pulse buffer? averager?

Cwejman, Livewire, TipTop Audio, Doepfer etc... Get your euro on!

Moderators: Kent, Joe., luketeaford, lisa

Post Reply
User avatar
neonmercury1
Wiggling with Experience
Posts: 381
Joined: Tue Jan 10, 2017 11:15 pm
Location: Berlin

pulse buffer? averager?

Post by neonmercury1 » Thu Feb 13, 2020 7:39 pm

SO. I have been working on a performance with a dancer where i use his voice as a main input into my modular then use the processed voice to evolve and generate the patches so that the music evolves in real time. To do this am routing one of the outs from the wireless mic to preamp to the lzx sensory translator which serves as part of the nervous system for the performance. One of the element of the patch that i have been struggling with is that i have been using the high freq gate out of the lzx to tap the tempo on my tempi which controls a number of elements. the problem comes with the fact that tempi chances tempo after two taps so i am sometimes having very drastic swings in tempo. My question is then is there a way to place a module in between that can for lack of a better term "slew" the incoming pulses so that the change of tempo is happens more gradually? i cannot think of how i would do it or even how to think about it. maybe a delay in line? (i just thinking now as i write. ill try tomorrow but im using both channels of the dld for audio of this) any suggenstions or thoughts? here is my Main rack and leftovers

User avatar
starthief
Super Deluxe Wiggler
Posts: 4812
Joined: Tue Oct 04, 2016 10:39 am
Location: St Louis, MO
Contact:

Re: pulse buffer? averager?

Post by starthief » Thu Feb 13, 2020 8:13 pm

If I were going to do something like this I'd probably use Teletype, and take a running average of the milliseconds between each trigger to set a predicted tempo. (Except at the beginning, when it needs to take the first two pulses as is.) It could also be smart enough to discard outliers from the average.

There might be some tap tempo modules that do similar averaging, I'm not sure.

Marbles is pretty smart about clock sync, but I haven't tested it with a case like that... depending on how off the input timing is, it might detect an accidental pattern and then give you a weirdly irregular clock.

If you had a PLL that locked in to such a low rate, that could theoretically work, but I don't think I'd want to rely on that in a live situation.

User avatar
moremagic
Super Deluxe Wiggler
Posts: 1184
Joined: Wed Jun 23, 2010 8:45 pm
Location: NC
Contact:

Re: pulse buffer? averager?

Post by moremagic » Fri Feb 14, 2020 1:34 am

the doepfer a-196 has a LFO mode and i recall someone doing something kinda similar, using it to sync triggers from a physical drumkit to tap tempo for a delay

andybizarre
Wiggling with Experience
Posts: 397
Joined: Fri Sep 30, 2016 2:21 pm
Location: Switzerland

Re: pulse buffer? averager?

Post by andybizarre » Fri Feb 14, 2020 1:59 am

I`m not in front of the modular ATM so I can`t check it, but my first guesses were gates into slew generator, from there into a comparator. Second guess: the Doepfer A-162 Trigger Delay "forgets" incoming gates if delay and lenth values get longer, which might help in your case.

User avatar
neonmercury1
Wiggling with Experience
Posts: 381
Joined: Tue Jan 10, 2017 11:15 pm
Location: Berlin

Re: pulse buffer? averager?

Post by neonmercury1 » Fri Feb 14, 2020 4:40 am

Thanks people,
starthief wrote:
Thu Feb 13, 2020 8:13 pm
If I were going to do something like this I'd probably use Teletype, and take a running average of the milliseconds between each trigger to set a predicted tempo. (Except at the beginning, when it needs to take the first two pulses as is.) It could also be smart enough to discard outliers from the average.

There might be some tap tempo modules that do similar averaging, I'm not sure.

Marbles is pretty smart about clock sync, but I haven't tested it with a case like that... depending on how off the input timing is, it might detect an accidental pattern and then give you a weirdly irregular clock.

If you had a PLL that locked in to such a low rate, that could theoretically work, but I don't think I'd want to rely on that in a live situation.
so i think teletype although undoubtably could do this I think is not the solution for me. I have friends who love it but i find im more into the idea of the PLL and trigger delay. I think ill try using my tides pll today and see if that yields desirable results.

In this performance i have it so the clock shifts from almost nothing to fast bursts, so the performer can really manipulated it as he chooses. i would just like the transition to go slower where there is not whiplash so to speak.
andybizarre wrote:
Fri Feb 14, 2020 1:59 am
I`m not in front of the modular ATM so I can`t check it, but my first guesses were gates into slew generator, from there into a comparator. Second guess: the Doepfer A-162 Trigger Delay "forgets" incoming gates if delay and lenth values get longer, which might help in your case.
this is interesting also and sounds like it could be close to the effect i am looking for. i will also see if i can borrow the 162.

my other idea while lying in bed last night was to patch the taps from the sensory into a side of the dld then add the inverted envelope out of the same band of the taps into the feedback so that when the gates stop there will be an echo.

doing a full day practice now so have alot to try. thanks

LunaticSound
Common Wiggler
Posts: 222
Joined: Tue Dec 25, 2018 3:50 pm
Location: Hamburg

Re: pulse buffer? averager?

Post by LunaticSound » Fri Feb 14, 2020 6:09 am

I think easiest might be to use a module that uses more taps to calculate the clock. Sadly, the only module I can think of is Fourmulator right now, which is not exactly, umm, handy, but there are surely more of them around.
WTB: ARC ANN

User avatar
Risc_Terilia
Wiggling with Experience
Posts: 338
Joined: Sun Mar 20, 2016 11:50 am
Location: UK

Re: pulse buffer? averager?

Post by Risc_Terilia » Fri Feb 14, 2020 6:13 am

Won't Peaks do this on tap mode?

andybizarre
Wiggling with Experience
Posts: 397
Joined: Fri Sep 30, 2016 2:21 pm
Location: Switzerland

Re: pulse buffer? averager?

Post by andybizarre » Fri Feb 14, 2020 3:52 pm

neonmercury1 wrote:
Fri Feb 14, 2020 4:40 am
andybizarre wrote:
Fri Feb 14, 2020 1:59 am
I`m not in front of the modular ATM so I can`t check it, but my first guesses were gates into slew generator, from there into a comparator. Second guess: the Doepfer A-162 Trigger Delay "forgets" incoming gates if delay and lenth values get longer, which might help in your case.
this is interesting also and sounds like it could be close to the effect i am looking for. i will also see if i can borrow the 162.

my other idea while lying in bed last night was to patch the taps from the sensory into a side of the dld then add the inverted envelope out of the same band of the taps into the feedback so that when the gates stop there will be an echo.

doing a full day practice now so have alot to try. thanks
So I did a quick patch this evening (imitating your scenario with "jumping" clocks from a heavy modulated 4MS QCD) The slew-comparator patch didn`t work as expected, as the comparator gate stays high during bursts. The A-162 did quite well, although you`ll probably use up both sections. As with the comparator, the gate of the A-162 stays obviously high during bursts for the selected gate length (or, as you call it, while "averaging"), the second section shortens the gate to whatever length you need. A don`t know if this will work for you though, because there`s no way of emphasising or prioritization the events that truly matter during the performance. Those are just fixed timing figures which may not fit in the feel of your dancer.

flashheart
Super Deluxe Wiggler
Posts: 1807
Joined: Mon Jun 01, 2009 7:58 am

Re: pulse buffer? averager?

Post by flashheart » Fri Feb 14, 2020 6:35 pm

I think what you need to do is frequency to V conversion, then slew that CV to control another clock module. Maybe the Doepfer PLL will help.
A-196 manual. It has a built in slew limiter, as well as being able to insert an external slew. I don't have one, so those that do could fill you in further. DOn't know how slow it can go either.
Actually another way I just thought of is to use a clock divider. By only taking 1 every n clocks this will average things out, then multiply that clock back to the original range by the same amount. This will lag the changes a bit, but for your use case this probably won't matter. The generated clocks between the divided one will will equidistant, so you'll need to experiment. with the best division.
I'm not buying a maths though, not my idea of fun...

Post Reply

Return to “Eurorack Modules”