KVR Audio

cron · Post by **cron** » Fri Sep 10, 2004 8:02 pm

Someone posted something about analog synths being capable of "instantaneous feedback" a little while back, and I'm not entirely sure what he/she meant. Feedback as in feedback loops, or feedback as in the tactile relationship with the instrument (e.g., a steering wheel giving you feedback about a road's surface).

The reason I ask is because I've been experimenting wih feedback loops in Audiomulch lately and, if the original poster meant the first type of feedback, what exactly am I missing through the lack of "instantaneous feedback"?

I'm pretty sure that the poster was referring to the zero latency tweakability of analog synths, but thought I'd double check

.

Edit: Had to change the topic to something a little more abrupt as the end was chopped off.

shamann · Post by **shamann** » Fri Sep 10, 2004 8:08 pm

cron wrote:Someone posted something about analog synths being capable of "instantaneous feedback" a little while back, and I'm not entirely sure what he/she meant. Feedback as in feedback loops, or feedback as in the tactile relationship with the instrument (e.g., a steering wheel giving you feedback about a road's surface).

The reason I ask is because I've been experimenting wih feedback loops in Audiomulch lately and, if the original poster meant the first type of feedback, what exactly am I missing through the lack of "instantaneous feedback"?

I'm pretty sure that the poster was referring to the zero latency tweakability of analog synths, but thought I'd double check .

Edit: Had to change the topic to something a little more abrupt as the end was chopped off.

Besides the notion of any kind of feedback immediately implies latency. It takes at least a modicum of time for something to go in before it can be fed back.

nuffink · Post by **nuffink** » Fri Sep 10, 2004 8:11 pm

Feedback is never instantaneous. There is always a phase shift and at some frequency this will become significant.

Robert Randolph · Post by **Robert Randolph** » Fri Sep 10, 2004 8:17 pm

Instantaneous feedback is called amplification.

and even then what nuffink said applies.

I have no idea what dr.wackler was talking about (he was the one who said it correct?)

cron · Post by **cron** » Fri Sep 10, 2004 8:19 pm

Thanks shamann and nuffink.

Any insights as to how this limitation can affect your sound? I'd have thought if there was a regular latency in the loop you'd encounter a sort of comb filter/tight delay effect at a constant pitch.

I shouldn't really worry as the stuff sounds OK as long as there's a limiter in there somewhere, otherwise the sound totally overshoots the headroom available and you're left with a near subsonic rumbling at best (which I'm assuming is something only encountered in the digital realm). It was just bugging me really.

Analog can't do that cool subsonic sound anyway.

cron · Post by **cron** » Fri Sep 10, 2004 8:22 pm

Robert Randolph wrote:Instantaneous feedback is called amplification.

Ah, so it's as water is to DHMO then.

Indeed it was dr.wackler who originally said it, just had a look through his posts.

valley · Post by **valley** » Fri Sep 10, 2004 8:35 pm

until Dr Wackler defines exactly what it is in the analogue world that exhibits "instantaneous feedback", I'd say the question is largely unanswerable.

I guess at a quantum level it is possible for effect to happen before cause, but I fail to grasp what relevance at all this has to anlogue audio circuitry.

meister eder · Post by **meister eder** » Fri Sep 10, 2004 9:37 pm

cron wrote:Indeed it was dr.wackler who originally said it, just had a look through his posts.

Hi there! Yes, it was me. So if you had a look through my post, just read on - I tried to explain what I mean in the very same thread. I can't do better than that, sorry.

cron · Post by **cron** » Fri Sep 10, 2004 9:55 pm

Thanks very much for replying.

The only other post from you in that thread is a pair of square brackets. Unless that's supposed to be a loop

, which Audiomulch is more than capable of (i.e., output goes into input goes though output goes into input etc ad infinitum).

Either way, it sounds nice, so who cares

.

meister eder · Post by **meister eder** » Sat Sep 11, 2004 12:46 am

http://www.kvr-vst.com/forum/viewtopic. ... 571#693571

http://www.kvr-vst.com/forum/viewtopic. ... 642#693642

http://www.kvr-vst.com/forum/viewtopic. ... 679#693679

http://www.kvr-vst.com/forum/viewtopic. ... 706#693706

http://www.kvr-vst.com/forum/viewtopic. ... 724#693724

shamann · Post by **shamann** » Sat Sep 11, 2004 1:25 am

Good Doctor,

Those explanations make no sense. A feedback loop starts when a signal passes through an input to an output to the input again. The initial signal is not in itself part of the feedback loop. Likely it's all faster than 1 sample (which on most systems is 1/44100 seconds, heaven forbid), but the delay is still there. If an original signal didn't get fed back into a loop, we would not call the phenomenon feedback.

cron · Post by **cron** » Sat Sep 11, 2004 1:29 am

Sorry, I only looked after the bit where you briefly mentioned it on page 21 or so.

I understand exactly what you mean by this in relation to guitars now, i.e., the string sympathetically vibrates the moment the air pressure hits it. The sort of feedback experiments I'm doing involve 'closed' systems though, e.g., flanger 1 into flanger 2 into flanger 1, so I'm not sure if this is still relevant to that area as there's no actual sound creating the feedback (though I guess the 1 sample delay still occurs between each device). I suppose I'll have to try these techniques in an analog circuit to appreciate the difference in sound this causes when experimenting with feedback (I'll no doubt learn all about this, and hopefully get the chance to do it when I begin my degree later this month

).

Thanks for your patience with my post blindness.

Edit: This seems to be a real point of contention

. Still, I should get to try all this myself soon.

meister eder · Post by **meister eder** » Sat Sep 11, 2004 1:35 am

shamann wrote:Good Doctor,

Those explanations make no sense. A feedback loop starts when a signal passes through an input to an output to the input again. The initial signal is not in itself part of the feedback loop. Likely it's all faster than 1 sample (which on most systems is 1/44100 seconds, heaven forbid), but the delay is still there. If an original signal didn't get fed back into a loop, we would not call the phenomenon feedback.

Thats right, a feedback loop starts when the original signal is sent into the loop, but the feedback itself happens the very moment it hits the startpoint of the loop again. Makes sense? I don't know how else to explain it.

nuffink · Post by **nuffink** » Sat Sep 11, 2004 1:39 am

dr.wackler wrote: Thats right, a feedback loop starts when the original signal is sent into the loop, but the feedback itself happens the very moment it hits the startpoint of the loop again. Makes sense? I don't know how else to explain it.

Where it's inevitably phase shifted to some degree

meister eder · Post by **meister eder** » Sat Sep 11, 2004 1:42 am

cron wrote:I understand exactly what you mean by this in relation to guitars now, i.e., the string sympathetically vibrates the moment the air pressure hits it..

Yeah, that pretty much describes it very well!

Now, what I say is not that this in detail is responsible for a certain difference in sound, but that this fact - on an abstract level - makes analog to what it is and what digital can not be. In that sense it is a (IMO very reasonable) answer to the question about what makes analog so analog - and thus also to the question about how we percieve analog sonically.

"Instantaneous feedback". Clarification?