DSP Robotics Support

I've noticed a lot of your schematics use those 'junctions' which as I recall during my FS demo period were just a 'straight wire' inside a module. What's the purpose of those?

Also - why are these being done with stream monos instead of polys?

Since I'm the dope without FS, let's do this the other way 'round. Here's a schematic of my situation. With a 0.5 float - perfect square, with a PW control, a saw on acid. Maybe it's not the same saw you guys are using.

Spogg wrote:I appear to have spogged one of Martin's schematics!

You are very welcome!
Funny thoung, I was tempted to apply the very same tweak this morning but then decided not to. The thing is you nail the min/max values so it looks neat and all but in doing so you are adding DC!

k brown wrote:Thanks for that, but I can't open the schematic I only have SM.

Ouch!

In general, is there an advantage to doing PWM with a steam subtract rather than mixing the phase-shifted saw through an inverter?

It's the same thing.

About the images I posted - there must something I'm not understanding. When I try to post larger, higher-res images I get a warning that they must 600 pixels on the largest dimension.

Actually 650. When reducing the image use better interpolation methods than nearest neighbor.

I've noticed a lot of your schematics use those 'junctions' which as I recall during my FS demo period were just a 'straight wire' inside a module. What's the purpose of those?

Mostly to tidy up the wiring. You can also use those for integer conversion or trigger extraction bu choosing a particular output pin type.

Also - why are these being done with stream monos instead of polys?

Just a demo - works also n poly.

The problem in your schematic was a faulty phase implementation Fix attached.

So!- the code I had was different; I thought I'd lost my marbles. Thanks so much for fixing that.

'Cause I'm not shy about begging - did you ever do one of these as a Triangle?

The DC problem you mentioned - is that also true of the 'fancier' PWM shown the schematic at the top of my post? That one also maintains a centered waveform at all pulse widths.

So, I tried your saw with the 'fancy' pwm (stays-centered) method and it didn't work! I noticed your oscillators have a different phase relationship to the stock oscillators. Here's how I modified the PWM to work with your oscs.



Does this have DC on the output?

You can easily measure DC by placing a lowpass with sufficiently low cutoff.

martinvicanek wrote:The thing is you nail the min/max values so it looks neat and all but in doing so you are adding DC!

Yes indeed! It's the problem of assymetrical waveforms having a dc component.
I guess it depends on what application for the PWM signal you have in mind.
Also, if you have a scope display on your synth, users might not like to see that apparent "offset" shifting as you change the pulse width. So as you say, mine looks better

Cheers

Spogg

I put a couple of u-he synths on a scope and they don't keep the pulse waves centered with varying width, so probably safe to say it's a non-issue in the real world. On some, the scope in the plugin itself shows the waveform staying centered, so it must be receiving a 'corrected' wave so it 'looks nice' on the scope. I do use the centered, unipolar version in LFOs though, so the pitch stays 'tuned'. It's interesting that there seems to be no consensus regarding unipolar vs bipolar on LFO waveforms in the synth world hardware or software. Dave Smith's LFO waves are all uni except Sine and Tri, none of u-he's (of the ones I checked) are uni, the ARP 2600 hardware - same as Smith's. Same with 'voltage' ranges of LFO waves; Smith's are +/- 5v on the bi-polars, 0-5v on the uni's while others' are 0-10 on unis, ARP Odyssey's are all over the place, etc.

Anywho, thanks again Martin for these and the correction - they sound great and do keep the CPU load at bay.

Great work, and I've given you credit on my website.

Wow so many cool sounds and good emotions. Thanks for this post