What is the simplest (and/or most CPU-efficient) delay module with polyphony? Doesn’t even need feedback, though it does need to be suitable for audio rates (bog CVD seems to not sound great on audio).
I’m impementing a multi-tap delay manually (like a totally sane person), and am currently just using the VCV stock delay for the taps. Which does work, but takes up lots of space for many features I don’t need.
The perfect solution would be a small module with nothing but in/out and modulatable delay time. Polyphony would be nice to save even more space, though I guess a 2hp module with the above features would work even if mono.
I do have a hunch that in the end I’ll just be using using Memory. But I thought I’d check.
Have a look at Stocaudio Polydelay, I use this a lot although CPU usage can be quite high. Maybe it’s better with fewer channels. Instant Delay is my favourite poly delay, but that’s premium.
The Nysthi ones are not polyphonic. Lots of effects modules are tagged as poly, but just accept a poly input and sum into mono or stereo audio outs. With delays, it’s not going to give you a different delay on each channel of audio.
I’ve thought of a few more though, Airwindows is properly polyphonic and has several delays, and there is also this one if you can get it on Mac:
Thanks folks! Those seem like interesting options. I also checked out the AS ones which are not poly, but have send/return slots which offers other interesting possibilities…
The memory ensemble solution is also looking good honestly. And another sensible choice might be to just learn the 4ms tap delay thing that popped up a couple weeks ago.
I am figuring out a way to implement a glitch/artifact-free way of modulating any delay.
I sometimes like to have the “instantness” of a digital delay, without the clicks and artifacts though, like the Instant Delay by Unfiltered Audio.
After reading about it, I use a switching/crossfading system between to parallel taps.
When you change the delay time, a slope detector triggers an envelop that controls a crossfader to the second tap, which is not changed yet.
The first tap then gets changed, so any artifact will not be heard or be present in the feedback path. After the change the signal is crossfaded back. Then the second tap gets updated (in silence again) to the first tap, using a eoc trigger to a sample and hold. You can set the crossfade/switch time to any time.
It works great I believe! When I am home tonight, I will share a screenshot/patch.
Not necessarily small -it needs a second delay module and the switching patch and you have to create in and out mixers externally, but very handy.
I think you can even implement a switch to turn it on/off so you can choose between tape-style delay or clean digital.
EDIT:
It will take a bit longer, it does not work perfect on every delay yet. It needs a bit more tweaking on the patch but when I’m done, I’ll post the patch!
I haven’t used it yet, but check out this module Rotor Model A in stoermelder: Utilities for polyphonic channels - #95 by Alphagem-O - this sounds pretty neat, and since they’re spread over poly channels you could route each delayed result independently.
This is a cool idea, but I am not sure it will work. Rotator spreads the carrier signal across channels, so I think the signal occurring after the apparent delay in channel 2 is a later instance of the source than the earlier one in channel 1. Might be worth experimenting nevertheless.
As it turns out, my way of externally crossfading isn’t precise enough. It was a nice excersize though…
But -again, I found the delay of Biset very good.
If you change the phase instantly (eg through CV with a slope detector and a sample and hold) it is really accurate and artifact free. When you slew the CV instead, you can easily create a smooth tape-style smearing/pitchshifting.
But I thought I’d check.
And another sensible choice might be to just learn the 4ms tap delay thing that popped up a couple weeks ago. 