Dynamically changing Delay Line

Jul 28, 2010 at 9:09pm

Dynamically changing Delay Line


Does anyone know how to create a dynamically changing delay line using tap in tap out where it doesn’t distort the audio too much. I am trying to create continuous non-linearity in the string. Excuse me if I don’t have the terminology correct.



Jul 28, 2010 at 9:27pm

Are you modulating the delay line at audio rate? That’s the first step. The second would be to smooth the changes in the modulating signal with onepole~/slide~/rampsmooth~ etc… if necessary. This depends on how you’re modulating though, if you’re using an lfo (cycle~/phasor~ etc…) then a smoothing filter won’t be necessary.

I’m not well-versed in physical string modelling (sounds like what you’re doing), so I can’t offer much advice on exactly how to go about achieving the non-linearity that you want.

Jul 28, 2010 at 9:40pm

Nice one Tim. I’ll give that a try. What I have tried to do is modulate the delay length of a Karplus strong with a another Karplus Strong so that it kind of fades out and eventually stops changing the delay length but it’s still a bit glitchy. Cheers anyway!

Jul 28, 2010 at 11:26pm

I’m afraid modulating the delay line with the audio rate distorts it even more.I’m going to have to figure something else out. Thanks for your help anyway.

Jul 28, 2010 at 11:50pm

Could you post an example? You should be able to modulate tapout~ at audio rate with no glitches/distortions. Maybe you’re modulating so fast you’re creating FM sidebands, hard to say without seeing a patch.

– Pasted Max Patch, click to expand. –
Jul 29, 2010 at 9:04am

tim is right as usual. in the meantime, the poletti m4l example is sitting in your patches folder and works great: Max5/patches/m4l-patches/Tools resources/M4L.vdelay~.maxpat

Jul 29, 2010 at 1:17pm

Nice one guys. The patch I have attached is part of a bigger patch. I’m trying to physically model a sitar as part of my masters thesis. All the main inner workings of my model are contained in the patch I have attached.

  1. string.maxpat
Jul 29, 2010 at 1:20pm

Actually might as well include the whole thing for clarity. At the moment I have it set to modulate the tapin tap out with not using the audio rate. Everything is in the string1 subpatch

Jul 29, 2010 at 1:53pm

I made some tiny changes that seem to have stopped the distortion, but I may be missing something; it’s a big patch! The bits I changed are outlined in red.

Jul 29, 2010 at 3:41pm

That’s perfect Tim!! Thank you so much, you’ve saved me many a panic attack. My thesis is due in a few weeks.

Sep 8, 2010 at 9:45pm

interested to hear the final thing did you do more to it?


Sep 15, 2010 at 4:31pm

I have indeed, submitted my thesis last week. Please find attached and enjoy.

  1. Sitar.zip
Sep 15, 2010 at 6:06pm


Sep 15, 2010 at 6:58pm

great stuff.

you forgot to include the “sitar.mid” file?…

Sep 15, 2010 at 7:01pm

It’s not worth listening to really, very simple demo of its sound. Although, I did do a composition using it. Its not very obvious I used it though because of all the time stretching. http://soundcloud.com/davidronan

It’s called “Called Anon Tiro”

Sep 15, 2010 at 7:02pm

really good work, thanks for posting. Will you make the thesis available too?

it’s a shame that the tuning is compromised in the higher registers. A bit better with sigvs @ 4, but then the cpu goes through the roof.


Sep 15, 2010 at 7:26pm

I will indeed, I still have a week to make a few changes to the writing, I’ll post it when it is finished.

Ya the thing eats cpu big time, it’s a pity.


You must be logged in to reply to this topic.