DSP Robotics Support

R&R wrote: My main (fundamental) Saw in my synth uses your syncable LFO ramp&saw oscillators equipped with the transition time input. The one inside the LFO that Spogg kindly posted here
[...]Would it be possible to make a version of that oscillator that has a fixed (bound) output to +1 -1 regardless of transition time? Similar to the WaveShape oscillator that Kevin uses?

Sure, no problem, here you go. The normlization costs some extra CPU though. If you don't modulate the frequency or transition very fast then you could hop some of the math (compute the normalisation factor only every 16 or so samples).

P.s. In the LFO link above the saw and ramp oscs are labeled wrong, it is the other way round.

P.p.s. I am surpried these things get used by others at all, given the crptic (read: nonexistent) documentation inside.

martinvicanek wrote:Sure, no problem, here you go.

I am very greatful!

martinvicanek wrote:P.p.s. I am surpried these things get used by others at all, given the crptic (read: nonexistent) documentation inside.

LOL
Your stuff is what makes FS fun...
Wouldn't even have bothered creating anything in FS if it weren't for your work and Spoggs... or even Kevins creations as reference for that matter. It's amazing how you all are able to create these things.

I still have zero clue about any theory (math&sound). FS is a nice pasttime mysterious puzzle

So then...
Adding this to the update todo list for my synth out with the old Saw of my LFO and in with this new one...

k brown wrote:[...]Spread Partials osc that could be phase modulated

Hmm, interesting. Phase modulation requires a phase to begin with. For periodic signals, the phase goes from 0 to 1 (in FS units) as we traverse one cycle. Easy. But what if the signal is not perodic, like when we choose a non-integer spread factor? Something I need to think about.

R&R wrote:Bruce.gif

martinvicanek wrote:

k brown wrote:[...]Spread Partials osc that could be phase modulated

Hmm, interesting. Phase modulation requires a phase to begin with. For periodic signals, the phase goes from 0 to 1 (in FS units) as we traverse one cycle. Easy. But what if the signal is not perodic, like when we choose a non-integer spread factor? Something I need to think about.

Feel like kind of a dope for not having considred that. At low settings the SP osc wave is quite periodic, though.

I realize that phase modulating the Spread Partials osc when many partials are present would probably not sound very pretty, but with fewer partials (soft timbres), could be useful.

No dopes here!
I think that for modulation purposes, we can interpret phase as time divided by the period of the lowest partial (and the integer part removed). For integer spread factors, this coincides with the phase of the (periodic) waveform , otherwise it is just a scaled time.
The fsm below implements this concept. Naturally, it uses more CPU than the phaseless version. I left it in modular rather than asm monolithic form so the structure can be grasped easier. (Truth is I was too lazy to optimize it. )
Have fun!

OMG !!

Thank you, thank you, t h a n k y o u!

Now fitting it into NeuOld FM - gonna call it EmVee FM. - absolutely loving it! Adding a third assignable EG.

Done.

Posted it in the NeuOld FM thread.

Héhé not only the osc !)
(Even if i have tendency to redo everything myself, i always need to see MV shematic to help and use other's.)
Very interesting to see innovative thing like this !!
I do not test it so much yet as i'm more in fx than synth but impatient to see some usage,
and i will probably try to understand more the math and concept behind !)