Speaking of delay apps..I have a delay question

FriedTapeworm

Is there a delay on iOS that has a retrigger type of effect. For instance, if I hit a note and then a second note a little later, the first note delay stops instead of continuing its full delay tail. (Sorry if my terminology sucks, I'm a musician but a production noob).

oat_phipps

The term that approximates what you are looking for is ducking. The delay more or less "gets out of your way" when an envelope follower detects a note being played and then is allowed to come back in to full wet level as long as you're not playing. However, it doesn't explicitly "re-trigger" per note as you describe (ie if the feeeback is high enough so that its still repeating when the delay kicks back in after your second note, you'll still hear the trails of the first note). I do not know of a re-triggering delay. Fabfilter Timeless 2 can do the ducking thing in Auria, but off the top of my head none other comes to mind. Someone will know for sure. I'm surprised the ducking is not a feature of more delays.

Some of the more well-versed users could probably help you set something up like described above by using the FAC envelope (Envolver) app in conjunction with any delay you have, but that kind of rigging, well it's not beyond me, but I'm too lazy for it.

FriedTapeworm

Ok thanks. I’ll take a look at the delays on the market specifically for ducking feature. I appreciate the detailed response. I would imagine that’s something I could figure out in Drambo with enough tinkering and a little research.

jolico

Sidechain

rs2000

@FriedTapeworm said:
Ok thanks. I’ll take a look at the delays on the market specifically for ducking feature. I appreciate the detailed response. I would imagine that’s something I could figure out in Drambo with enough tinkering and a little research.

You can build all kinds of delays (except reverse delays) in Drambo.
From your description it's not clear to me though what exactly that delay should do.

espiegel123

@rs2000 said:

@FriedTapeworm said:
Ok thanks. I’ll take a look at the delays on the market specifically for ducking feature. I appreciate the detailed response. I would imagine that’s something I could figure out in Drambo with enough tinkering and a little research.

You can build all kinds of delays (except reverse delays) in Drambo.
From your description it's not clear to me though what exactly that delay should do.

He wants echoes of the old material to stop when the input signal crosses a threshold so that the echo of the new signal will just be from the new notes.

One way of accomplishing this would be an envelope follower, two separate delays and a Mozaic or screenbyter script to do some clever switching.

Setup you could use, input into an envelope follower. Input also sent to Delay A and Delay B.
Initially delay A is active and delay b is bypassed. When the input crosses the threshold, a script toggles A and B so that B is active and A bypassed. The script would also set Delay being bypassed’s feedback to 0 then restore it to its original value (in order to kill old echoes).

When switching, one might he want a crossfade rather than a sudden switch.

rs2000

@espiegel123 So if you have 10 new notes in any succession above that threshold, you'd need 10 delays?
Or rather a delay that can simply have its buffer cleared by an external trigger?

espiegel123

@rs2000 said:
@espiegel123 So if you have 10 new notes in any succession above that threshold, you'd need 10 delays?
Or rather a delay that can simply have its buffer cleared by an external trigger?

With what I outlined, you switch back and forth between two delays. The reason for using two rather than just clearing and resetting the buffer is to provide a way to soften the transition with a crossfade. Also because the slight latency might result in some of the attack being cut off if you simply use one delay and reset the buffer.

rs2000

@espiegel123 said:

@rs2000 said:
@espiegel123 So if you have 10 new notes in any succession above that threshold, you'd need 10 delays?
Or rather a delay that can simply have its buffer cleared by an external trigger?

With what I outlined, you switch back and forth between two delays. The reason for using two rather than just clearing and resetting the buffer is to provide a way to soften the transition with a crossfade. Also because the slight latency might result in some of the attack being cut off if you simply use one delay and reset the buffer.

Sounds reasonable but then you'd need a delay that can have its buffer cleared.

espiegel123

@rs2000 said:

@espiegel123 said:

@rs2000 said:
@espiegel123 So if you have 10 new notes in any succession above that threshold, you'd need 10 delays?
Or rather a delay that can simply have its buffer cleared by an external trigger?

With what I outlined, you switch back and forth between two delays. The reason for using two rather than just clearing and resetting the buffer is to provide a way to soften the transition with a crossfade. Also because the slight latency might result in some of the attack being cut off if you simply use one delay and reset the buffer.

Sounds reasonable but then you'd need a delay that can have its buffer cleared.

I think setting feedback to zero effectively does that..though this is a thought experiment and I could be wrong. I am not in a situation where I can test at the moment.

aleyas

@espiegel123 said:

@rs2000 said:

@espiegel123 said:

@rs2000 said:
@espiegel123 So if you have 10 new notes in any succession above that threshold, you'd need 10 delays?
Or rather a delay that can simply have its buffer cleared by an external trigger?

With what I outlined, you switch back and forth between two delays. The reason for using two rather than just clearing and resetting the buffer is to provide a way to soften the transition with a crossfade. Also because the slight latency might result in some of the attack being cut off if you simply use one delay and reset the buffer.

Sounds reasonable but then you'd need a delay that can have its buffer cleared.

I think setting feedback to zero effectively does that..though this is a thought experiment and I could be wrong. I am not in a situation where I can test at the moment.

I believe feedback must be reduced to 0 for the same duration as the delay length itself for the buffer to be completely cleared. Reducing feedback to 0 momentarily would only erase a portion of the feedback.

Perhaps 3 or 4 delays connected to a 1-N switch, fed round robin style would be sufficient for the buffers to be cleared in the preceding delays..

rs2000

@aleyas said:

@espiegel123 said:

@rs2000 said:

@espiegel123 said:

@rs2000 said:
@espiegel123 So if you have 10 new notes in any succession above that threshold, you'd need 10 delays?
Or rather a delay that can simply have its buffer cleared by an external trigger?

With what I outlined, you switch back and forth between two delays. The reason for using two rather than just clearing and resetting the buffer is to provide a way to soften the transition with a crossfade. Also because the slight latency might result in some of the attack being cut off if you simply use one delay and reset the buffer.

Sounds reasonable but then you'd need a delay that can have its buffer cleared.

I think setting feedback to zero effectively does that..though this is a thought experiment and I could be wrong. I am not in a situation where I can test at the moment.

I believe feedback must be reduced to 0 for the same duration as the delay length itself for the buffer to be completely cleared. Reducing feedback to 0 momentarily would only erase a portion of the feedback.

Perhaps 3 or 4 delays connected to a 1-N switch, fed round robin style would be sufficient for the buffers to be cleared in the preceding delays..

Yes, that's the point.

espiegel123

@rs2000 said:

@aleyas said:

@espiegel123 said:

@rs2000 said:

@espiegel123 said:

@rs2000 said:
@espiegel123 So if you have 10 new notes in any succession above that threshold, you'd need 10 delays?
Or rather a delay that can simply have its buffer cleared by an external trigger?

With what I outlined, you switch back and forth between two delays. The reason for using two rather than just clearing and resetting the buffer is to provide a way to soften the transition with a crossfade. Also because the slight latency might result in some of the attack being cut off if you simply use one delay and reset the buffer.

Sounds reasonable but then you'd need a delay that can have its buffer cleared.

I think setting feedback to zero effectively does that..though this is a thought experiment and I could be wrong. I am not in a situation where I can test at the moment.

I believe feedback must be reduced to 0 for the same duration as the delay length itself for the buffer to be completely cleared. Reducing feedback to 0 momentarily would only erase a portion of the feedback.

Perhaps 3 or 4 delays connected to a 1-N switch, fed round robin style would be sufficient for the buffers to be cleared in the preceding delays..

Yes, that's the point.

Oops, you are right. If the notes are going to come in with space between them longer than the delay time then when toggling a delay off, one can set both feedback and delay time to 0 and wait a buffer's length of audio to set them back. But that obviously doesn't work if the notes are coming in with closer spacing.

rs2000

@espiegel123 said:

@rs2000 said:

@aleyas said:

@espiegel123 said:

@rs2000 said:

@espiegel123 said:

@rs2000 said:
@espiegel123 So if you have 10 new notes in any succession above that threshold, you'd need 10 delays?
Or rather a delay that can simply have its buffer cleared by an external trigger?

With what I outlined, you switch back and forth between two delays. The reason for using two rather than just clearing and resetting the buffer is to provide a way to soften the transition with a crossfade. Also because the slight latency might result in some of the attack being cut off if you simply use one delay and reset the buffer.

Sounds reasonable but then you'd need a delay that can have its buffer cleared.

I think setting feedback to zero effectively does that..though this is a thought experiment and I could be wrong. I am not in a situation where I can test at the moment.

I believe feedback must be reduced to 0 for the same duration as the delay length itself for the buffer to be completely cleared. Reducing feedback to 0 momentarily would only erase a portion of the feedback.

Perhaps 3 or 4 delays connected to a 1-N switch, fed round robin style would be sufficient for the buffers to be cleared in the preceding delays..

Yes, that's the point.

Oops, you are right. If the notes are going to come in with space between them longer than the delay time then when toggling a delay off, one can set both feedback and delay time to 0 and wait a buffer's length of audio to set them back. But that obviously doesn't work if the notes are coming in with closer spacing.

I've already requested "buffer clearance" and drop-off crossfade as a new delay feature, let's see what happens 😊

branis

A simple way to achieve this is to use velocity to fake the delay effect:

dobbs

or, for fraction of a second, set both feedback to zero and delay time to as small a value as possible. Or. make a new note deactivate and reactivate the plugin... I think some clever fiddling with envolver/mosaic/... might get you there...

colonel_mustard

It sounds like you want a 'note repeat' midi effect, rather than a delay - assuming that you're using software instruments and not tracking from a guitar/mic. It might be worth looking at arpeggiators too, for these kind of rhythmic effects.

I love delays, but midi fx seem a bit neater for this use.

Would definitely try out @branis' suggestion too. If you know how you want the repeats to sound, automation can probably get you there.

espiegel123

@dobbs said:
or, for fraction of a second, set both feedback to zero and delay time to as small a value as possible. Or. make a new note deactivate and reactivate the plugin... I think some clever fiddling with envolver/mosaic/... might get you there...

Btw, the three delays I tested on (two from Nembrini and wrapped AU delay) don’t clear the buffer if you set feedback and delay to 0 only briefly before resetting to the desired setting. It seems that need to remain at 0 for a buffer length at the size you set it back to.

dobbs

Huh that's interesting. I also just noticed the delay/reverb tail reappearing in velvet when modulating its duration knob...

FriedTapeworm

Thanks for the discussion! I will give some of these things a try in Drambo and see how it turns out.

FriedTapeworm

I haven't tried it yet, but Nembrini's Delay3000 has a ducking feature according to the app store description.

https://apps.apple.com/app/id1503345580#?platform=iphone

eross

i’d totally buy a delay like that

NeuM

@jolico said:
Sidechain

That’s what I was going to say also.

wim

On of my first Mozaic scripts was MIDI echo Mozaic that might be of interest for this. As others have mentioned, one way to go about this is to use notes triggered at decreasing velocity rather than an audio echo. Velocity Echo re-triggers notes after the intervals you enable and at the velocity you set. This can be used to make interesting rhythmic variations.

But the part that made me mention it is: If a new note comes in, it interrupts the echo cycle of the one before it. This ensures that the second note's "echos" are heard and the first are ended.

https://patchstorage.com/velocity-midi-echo/