I’ve skirted around the edges of “noise” for a while, I’d like to start using more system feedback in my recordings, but am a bit nervous about some of the risks. Mostly blowing out speakers, my ears, or burning out equipment (is this even possible?) I’m planning on using a passive matrix mixer with some delay pedals, OP-1 and a modular system (Phonogene, Echophon and Clouds are high on the list of feedback / sound sources).
One immediate question, I don’t have a hardware compressor, which would help me cut down on crazy feedback. If I run my feedback related signals through a ES-8, and use software compression, am I risking anything? Do I need to keep my feedback taming compression in its own box?
Posting this to studies as I’d love to see some ongoing noise-wrangling information.
The main thing to do to protect your speakers, keep the volume low and be especially careful with high-gain low (sub-50Hz) frequencies.
The main difference with using a hardware compressor and software will be where it sits in the signal chain. A hardware compressor or limiter before the inputs to your audio interface will help prevent clipping, but, as long as you’re careful and keep gain levels below the clipping point, it should be fine.
Never heard of anyone burning out equipment with audio feedback, I would think avoiding shorts would be the primary concern, so don’t mix output to output.
Here’s a PCB design for 2 channels of summing amplifier for converting bipolar 5V to unipolar 5V and back, for interfacing Eurorack with Teensy:
I originally got interested in scaling/attenuverters for the purpose of doing something similar, but I can see the utility for all sorts of purposes, if the module was flexible enough and supported enough channels. I’d love to have a module just like SSF’s gnd-ctrl that supported 8 channels.
Yeah, my read on the difference is that compressors are designed to take your signal and maximize its use of the spectrum, while limiters do so while also avoiding clipping. (that was a bit of a handwave and I suspect not entirely accurate).
So, I tend to think of compression as having a more audible effect of increasing overall loudness, while limiters do the same but to a more limited degree that has more to do with safety than alteration of the sound.
But I’m talking completely out of my derriere here. Someone else almost certainly knows more about this topic.
a compressor will take incoming audio and reduce it’s level by a specified ratio above the threshold point. for example, if you have a vocal track that has a wide dynamic range and would like to “even it out”, you could reduce the signal above X db by Y%. as the signal gets louder (assuming Y isn’t 100%) the outgoing signal will also get louder, just less so.
a limiter will reduce all incoming audio by 100% past the threshold, so nothing passes a fixed point.
so yes! a square is a rectangle, but a rectangle is not a square. a compressor can be a limiter, if the attack and release are set to zero and the ratio is set to infinite. a limiter cannot become a compressor though, as it has far less functionality.
in this case i’d think a limiter is what you’re looking for, set at a pretty high threshold so that it only engages when you’re about to feed back into the danger zone. maybe some kind of visual feedback would be helpful too, as you may not immediately notice if the limiter engages.
I love feedback and have been using it in a variety of contexts for a few years now.
The only safe way is to use feedback is to have something to completely shut your signal down that occurs after your feedback loop.
For example: let’s say you construct a feedback loop with a mixer, by plugging the headphone out back into an input channel. Eventually as you turn up volume at various points (the headphone out, the feedback input) you will get an oscillation. The gain level at this point is going to BLOW UP. If you have speakers after the mixer you will be pumping a ton of gain into them instantly (possibly damaging the speakers), headphones you will hurt your ears. So here it is ESSENTIAL to have another piece of gear after your feedback loop to bring the gain back down to a normal level. I have a distortion pedal I use and a reverb pedal, both have a ton of gain reduction that will bring things to a safe level.
Compressors won’t be adequate because it will be so loud that you would need to set the compressor to act as a limiter anyways. I have read that since feedback can get chaotic so quickly that it could potentially blow past a limiter, but I have not experienced this first hand. Again I have a couple pedals will enough gain reduction to make it safe.
Also I have not had any success with feedback in the digital realm, because as soon as you clip a digital signal it (to me) becomes unusable because it doesn’t have an interesting quality. It just sounds broken. So trying to create a feedback loop digitally results in too much unusable gain. Maybe if your audio interface had enough gain reduction you could make something happen.
Basically… using a hipass filter will do the trick (removing frequencies below say… 10 hz or something like that)
I use the SuperCollider audio programming language which has LeakDC.ar() which does the job also…
Recently I’ve been revisiting the use of microphone feedback in my work, with the goal of having (reasonable) control over both the pitch and volume / amplitude of the output. To be clear, there’s no digital components in the setup I’m discussing.
I’ve been having some success with pitch manipulation, but am struggling to be able to control volume such that the feedback itself is maintained, without also jumping to ear-splitting levels. I’ve done some reading on artists like Lucier, Tudor, Nick Collins and others, who mention the use of hardware limiters in their signal chain, presumably to cap the volume of the feedback at a given range.
To experiment, I purchased a relatively inexpensive Boss RCL-10 Micro Rack Compressor/Limiter, but when putting between the microphone and loudspeaker it doesn’t seem to be working as anticipated, i.e. taming excessive volume levels.
Does anyone have experience working with feedback in the way I’m describing, particularly in terms of being able to maintain a reasonable volume level? Understood that ‘reasonable’ here is certainly subjective, but looking for something that’s, say, about the same volume as if you were to be listening to music in your living room (again, subjective I know ).
Should microphone -> limiter -> amp / loudspeaker work for what I’m describing, or am I missing something about how those artists are utilizing hardware limiters in their setup?
Any guidance would be appreciated - thanks so much!
Thanks for the recommendations! Based on the description, it does sound like the tanh would work well in this context.
I am still curious, though, assuming again that the artists I’ve looked into (Lucier, etc) were using (or creating) equipment more similar to something like a standard limiter, why that standard piece of equipment doesn’t seem to have been effective in my case. I see mentions specifically of voltage controlled limiters, and wonder if that’s the difference between my little Boss limiter and what they were/are using…
Of course, if it’s that I need something more sophisticated then I’m all for it.
Just wanted to say no-input mixers aren’t a analog only thing — you can get really interesting soundscapes just from feeding sends into themselves on Ableton, and it’s not bound to sterile, «perfectly digital» sound if you know how to create instability and chaos. You also get very precise control and nearly infinite headroom, a stereo signal path, and a lot of flexibility. And you get as many effects as your computer affords you to run in real time, which is usually a lot.
EQs allow easily creating oscillating peaks. Frequency shifters with very small shift frequencies or granular delays with very small pitch shifts provide an interesting way to disturb stable oscillations created by resonant filtering, creating phasing patterns and slow timbral changes. Granular delays that only alter the signal by slightly jittering the delay amount by a few milliseconds can also give the sound a sense of organic movement. Saturation, erosion and vinyl distortion allow the addition of grit. A limiter at the end of each feedback path set to 0dB with some amount of gain is all you need to maintain oscillation indefinitely without blowing up.
In modular, I like to maintain infinite feedback loops with an envelope follower, inverted, controlling a VCA. There’s a careful balance of gain, VCA bias etc. but it should be pretty much the same as using a limiter.
I used to use a Bastl Dynamo plus a VCA, or Maths and a VCA, but now I use a custom ER-301 unit for that.
).