FM Synth with DC Offset?

maybe this is not the best idea, but i have been wondering if this is something i could accomplish with renoise: creating an FM Synth using the ‘DC Offset’ DSP in conjunction with a sample consisting of a single 0-crossing silence, looped.

maybe “FM Synth” is too heavy a term, so let me just focus on getting a sine wave.

here’s what i’ve tried to do thus far, and haven’t gotten it to work how i was hoping, please let me know if i’m just certifiably insane:

  • ‘Create’ a new sample, but leave it as is, so it’s just silence.

  • add the following DSP Chain: Key Tracker -> Formula -> DC Offset.

  • Key Tracker set to a range of C-0 to B-9, scaled linear, destined to Formula’s ‘B’ input.

  • i left the Formula default function pretty much as-is for now, as i’m trying to just Get Sound to happen, but i threw in a little ‘* 440’ in the sin(SAMPLECOUNTER/SRATE…) line. again, i know this isn’t all that is required to get Correct Notes to happen, but i’m just trying to get audible sound first.

  • Formula DSP’s output is destined to DC Offset’s “DC Offset”.

then i play notes.

i can clearly see the VU indicator moving in a sinusoidal manner, and the rest of the system’s visual outputs show me i’m emitting something, and for the most part, the rate at which the VU jobby is zinging back and forth does appear to increase as i key in higher pitched notes, but overall i can never manage to get anything beyond a very low/quiet buzz.

no matter how much i jack up the numbers inside the formula() in order to try and get it be jiggering the DC offset faster so that i can get a higher pitch, i never end up hearing anything like a useful tone.

there is also a very “visually obvious” interpolation or harmonic or something in that some notes end up moving the DC slider in a very slow fashion, or are completely still, in despite of their neighbour notes being very rapid.

i’m beginning to speculate that the DC Offset DSP can simply not be used in the manner i’m trying to use it, in despite of how theoretically sound my reasoning seems to be that it ought to be possible to use it to essentially create a waveform. i’ve tried different sample rates for my audio hardware (old macbook), but it doesn’t yield any substantial difference in the outcome.

in specific i’ve wanted to avoid VSTs or Tools or other external things that just Don’t Come with the default install of Renoise, but also have been trying to create a ‘non-sample’-based way of emitting FM-like sounds from the platform.

thoughts?

(using 3.1 32b btw)

In short, It’s not that you can’t make this work to some degree with some crazy settings (this includes a very high BPM/LPB). But you really should not bother.

If you want fm you want fm, maybe you should explore the possibilities of the sampler more, work with samples with few harmonics etc.

P.S. you’ll get AM with the dc offset.

lol it is AM, thanks for the correction!

So you don’t want to use synths but want a synth ?

Not sure what the fascination is with that ?

For very basic oscillator shapes you can use the am filter in modulation (plus a dc sample loop) or the ringmod device or the am filter in the chorus. I find sampled wavecycles in a loop are much more versatile & can sound better. The formula device is for modulation of parameters at tick rate (slow), not usable for audio generation.

The fascination is you don’t need VSTs for simple substrative synthesis, renoise is very performant, and you can share the modules with anyone. And you place a limitation on your tools, some people seem to like this for optimizing their creative mindset and throughput.

Well if you want AM, by sacrificing stereo, you can get RM by multiplying the left sample frame with the right sample frame, then sum the output with the left or right original sample and you will get AM, can be done with a scripting plugin like protoplug. This way the plugin only performs one operation while renoise performs everything else.

For very basic oscillator shapes you can use the am filter in modulation (plus a dc sample loop) or the ringmod device or the am filter in the chorus. I find sampled wavecycles in a loop are much more versatile & can sound better. The formula device is for modulation of parameters at tick rate (slow), not usable for audio generation.

The fascination is you don’t need VSTs for simple substrative synthesis, renoise is very performant, and you can share the modules with anyone. And you place a limitation on your tools, some people seem to like this for optimizing their creative mindset and throughput.

Point being, you don’t need to do any of that, a single cycle can be taken from any recording, can be drawn, and either of those will create much more complex starting points than the suggestions here, but hey its a tracker and all that that entails, i guess the hard way to do everything is the flavour of the week haha