Envelope-controlled fade-in of reverb/echo.

[Mark Hammer]

I just picked up a copy of Eric Johnson's most recent release, "Bloom".  One of the things I noticed about his tone is his use of delays.  In many instances, the wet signal only comes in once the main note has decayed substantially.  I started thinking about this, and it occurred to me that a suitable add-on box for accomplishing this would be pretty trivial.

Here's the idea:

• Straight guitar goes into inverting op-amp and into inverting op-amp mixer stage (i.e., comes out in phase with input)
• Feed from delay unit goes to second inverting op-amp gain stage, and from there into same mixer
• Tap from clean input stage goes to gain stage and envelope follower to opto-isolator (NSL32, CLM6000, etc.)
• LDR half of optoisolator goes in feedback loop of delay unit stage

One of the properties of inverting op-amps is that the gain is a function of Rfeedback/Rinput.  So, as the feedback resistance drops below the input resistance, the gain becomes less than 1.  I see the wheels starting to turn in your mischievious little heads already.

So, imagine we have the op-amp stage for the delay input set for a gain of 2 with fixed resistors (let's say Rf=100k and Ri=47k).  When the guitar is silent, the LDR resistance sneaks up into the megohms, so that when it is paralleled with Rf, the feedback resistance is still pretty much 100k.  Play something, though, and the LDR resistance will be made to go low, which will reduce the combined parallel resistance of the LDR and 100k resistor.  That will have the net effect of attenuating the delay signal being fed to the mixer stage.  Stop playing, and the LDR resistance gradually rises, bringing the gain of that delay input stage back up.

There you go.  Automatic delayed delay.  I can see where this would be useful for fading in reverb or echoes.  Incidentally, since a great many delay pedals have separate dry and wet outputs, this conceptual circuit could easily be an add-on box.  The resulting output would be mono, but thesonic flexibility of the delay pedal would be extended and permit useful decluttering.  The Tone Core (Line 6) Echo Park and the Delay Modeller already have this capacity as the "ducking" program, but there is no reason why those of you/us who love analog shouldn't have that creative tool as well.  A Deluxe Memory Man, DM-2, PT-80 or AD3208 need not be denied this capacity.

I'll need to whip up a circuit and post it, but there have to be at least a few dozen among you that could probably beat me to it.  This should not require more than a quad op-amp.  Input impedance may be an issue but that can easily be solved with a discrete stage or another dual op-amp for input buffers.

(BTW, I'm not exactly batting 1000 with respect to remembering things I've posted here before.  Have I already suggested such a circuit and merely reinvented a wheel?  )

[StephenGiles]

What's this Mark - no gardening to do!!!!! Sounds like a great idea.

[Mark Hammer]

Did it all last week.  Time to play now.

Actually, now that I think of it, fading in of a second channel needn't be restricted to delay.  With the threshold set right, and a fairly fast decay, one could hold off on, say, the phase-shifted signal for a moment or two and then bring it in for the remainder of sustained notes.  That sounds kind of interesting.

[A.S.P.]

http://www.diystompboxes.com/smfforum/index.php?topic=34918.0

[Processaurus]

You got me thinking about a couple variations on the theme of envelope controlled delay.  One is using the envelope to control the delay's feedback, so when you're playing, its low and things don't get too busy and garbled, and when you stop, it gets cranked up for like the guy in dub bands that just plays with knobs on the echoplex would do. 

The reverse of the ducking delay would be great too, where the delay signal gets mixed in when you're playing, but when you stop it drops out, to get the big delay dimension (which can be natural sounding) when you're playing, but avoid the telltale delay trail, that says, check it out, I use delay.  Gated delay?  Sidechain expanded delay?  I kind of try to do that to vocals if someone's left me in charge of running the board at shows.  Commercially the EH holier grail reverb has an option that does that, which is what sold me on it, but the gate was too awkward sounding.

RG's current controlled amplifier article at Geofex could help these ideas come to light.

[Mark Hammer]

http://www.diystompboxes.com/smfforum/index.php?topic=34918.0

Again?!  Jeez Louise, we gotta stop meeting like this. 

The proposed circuit is certainly a cousin of what is in that thread, but it provides envelope-controlled mixing of one channel, rather than envelope controlled panning between them.

Okay, here is a rough draft of what would be involved: http://img.photobucket.com/albums/v474/mhammer/DelayFade-in.png
Like I suggested earlier, phase coherence is maintained by using inverting stages throughout.  Given that input impedance is better for non-inverting configurations, I ought to have usednon-inverting inputs.  The thing of it is that we want the LDR to produce gains LESS than 1 and the noninverting configuration won't do that - but the inverting one will.  Once you have one section inverting, you sort of get roped into making the others inverting as well.

Note the following suggested controls.  First, you have level controls for each of the two channels.  Second, you have a pot for setting the gain/sensitivity of the envelope follower.  Third, you have a large smoothing capacitor with a Release-time (Decay rate) pot.  This actually sets the fade-in time.

The way it is proposed to work is by sudden attack transients from the dry level immediately lighting up the LED and driving the LDR resistance low.  This effectively attenuates the second channel almost completely, depending on the sensitivity setting.  As you play softer, the current stored in that electrolytic smoothing cap after the diode slowly or quickly drains off, depending on the release setting, lighting the LED less and less.  As the LED goes dark, the LDR half of the optoisolator goes up in resistance again, fading in the second channel, to some level determined by the pot.  Component values remain to be worked out but the design should work, in principle.  Missing from the draft is all of the power supply related circuitry.

[nelson]

Could quite easily be an envelope controlled panner.



Just duplicate the effect in inverting opamp section for the dry signal and put the LDR as Rin.