Synthesized feedback and simulated PWM (LOUD)

A trick I picked up on how to make feedback type sounds without actually using feedback. 

The Scale module mixes the voice with an Offset that is being modulated at audio rate. Lowering the Amount knob allows the modulated voice to break through more resulting in what sounds like feedback. With the Saturation maxed out the modulated voice ends up only being audible on the edges of the cycle which has an unusual sound. 

Also worth noting is the Pulse Width Modulation that's achieved by raising and lowering the Offset amount.


  • Cool idea! Love it!

  • edited July 1

    Ooh 😮 me likes!

    edit: I get how it works now, really cool idea! 👍

    now to try this with audio!

  • Impressive, bravo

  • Every now and then a thread like this will pop up and inspire some modular exploration and experimentation.

    Thank you @Johnisfaster

  • Its pretty neat right? Even if you’re not shooting for a feedback sound its really neat how the modulation gets shoved to the edges of the cycle.

    This trick was used on Ventolin :)

  • Man that PWM with the offset sounds so freakin good. The slight curve makes it warmer than a square.

  • Yours definitely sounds better than the screechy Aphex Twin siren 😉

  • I haven’t even tried (too much work, little stamina left), but I took a reference pic so far

    and here’s some unwarranted thoughts:

    DC offset is all over the place, it’s kind of the point I know, but what would result of filtering off zero frequencies at the end ? (HPF24 at minimum freq).., Does it still roar?

    Using WT instead of simple OSC

    Or FM operator (boom!!)

    Using audio like a flexi (gotta try that amen break)

    What about two (or more) scale+offsets

    And cross modulating these?

    Before I even try anything I need to slsp a limiter and DC remover at the end

  • Cutting off the offset with a highpass filter doesn’t seem to alter the sound in any noticable way.

    i wonder about DC offsets being in the mix sometimes. Is there a reason it shouldn’t be done? My thinking is if it sounds good it is good but I don’t know if theres a technical reason behind it.

  • 20 years ago I’d listen to Ventolin like it was just another song. Now I can only take about a minute. It sounds neat though, just kinda hurts.

  • I think the main reason is it blows your speakers, and can’t be good for eardrums either if you’re on phones

  • Amps usually have multiple DC cut filters on board so the speakers should be safe, same on iDevices (plus some DSP in place).

    Things might be different when using a headphones jack because iDevices have DC coupled outputs indeed and your headphones might suffer from DC offset, including distortion because of the nonlinear behaviour of the transducers.

  • That's what I don't get though. It's safe for an amp to go +/-1 at audio rate but it's somehow bad if it's lfo or fixed?

  • If a DC coupled power amp delivered a substantial DC offset to the loudspeaker, that would put constant high power on the bass speaker's voice coil that you don't hear. You would only see the speaker membrane constantly moved outwards or inwards. That's the risk: High power but you hear nothing. It could burn the voce coil if DC-induced power is high enough, and you might wonder what happened.

  • edited July 3

    I don’t mean to just be argumentative but Im kinda just a skeptic. Ive been hearing about the risk of sending offsets to your speakers for like 25 years. I used to have a Doepfer filter that actually added a DC offset on certain settings (a-106-6). When I was young I would mess around in Cool Edit and made beats that had offsets in them I didn’t even know what they were then.

    It kinda begs the question “how slow is too slow?” Like if I have a Square just buzzing away and I start slowing it down until its just clicks and then its just occasional clicks at what point did that Square start potentially damaging a speaker? Why can a speaker change very fast but not slow?

  • Is it just cause it can be deceptively loud? Like you wouldn’t even hear a Sine Lfo so maybe you’d crank the volume till your speaker blew?

  • edited July 3

    The answer is simple: If the DC causes higher RMS power in the speaker's voice coil than it's rated for then it will likely burn sooner or later. For that reason, power amps not only usually block DC but also have additional protection circuits that disconnect the speaker terminals from the amp's power stage in case one of the power transistors burns and causes a short circuit to the positive or negative power supply rail.

    DC power that equals the (moderate) RMS power of your music won't cause any damage of course. The huge difference is that you won't hear accidentaly applied DC power.

    I guess it would be easier to understand if you spent some time building amps, protection circuits, speakers and measuring chassis parameters 😉

  • So it's about the Max RMS. So if the Offset is in a musical context you can hear wouldn't that remove the risk of going over max RMS? Like below, if you were to put that on you would hear the high frequency which would keep you from cranking the volume way past your speakers max RMS wouldn't it?

  • edited July 3

    I don’t understand the technical part like @rs2000 explained, but think of it this way: the speaker membrane is oscillating at the rate of the audio, and is designed to do so around the zero (rest position). If you have a dc offset, you’re forcing it to operate at a point at the inside or outside of the middle, like being always out never crossing the middle (rest) position. That’s not natural..

  • Sorry new here. Can you clarify, are you saying building and using the set up in the OG post could cause damage to iPhone, iPad and any connected external speakers or have I read that wrong?

  • edited July 3

    Only some speakers and only at loud volumes. Add a High Pass filter with the cutoff set real low at the end will remove the DC offset.

  • edited July 3

    Dc offset is a standing wave at 0hz, as Im sure you know. You can see the “flat line “ even screenshoting

    You can’t hear it (like phase) but it can ruin a mix

    Btw had this saved, only difference is 24db

  • HP24 must be in beta

  • No, it’s not going to damage anything, but if you want a consistent mix at the end, these are things to keep in mind

  • edited July 3
  • So if the Offset is in a musical context you can hear wouldn't that remove the risk of going over max RMS? Like below, if you were to put that on you would hear the high frequency which would keep you from cranking the volume way past your speakers max RMS wouldn't it?

    Yes indeed, as long as the offset arriving at the loudspeaker is low enough in comparison.

    Still, it would add unwanted nonlinearities (although, again, that's unlikely to happen with common amp designs).

Sign In or Register to comment.