks,

[Vsat]

Mr Coffee & list,
Harry forwarded me the schemo for the CD4053 version of his envelope follower and asked if I could post it (he tried but his machine crashed in the process).

Now just how do you post the gif....
Anyone?
Cheers, Mike

[H^)harry]

Hola Mr. Coffee

Once Vsat or I figure out how to post, I'll send the schematic.  I use the 4053 to pass the
signal to the hold caps, and the other part of the switch to dump the charge. It requires some
low leakagew diodes to work really well, but does not have the voltage drop of my original circuit. That might be a blessing or curse.

I'm using a 4093 and a 4017 for the gating signals, you might do better. Still looks like 4 ICs so its
not stompbox ready.

Bop over to wiseguysynth.com and get the schematic for my "morphlag" circuit to finish the job. Do it
quickly because the owner (Larry Hendry) passed away in a tragic motorcycle accident and it might disappear any time.  Larry was a great member of the DIY scene and he'll be missed, and often.

H^)harry