An intriguing idea for using sample & hold

Started by Mark Hammer, December 24, 2021, 12:24:05 PM

Previous topic - Next topic

Mark Hammer

Too many things to get to for me to do the R&D, but I thought I'd throw it out there for those who are adventurous enough.  I have to credit Moritz Klein's excellent Youtube channel for jogging my thinking.

Sample & hold units traditionally use a clock to feed a fixed-width pulse to the gate of a JFET.  When turned on, the gate allows a storage cap to charge up and hang onto that voltage until the next clock pulse/step.  We are most familiar with the use of such circuits in filter units, like the Maestro/Oberheim circuit, and more recent doodads like the various "robot" pedals from Sub Decay and others.

The strong-arm tactic these pedals use is that the "step-rate" is dictated by the on-board clock.  That, in itself is neither terrible or unmusical, but it IS limiting.  What if the clock rate varied with one's picking strength?  Pick/strum hard,and the filter cycles through steps more quickly.  Or better yet, what if there was a step for each note picked, as detected by an envelope-follower/trigger generator?  So, note X has one filter setting, note X+1 has another, and so on.    Or what if the sampling was dictated by a sequencer-like circuit, not unlike what Zach Vex uses for his various "Seek" effects, but instead of the little pots setting the level of an effect, they set the time a sample is held until the next step.  The noise source to be sampled remains the same, and the degree of randomness is the same, but the timing of the sampling is at the command of the player. 

How does that sound to you?  I hate having ideas that intrigue the heck out of me, and exceed the time and/or skill I can bring to the fore.  But recent off-line interactions with folks I respect tells me there are clever people out there waiting for intriguing ideas.

Merry Xmas.

EBK

Quote from: Mark Hammer on December 24, 2021, 12:24:05 PM
I hate having ideas that intrigue the heck out of me, and exceed the time and/or skill I can bring to the fore.  But recent off-line interactions with folks I respect tells me there are clever people out there waiting for intriguing ideas.
There's a thread full of such ideas looking for people.
https://www.diystompboxes.com/smfforum/index.php?topic=121960.0
  • SUPPORTER
Technical difficulties.  Please stand by.

mozz

Probably nothing about what you are thinking about, I've got some of these if anybody wants to try. Some funky waveforms might be possible. Gated op-amp could be triggered somehow by level or freq?
https://www.mouser.com/catalog/specsheets/intersil_fn2856.pdf
  • SUPPORTER

RickL

QuoteOr better yet, what if there was a step for each note picked, as detected by an envelope-follower/trigger generator?

I'm pretty sure Boscorelli released a project that did something similar to this. I think it was a panner that sent each note (or, selectably, each 2 or 4 notes) to alternate outputs. IIRC I tried to build it but didn't have much success.

Ripthorn

I think EQD do something in the night wire trem that changes rate with picking strength, so your idea is not to far out there.
Exact science is not an exact science - Nikola Tesla in The Prestige
https://scientificguitarist.wixsite.com/home

iainpunk

now i want to experiment with S&H again. i did a snare drum sound once with S&H where an audio freq osc sampled a MHz range sqare wave osc for a ''random high/low'' computer 8 bit distortion, but the biggest problem was the volume, it was constant.
a similar system with a guitar signal's every positive zero-crossing triggering a sample of either white noise or that MHz osc. would probably sound like a more chaotic octave down divider.

cheers
friendly reminder: all holes are positive and have negative weight, despite not being there.

cheers

parmalee

Quote from: Mark Hammer on December 24, 2021, 12:24:05 PM
Sample & hold units traditionally use a clock to feed a fixed-width pulse to the gate of a JFET.  When turned on, the gate allows a storage cap to charge up and hang onto that voltage until the next clock pulse/step.  We are most familiar with the use of such circuits in filter units, like the Maestro/Oberheim circuit, and more recent doodads like the various "robot" pedals from Sub Decay and others.

The strong-arm tactic these pedals use is that the "step-rate" is dictated by the on-board clock.  That, in itself is neither terrible or unmusical, but it IS limiting.  What if the clock rate varied with one's picking strength?  Pick/strum hard,and the filter cycles through steps more quickly.  Or better yet, what if there was a step for each note picked, as detected by an envelope-follower/trigger generator?  So, note X has one filter setting, note X+1 has another, and so on.    Or what if the sampling was dictated by a sequencer-like circuit, not unlike what Zach Vex uses for his various "Seek" effects, but instead of the little pots setting the level of an effect, they set the time a sample is held until the next step.  The noise source to be sampled remains the same, and the degree of randomness is the same, but the timing of the sampling is at the command of the player. 

Interesting.  Not sure if I'm completely out to lunch here, but your description actually reminds me a bit of Fairfield Circuitry's Shallow Waters <<< pedal.

Obviously, it takes a very different approach--using a delay (a BBD, I think), with timing controlled by pseudo-randomness (an lfsr, presumably), and a voltage controlled gate/filter--but the results might not be wholly dissimilar.

OK, I suppose the results are apt to be entirely dissimilar, but not entirely in a different ballpark (?!).  An intriguing idea all the same.

Digital Larry

This sort of stuff is easy to do on the FV-1. In one case I took the sampled random value, low pass filtered it to introduce a little lag, then mixed that back into the signal controlling the ramp LFO rate.  I also made a 4 "phase" sample/hold that updated one thing after the other.  You can send any value to anything - filter freq, Q, delay time, LFO rates, etc. 

I'm not sure if something like that results in better music or not.   :icon_mrgreen:
Digital Larry
Want to quickly design your own effects patches for the Spin FV-1 DSP chip?
https://github.com/HolyCityAudio/SpinCAD-Designer

Mark Hammer

You shame me, sir!  :icon_cry:  Man oh man, do my digital chops need to catch up.

parmalee

Quote from: Mark Hammer on December 28, 2021, 04:31:18 PM
You shame me, sir!  :icon_cry:  Man oh man, do my digital chops need to catch up.

Sure, but after the EMP--or, I suppose, a continuous succession of EMPs--who is going to get the last laugh?

Mark Hammer

If I'm not mistaken, our friend Ton "Puretube" Barmentloo.

ElectricDruid

Quote from: Mark Hammer on December 28, 2021, 05:22:37 PM
If I'm not mistaken, our friend Ton "Puretube" Barmentloo.

Don't forget Metasonix!

Mark Hammer

Given some of the names of their products, Tom, I think they've had their laugh already!