delay pedal only active when pressed...?

Started by joySeeing, November 27, 2015, 05:32:48 PM

Previous topic - Next topic

joySeeing

I run woodwind instruments through effects. One annoying thing about guitar effect pedals in general is they use on / off switches that must be pressed for each state. I want the functionality of a piano pedal where the effect is only on with the pedal is down, giving a lot more control to the user.

Something tells me this is very easy, a matter of modding the switch mechanism. Although I don't really know because I've never done this before. Would it be fairly easy to mod an existing delay pedal to somehow accept a 1/4 piano sustain pedal as the device to complete the circuit instead of a on / off stomp switch?

Or maybe there is already a pedal to do this?

Or maybe better to whip out my arduino + Macbook Pro + Supercollider Lang?


R.G.

It's a common, if seldom recognized issue.

Down at the meta level, footswitches combine some amount of switching, all mechanically ganged together, with either memory (that's the alternate action, on-punch-off-punch-on... version) or no memory (i.e. what you're asking for, on only when pressed). You're asking for what is known as "momentary" switching, not "alternate action".

Whether it's easy or needs a substantial reworking for the pedal to do what you're asking for depends on the specific pedal. Different ones use different ways to get signal switching. If the switching is all wrapped up in one mechanical footswitch, then your problem reduces down to finding a similar switch but with alternate action. That can be easy, difficult and expensive, or impossible, depending on the exact mechanical footswitch. If your pedal uses some form of electronic switching triggered by the switch mechanism, it gets easier, but no more definitive. Each pedal designer dreams up his own circuit for doing bypassing, and the electronic versions may have many different ways of doing this.

So the answer is, sadly, it depends on the pedal. There isn't a simple one-size-fits-all setup other than making an external box that does what you want and leaving the pedal "on" all the time.

Which pedal do you have?
R.G.

In response to the questions in the forum - PCB Layout for Musical Effects is available from The Book Patch. Search "PCB Layout" and it ought to appear.

drummer4gc

You could make a simple true bypass loop pedal with a momentary DPDT switch. Like this: https://www.pinterest.com/pin/413557178257626646/ (anyone feel free to edit to embed the picture, can't seem to do it from my phone)

Keep the effect pedal always on, stomp the looper when you want to hear the effect.

ashcat_lt

Quote from: drummer4gc on November 27, 2015, 06:02:06 PM
You could make a simple true bypass loop pedal with a momentary DPDT switch. Like this: https://www.pinterest.com/pin/413557178257626646/ (anyone feel free to edit to embed the picture, can't seem to do it from my phone)

Keep the effect pedal always on, stomp the looper when you want to hear the effect.
This is very close to the "one-size-fits-all" answer.  Problem is, though, that with a delay you probably want the tails to keep going and fade out naturally after you take your foot off.  That is going to require a bit more than a passive switch and some jacks.  It'll want at least a buffer and a passive mixer, but probably prefer an active mixer also.  It's not super complicated, but...

joySeeing

Thanks for the answers.

Yeah, with a bypass, the tails will always be created.

Think like a piano. The second you release the sustain pedal, that sound disappears for ever. It's a hard-cut off.

I have a Boss DD-5. Even if that was easy to mod with a different switch, I guess I would still have a problem with tails being pulled over into the next delay activation. That is... you have delay active and create this big wash of sound, then turn it off and everything starts to fade away naturally (not hard-cut off). But if you reactivate it fast enough, the stuff that was fading out will be thrown in the next delay trail, which is unfortunate. I wish this would not happen. Although I may be able to live with that by adjusting feedback, delay length settings to manage it better.

The ideal pedal for me would be a boss dd-5 except one that has a "momentary" switch instead of the stomp on stomp off, adjustable fade out time when deactivated (say 0.00 seconds (hard cut off) to 5.00 seconds or until it naturally expires), one that doesn't bleed a previous trail into the next one upon reactivation of the effect.

Maybe there is already a pedal that does that but has to be modded with a momentary switch (since basically no delay pedal uses this method, it seems)?

R.G.

Your preferences for tails makes a great deal of sense musically. But it's fairly complicated to do (or at least do well) in the guts of the bypassing circuits. There are several subtle things to consider in doing this properly.

So it's hard. I don't know of any pedal that does this as you ask, or is simple to mod for it, other than the possibilities of reprogramming the DSP in some of the digital modelling delays - and the makers of these don't want to talk about how to do that.

You said:
QuoteThe ideal pedal for me would be a boss dd-5 except one that has a "momentary" switch instead of the stomp on stomp off, adjustable fade out time when deactivated (say 0.00 seconds (hard cut off) to 5.00 seconds or until it naturally expires), one that doesn't bleed a previous trail into the next one upon reactivation of the effect.
Doing tails on a delay requires you mute the input to the delay section, but leave the output of the delay still mixed in with the "dry" signal until the tails get below some very small level. That might seem like it would be one pass through the delay time, but almost all delays feed back some of the delayed signal to get a longer and softer fade out. This is controlled by something like 'repeats'. So the circuit doesn't know how long the tails will last.

You could just let the input to the delay line stay muted all the time when the effect is "bypassed", but then the residual noise of the delay line stays in the output too. Without delays to mask it, this can be significant. You could do something like a noise gate to sense when the output of the delay line has fallen below X volume, then cut off the delay line output, but you'd also have to switch this out when the delay is switched on.

Deciding how to do tails neatly is much more complicated than simply bypassing the delay line back to dry. Not impossible, but it's far more than cutting a few wires and modifying some component values in all the delays on the market that I know of.  I know of one commercial pedal that came close, but the circuitry to add what you want would have added about 25% more to the entire circuitry of the pedal.
R.G.

In response to the questions in the forum - PCB Layout for Musical Effects is available from The Book Patch. Search "PCB Layout" and it ought to appear.

Mark Hammer

This is where I like the Anderton CMOS switches from the EPFM books.  They can be used in latching OR non-latching modes, such that you can engage the effect for as long as you hold your foot down, and it reverts back when you lift your foot off.

Since the actuator for the circuitry is simply a SPST switch, I parallel a latching stompswitch and a momentary switch.  If I leave the latching one open, I can use the momentary.  And if I don't feel like holding my foot down, I use the latching switch.

ashcat_lt

Quote from: joySeeing on November 27, 2015, 07:08:49 PM
...I guess I would still have a problem with tails being pulled over into the next delay activation...
I don't think this is the issue that you think it is.  Given that you haven't changed the Repeats (Feedback, whatever your pedal calls it) knob since you bypassed it, these previous tails will just continue to fade out at the same rate even after you engage the effect.  I can't see how doing it any other way would sound anywhere near as natural and intuitive.

Now, if that delay does trail off when you hit bypass, it actually might not be (theoretically) so hard to do what you want.  The switch is already a momentary switch which is activating some active flip-flop thing that makes it act like latching.  You should be able to bypass that flip-flop part that stores the state of the switch, but of course details matter.  Schematics would help.

slacker

If you don't fancy or can't modify your DD5, here's a couple of loopier ideas that will give you tails from any delay pedal, they can use either momentary or latching switches or both so you have the option of the effect being on only when the switch is pressed. They don't solve the possibly problem of previous delays still recirculating next time you engage it.

http://www.diystompboxes.com/smfforum/index.php?topic=90340.0
http://www.diystompboxes.com/smfforum/index.php?topic=90460.msg769328#msg769328




joySeeing

I've decided to move this over to a software based pedal for now to achieve the type of control I'm looking for. And I'm more familiar with this than circuit boards (for now).

I'm still learning about how delay can work in general, so let me know if I'm missing something.

I spent last night ripping apart an old effects pedal to extract the 1/4 inch jack and mod it to my arduino so that in Pure Data, I can simply get a on/off 1/0 value if my piano pedal is depressed. And this all works, whew.

As far as the mechanics of the delay... I'm going to try to implement a delay that activates when you press the pedal (of course) and when you release the pedal, it uses that same chunk size to "fade out" or get out of the way. So the deactivation does more than just "stops the creation of new delay lines" but also fades out all previous lines. This way, even if it is immediately reactivated, it has that specific chunk size amount of milliseconds (or whatever) to allow 1. fade the olds, 2. new to be record and there is no overlap.

PD / Max/msp / Supercollider (haven't settled on one yet) will handle only the wet signal and will not be in charge of the dry because of the latency that will occur with dry. Thankfully, delay is one technology where latency hardly matters (because it is being delayed anyways). Also this complicates things less for me. The wet will be mixed back in with stereo.

ashcat_lt

So you're going to create a new buffer with each activation?  How much memory are we working with here?

Anyway, I still think you're overstating the "problem", looking for a solution that isn't really necessary, and that your proposed solution is going to be wonky and sound unnatural.

joySeeing

Ok, after days of messing this and trying to learn Max / MSP and some cabling.... I made THIS: https://www.youtube.com/watch?v=bauob6T6rYU

Quackzed

thats pretty good thinking, from your you tube page i gathered you used 4 different 'memory' buffers and cycled through them? so you can go A...a....a...B...b...b...b....b...C...c....c...D...d.....d.....d.....A..a... etc? each press clears the previous memory and uses the current cleared memory. very clever!!! and more musically usefull as you mentioned...  now someone figure out how to do this with analog delay chips already...  :icon_redface: no, not me.
thats pretty slick thinking.  8)
nothing says forever like a solid block of liquid nails!!!

samhay

Very nice.

> I'm going to try to implement a delay that activates when you press the pedal (of course) and when you release the pedal, it uses that same chunk size to "fade out" or get out of the way.

Are you using ring buffers or is there a maximum length you can record for (if so, what happens if you exceed this?).

If you can record for a long time, it might be nice to be able to set the maximum 'fade out' time.

I'm a refugee of the great dropbox purge of '17.
Project details (schematics, layouts, etc) are slowly being added here: http://samdump.wordpress.com

joySeeing

Quote from: Quackzed on December 03, 2015, 12:15:22 AM
thats pretty good thinking, from your you tube page i gathered you used 4 different 'memory' buffers and cycled through them? so you can go A...a....a...B...b...b...b....b...C...c....c...D...d.....d.....d.....A..a... etc? each press clears the previous memory and uses the current cleared memory. very clever!!! and more musically usefull as you mentioned...  now someone figure out how to do this with analog delay chips already...  :icon_redface: no, not me.
thats pretty slick thinking.  8)

Something like that. Once a pedal is released, it fade's out that delay unit over the course of time you specify and then the buffer is immediately cleared for that delay unit and it is prepped to begin again. MEANWHILE, if you push the pedal again during that time, it uses the next delay unit. Maybe there should be more than 4 just in case. There's probably a simple non-verbose programatic solution that creates infinite new ones and deletes the other from memory when they are done but I'm only bright enough to suspect their existence.


Quote from: samhay on December 03, 2015, 04:21:54 AM
Very nice.

> I'm going to try to implement a delay that activates when you press the pedal (of course) and when you release the pedal, it uses that same chunk size to "fade out" or get out of the way.

Are you using ring buffers or is there a maximum length you can record for (if so, what happens if you exceed this?).

If you can record for a long time, it might be nice to be able to set the maximum 'fade out' time.

I'm not sure what you're asking. It's all done in Max. So you can control everything with tapin~ and tapout~ objects, basically.