FM via Layer B's output into CV in - Urban legend or real?

[lnetzel]

The idea has been around for years in the "other" forum but I've never gotten it to work and the users writing in that thread are not present anymore so I want to blow life into this question again or get to the bottom of this, since this method is said to provide the best keytracking (and is seems logical).

The idea:
1. You have to be in stackmode so you can trigger both Layer A and B with the same key.
2. Patch Layer B's left output into CV-IN and Layer B's Right output into "space" (leave it hanging to get rid of Layer B being mixed through the normal output)
3. On Layer A, use a Mod slot with CV In as Source (Breath or Expression, don't remember which right now) and OSC 1 FREQ as Destination
4. On Layer B, mess around with the VCA volume and different frequencies since the CV-IN is "heavily filtetred" until you get the desired modulation.

My trial:
I know the Mod slot with CV-In works because I've plugged external gear into it. Korg MS10's LFO (called MG, Modulation Generator) output is a 5 volt output and I've had no problem getting that to control the OSC1 FREQ on the Prophet '08.
I've used all sorts of settings and wave forms on Layer B but never gotten any effect on Layer A with this method.

I asked the DSI support and got this reply Aug 14 2015

Thanks for contacting us here at DSI. Reading through the thread (rem: he is referring to the other forum I linked to in my question), I tested out a Prophet 08 here in the office and can confirm that I could NOT replicate the FM process (layering method from the Output B into the CV port).
...
If you do find an alternate way of doing this, please let me know and if you have any other questions, feel free to ask.

So this obviously seems like an urban legend to me now...

Does anyone have any proof this works, like a patch to download and deeper knowledge on what you have to do on Layer B to do this. Or, can this method be killed off as a hoax?

[Pym]

Technically yes but it's severely limited in terms of how fast the modulation can happen, eg not high audio rates

[lnetzel]

Technically yes but it's severely limited in terms of how fast the modulation can happen, eg not high audio rates

Can you elaborate on this? I off course don't expect the P'08 to be able to sound like a DX7 (or even remotely) but what settings on the Layer B (and what audio range) would actually work to make CV-in have an affect on the OSC FREQ on Layer A?

I'm not an electrician and I understand that an -5/+5 V LFO is not the same as an audio signal, but it's still electricity coming out of Layer B outputs (even if it's meant for a "speaker"), right, or am I completely wrong here?

[cr73645]

I remember reading an extensive post about it, but never tried it myself.

I really enjoy FM sounds, but I've learned that FM is quite complicated with analog oscillators. This is where the P12 is probably way better than the P08. It has the availability of Fm in the first place... with the P08 I accept that I'll just have filter FM.

Anyway, it's an interesting experiment... maybe I'll give it a try.

Cheers!

[lnetzel]

I remember reading an extensive post about it, but never tried it myself.

I really enjoy FM sounds, but I've learned that FM is quite complicated with analog oscillators. This is where the P12 is probably way better than the P08. It has the availability of Fm in the first place... with the P08 I accept that I'll just have filter FM.

Anyway, it's an interesting experiment... maybe I'll give it a try.

Cheers!

Yes, there is an extensive post in the other forum but it's dead and the few sysex examples are dead links and noone has provided any audiable proof of this and believe me, I'm trying to get it to work as described.

[lnetzel]

Actually, I have gotten some effects now, not very FM-like, more like the unstable Boards Of Canada:ish sound, which is not bad, not bad at all... but I will explore this more the next few days and upload a patch later.

[lnetzel]

OK, so finally some audio and documentation of my explorations.

I've attached the sysex for my patch "Layer B -> CV in", Screen dumps from the Sound Tower application for both Layer A and B (in case you don't feel like loading it up into your P'08), a screenshot of the Midi I've recorded in Cubase and 2 mp3s. One with the Layer A only (no modulation) and then a Stacked version where Layer B is modulating Layer As OSC FREQ.

I've used all 4 Mod slots with the same setting on Layer A
Source: Midi Exp
Dest: OscAllFreq
Amount: 127

For each mod slot used, the modulation is stronger.

So to make this patch work:
Plug Output B Left output -> Pedal/CV
Plug Output B Right output -> Leave it hanging (to kill the Output B right from being mixed into the Main output as well)

Since the Master Volume also affects the Output B, you need to keep a pretty loud setting. With a low Master Volume, volume will be to low to cause modulation.

It sounds horrible but proves it's possible to do Frequency Modulation the stacked layer way. If anyone have tips to improve this and actually make this into a useful patch that you can play with I'd be very grateful. especially on the OSC FREQ settings on Layer B... cause I guess there's where the FM ratio hidden is right?

[MartinM]

Moinmoin,

this method will not lead to real FM, as the CV is low pass filtered. In theory the manual wording ("heavily filtered") may be subject to interpretation, but real world shows a cutoff far below the LFO's maximum frequency, which is 261Hz (~middle C).
So "FM-sounds" via CV-input will be even "less FM" than setting LFO3 destination to VCO frequency.

For the P'08 I see no way to supply or even feed back general audio signals as control voltages without circuit bending. Audio Modulation remains the only way to get close to FM without using a soldering iron.

Martin

[lnetzel]

...this method will not lead to real FM, as the CV is low pass filtered...

Aha, that was good info to have! Thank you!