An idea for simpler-type delays some may want to try

[Mark Hammer]

Some digital delay pedals come with what is referred to as "ducking delay".  "Ducking" devices were traditionally used by radio announcers/DJs.  The device monitored the voice-mic level, and when it was above some threshold the music level would be turned down, returning to normal broadcast level when the announcer stopped talking.  Much like broadcast ducking, the ducking setting on a delay turns the wet signal down as you play, and brings it back up when you stop playing such that repeats are heard when there is sufficient "space" to hear them.  At the same time, it doesn't let the sound get too cluttered if you're playing lots of notes that get mixed up with lots of repeats, leaving the repeats for when they can best be appreciated.

I was thinking this morning that a simple envelope follower, tapped off the input stage of a delay, whether analog or some simple PT2399-based circuit, could drive a vactrol (homebrew or commercial) that would govern the output level of the wet channel, such that DIY delays could avail themselves of the advantages of ducking.

I guess the idea is that the wet path has an inverting op-amp stage, just before the "mix"/wet-level control.  Inverting op-amps can achieve "negative" gains when the feedback resistance is greater than the input resistance.  So, if one had a unity-gain inverting stage with input R = 100k and feedback R = 100k, an LDR in parallel with the feedback R could reduce the gain below unity, as the LED got pushed brighter and drove the LDR resistance downward.  Naturally, the time constants - especially release/decay - would have to be selected prudently such that the return of the wet signal was not too abrupt, and the LDR dark value selected so that the combined parallel feedback resistance, when not playing, remains very close to the nominal value of the fixed resistor.

One of the perks this adds is that typical delay pedals that have a ducking option do not let one determine how much the wet signal is reduced, in proportion to picking strength.  Having a "sensitivity" control for the envelope follower, or a variable resistance in series with the LDR, to set maximum feedback reduction, would permit finer adjustment  of how the wet signal is combined with dry.

What I have not yet wrapped my head around is what use of an inverting stage might complicate.  Normally, if one combines an inverted and non-inverted version of a signal, they cancel each other out.  But delaying the inverted version means there is never two identical-but-opposite signals to combine.  Nonetheless, there might be some unwanted cancellations.  I guess that would be overcome by simply using a DPDT toggle to completely bypass the inverting stage and revert back to the original circuit.

I'm up to my eyeballs in current projects, soI won't embark on this any further at the moment.  But I thought I'd mention it so that those of you more intrepid types with a little more time and a much shorter to-do list might explore it.

Thanks for your ear.  Now back to my very first attempt to tweed-cover a cabinet.

[Ripthorn]

Very cool idea. I'm working on an crazy PT2399 design right now, so this won't make it into my queue, but I love the ingenuity.

[anotherjim]

The envelope detector might use the spare pins13,14 opamp in the PT2399, since they aren't exactly hi-fi parts but, being CMOS, are rail-rail. The audio path can then use a better audio opamp for the post-filter from pin12.

[ElectricDruid]

Yeah, I agree. Interesting idea, thanks for sharing it. It doesn't sound so hard to do, either.

I also agree that I don't think there's a problem with putting an inversion in the wet path. After all, anything over a few milliseconds will be more than half a wavelength for our audio frequencies anyway. Example: 50Hz = 20msecs period, therefore 10msecs= out of phase. And that's at the low end where it's more noticeable. As soon as the delay is over 50/100msecs or so, what "phase" it is doesn't really make sense to ask. It's turned into a comb filter at that point and it'll be in phase at some frequencies and out of phase at others, but as long as the delay is long enough, we'll hear it separated in time and won't hear the comb filter effect so much.

[Sooner Boomer]

Sounds a lot like how Dolby noise reduction works.

[PRR]

..."Ducking" devices were traditionally used by radio announcers/DJs......

Supermarkets. Duck the Muzak when announcements are made. Airports, snack bars, bus depots....

Not radio stations. They were required to have a technical person on site. FCC/CRTC, also union contracts, and just because manual balance control is better; and radio revenues can cover it. (Yes, today is a different world; likely to lose the human DJ, everything an algorithm.)

[Nasse]

Little OT but for some time I have been using compressor again (for years did not) and local instro forum said put the compressor AFTER the tape delay. Works nice imho

[Mark Hammer]

..."Ducking" devices were traditionally used by radio announcers/DJs......

Supermarkets. Duck the Muzak when announcements are made. Airports, snack bars, bus depots....

Not radio stations. They were required to have a technical person on site. FCC/CRTC, also union contracts, and just because manual balance control is better; and radio revenues can cover it. (Yes, today is a different world; likely to lose the human DJ, everything an algorithm.)

Quite right that anywhere that a person might speak into a mic, and override whatever else was playing in the background, would be likely to use a ducker.

But while FCC/CRTC (thanks for the Can-con addition!  ) regs might have mandated someone on hand, based on my own recollection, they'd have to have lightning-fast reflexes to manually adjust levels with the sort of interruptions I was used to hearing.  I'm sure there were circumstances where they actually did ride the faders, but little snippy comments by DJs, over records, would happen too fast, and subside every bit as quickly, for any human to make suitable adjustments in time.  Had to be an automatic device/circuit in those instances.

Which brings me back to the point that, if one does attempt such an add-on, the onset of the wet signal (which is actually the release time of the wet gain-reduction/suppression) has to be set judiciously.  Onset can be quick, but return has to be somewhat gradual. 

In some respects, a perfect version would adjust return/release time contingent on delay.  So short delays would require a shorter release time such that repeats could be heard before they completely fade out, while repeats of longer delays could tolerate a longer "waiting period" before they faded in.

[anotherjim]

There's no such thing as live radio anymore in the truest sense. Some delay (seconds) is used so they have a chance of cutting anything bad before we hear it. This began with phone-ins but if you can do it for one show, why not do it for all of them? This means that technical glitches can often be ironed out.

[imJonWain]

Off topic but that's not totally true.  Most college and community radio stations still run live outside of DJs who want to pre-record/mix their shows.  The station I volunteer and sometimes DJ at is definitely done live and we do all fading/leveling manually.

[Mark Hammer]

On-topic, can folks suggest any other kinds of effects in which a ducking function might add some value?  That need not be a matter of fading in the effect overall, in the absence pf picked/strummed notes.  For instance, a ducking circuit could be used to add a voice-like emphasized vibrato after one stops strumming and lets the strings ring.  Ducking can be used to make many sorts of modulation effects back off and then become more evident as one gives them space.

One interesting quasi-ducking effect is the old Gretsch Contrafuzz.  Normally, notes played through a distortion or fuzz started out having gobs of added harmonic content, and then simmer down as the string dies out.  The Contrafuzz mixed a clean signal and a seriously boosted/clipped version of it in a way that started out clean but then appeared to get fuzzier as the string/note decayed (hence the name).  It did notuse any sort of sidechain to accomplish this, but the effect was almost the same: the fuzz came into the picture after you stopped picking.

My point is that ducking is not only applicable to delay, but to other categories of effects as well.  Let's keep this conversation going.

[Fancy Lime]

Long, intense reverb seems to me to be an even more obvious application for ducking than delay.

Andy

[Mark Hammer]

I agree.  After all, why else do the really good ones often include a pre-delay control?

[PRR]

> FCC/CRTC (thanks for the Can-con addition!  ) .... lightning-fast reflexes .... little snippy comments by DJs, over records

Ah. That's just the master limiter. Pot the record to +4VU. Yell over that, say +10VU. The console has the headroom. The transmitter does not, but all broadcast radio has a fancy limiter before the transmitter.

The CRTC is news to me. I know where the FCC fits in the USA. Someone has the same functions in Canada. It was amusing to learn that at first the main broadcaster (what we now know as CBC?) made the rules.

[Mark Hammer]

The "main broadcaster" was the public broadcaster.  So, in many respects it was only logical that the broadcaster citizens were paying for (remember, no commercials/ads at that time) set out the ground rules.  When I was a kid, we had 3 television networks: CBC, French-language CBC, and newcomer CTV.  In town we had maybe 4 or 5 AM radio stations; again English and French CBC, and a few others.  With the introduction of more competition, as well as more communication mediums (FM radio, internet, cellphones, etc.), CBC became one small voice in a crowd.  I know the person who was, when I last looked, the legal representative for the CRTC, having moved there from a similar position at the Competition Bureau (an agency devoted to preventing monopolies).

[PRR]

> When I was a kid

That was just about the time they quit self-regulating. (See US AEC/NRC both promoting and regulating atoms; CAA/FAA doing as much for air travel.)

The CRBC took over a network of radio stations formerly set up by a federal Crown corporation, the Canadian National Railway. ...On November 2, 1936, the CRBC was reorganized under its present name. While the CRBC was a state-owned company, the CBC was a Crown corporation on the model of the British Broadcasting Corporation....
For the next few decades, the CBC was responsible for all broadcasting innovation in Canada. This was in part because, until 1958, it was not only a broadcaster, but the chief regulator of Canadian broadcasting. It used this dual role to snap up most of the clear-channel licences in Canada....
https://en.wikipedia.org/wiki/Canadian_Broadcasting_Corporation#History

I did not know that greedy US broadcasters had left ANY clear channels for CBC to snap up.

[bartimaeus]

I agree.  After all, why else do the really good ones often include a pre-delay control?

getting further off topic, but pre-delay is there to change how large the reverberant space feels, not to make space for the dry signal. it emulates how, in larger rooms, it takes longer for any sound to reach the walls and reflect back to the listener. of course, that doesn't mean it can't be used for ducking!