Prophet 12 FM examples (DX7 style sounds)

[joosep]

Could 6-operator FM not be had by using a stacked patch, one having a 4-op and the other having a 2-op?

Not unless you can modulate one layer with another (assuming that's what you mean by "stacked").  6op means you have 5ops modulating 1op in some configuration (algorithm).  A 4op patch simply played simultaneously with a 2op patch might yield interesting results, but unless you can put the 2ops "in line" (thus "linear FM") with the other 4ops, it won't give you 6op FM.

Linear FM does not mean "in line" like you refer.
And ofcourse you can have a 6 operator FM patch with the P12. It just depends on what algorithm you want to achieve.
Here are all the DX7 algorithms:


As one can see, almost all algorithms can be achieved on the P12.

Algorithm 1: 2 layers. Layer 1 is oscillator 2 modulating oscillator 1. Layer 2 is 4 oscillators, each modulating the next and the last one also modulating itself.
...
Algorithm 28: 2 layers. Layer 1 is oscillator 2 modulating oscillator 1. Oscillator 6 without modulation. Layer 2 is 3 oscillators modulating the next and the last one modulating itself.
...

I think out of all of them, only 16,17 and 18 cannot be done on the P12.

[Paul Dither]

6op means you have 5ops modulating 1op in some configuration (algorithm).  A 4op patch simply played simultaneously with a 2op patch might yield interesting results, but unless you can put the 2ops "in line" (thus "linear FM") with the other 4ops, it won't give you 6op FM.

The term "linear FM" does not indicate in what order a couple of operators modulate each other. The attribute "linear" refers to the relationship between hertz and volts in an oscillator circuit (Hz/V). No matter how much input voltage you apply, the frequency will always vary in a linearly proportional manner. What is doubled or halved in volts, will be doubled or halved in frequency: 1V = 55Hz, 2V = 110Hz, 4V = 220Hz, 8V = 440Hz, 16V = 880Hz. So in short: linear FM means that you modulate in hertz. There are still plenty of different linear FM versions, though.

This linear relation between hertz and volts stands in contrast to the logarithmic or exponential relation between frequency and pitch, as per octave the frequency is doubled. As an equivalent to this, you have the standard V/Oct with the following relations: 1V = 55Hz, 2V = 110Hz, 3V = 220Hz, 4V = 440 Hz, 5V = 880Hz. The V/Oct standard is what exponential FM is based upon. LFOs also typically work this way.

Both is different from the kind of frequency modulation Yamaha utilized: phase modulation. In phase modulation not the frequency of a carrier is linearly modulated, but rather its playback speed, or – to be more precise – the phase value of the carrier waveform. If you rise the modulation signal, the carrier will be accelerated, if you lower the modulation signal, the carrier will be slowed down. This was done to avoid phase shifting, which would lead to dissonances and not desired tunings, particularly in complex FM situations where you use more than a pair of oscillators.

[Quatschmacher]

Could 6-operator FM not be had by using a stacked patch, one having a 4-op and the other having a 2-op?

Not unless you can modulate one layer with another (assuming that's what you mean by "stacked").  6op means you have 5ops modulating 1op in some configuration (algorithm).  A 4op patch simply played simultaneously with a 2op patch might yield interesting results, but unless you can put the 2ops "in line" (thus "linear FM") with the other 4ops, it won't give you 6op FM.

Good point. I was taking the above example of the Rhodes sound where it seemed to be 3 pairs of operators running parallel to each other so at least that ought to be possible.

[Sulqué]

It would nice to see a good video of the P12's FM possibilities one day. Just one. I did a test on Youtube, Prophet 12 FM, and Novation Peak FM. I encourage you to do the same. I've read a lot about how cool, or not so cool, great sounding, or not great sounding FM on the P12 is, but virtually no examples. Where can I hear patches and what algorithm was used and how?

[proteus-ix]

...

I think out of all of them, only 16,17 and 18 cannot be done on the P12.

Interesting points, in fact I was thinking of algorithms like 16-18, where there is one output and essentially all of the other oscillators follow a line of modulation relating back to the base oscillator, but I see now how that is only a limit for that type of "tree" algorithm.  Thanks for explaining this, it means the P12 is even more powerful for FM than I thought!

Sorry System-8, you'll just have to do my Lately Bass and leave the fancy stuff to my P12. 

[SandyS1]

I'd think you couldn't really get any of the ones where operators feed back to themselves. In fact, given that you only get FM from 4 -> 3 -> 2 -> 1 [-> 4], many of those algorithms wouldn't be reproducible.

[proteus-ix]

I'd think you couldn't really get any of the ones where operators feed back to themselves. In fact, given that you only get FM from 4 -> 3 -> 2 -> 1 [-> 4], many of those algorithms wouldn't be reproducible.

Couldn't you setup Osc 1 source -> Osc 1 FM destination?  Or would that just be FM mod but not really a feedback loop to where Osc 1FM output doesn't really then become a source for Osc 1 again?

[SandyS1]

I'd think you couldn't really get any of the ones where operators feed back to themselves. In fact, given that you only get FM from 4 -> 3 -> 2 -> 1 [-> 4], many of those algorithms wouldn't be reproducible.

Couldn't you setup Osc 1 source -> Osc 1 FM destination?  Or would that just be FM mod but not really a feedback loop to where Osc 1FM output doesn't really then become a source for Osc 1 again?

I don't know. I was thinking of the interface through the Oscillator section, not the mod matrix. It's worth a shot. If that works, then more of those algorithms would open up.

[Quatschmacher]

I'd think you couldn't really get any of the ones where operators feed back to themselves. In fact, given that you only get FM from 4 -> 3 -> 2 -> 1 [-> 4], many of those algorithms wouldn't be reproducible.

Couldn't you setup Osc 1 source -> Osc 1 FM destination?  Or would that just be FM mod but not really a feedback loop to where Osc 1FM output doesn't really then become a source for Osc 1 again?

Or by using the audio out as a mod source and feeding it back into the oscillator?

[joosep]

I'd think you couldn't really get any of the ones where operators feed back to themselves. In fact, given that you only get FM from 4 -> 3 -> 2 -> 1 [-> 4], many of those algorithms wouldn't be reproducible.

Incorrect. All oscillators are sources and also destinations. You can easily use the modulation matrix to create FM. The 4->3->2->1 is just a pre-defined shortcut, it works 100% the same as the modulation matrix.

From the P12 manual:

[jdt9517]

The P-12 FM is different than the DX FM.  I did some experimenting with modulating sine waves, and it does not modulate like the DX.  Using sine waves, sidebands over 4k hz were almost non-existent.  The DX will produce sidebands well in excess of 12k hz using the same carrier/modulator ratios.   Haven't dug deep into modulating the more complex waves yet.  However, I am hopeful that I can get the higher timbres going that direction.

Weird I wonder why this is the case.  Do they just not run fast enough or modulate tight enough?

Have you been able to get higher timbres with it's FM yet?

Not through the FM.  Other ways yes, but with the FM - no.  I have found the FM does its best in bringing ring type modulation to the synth or bringing a dirtiness to the sound that can emulate analog.   But the oscs will not follow Chowning Theory fully in programming.  Possibly not enough processing power to push all the sidebands out.

While it was 35 years ago, I remember that processing power was a big issue to make the DX fully perform.  The computer in the DX was monsterous for the day.  I anticipate whatever the replacement for the P-12 will be,  it will have a much further developed and robust FM section to it - but look how big the Montage had to be to get all the processing power in to it.

However, it's a very capable synth and is one of my "go-to" instruments.  I've learned to work within those capabilities.  I know the limits there and bring out the musical colors in other ways.  It's definitely a keeper.

[SandyS1]

Incorrect. All oscillators are sources and also destinations. You can easily use the modulation matrix to create FM. The 4->3->2->1 is just a pre-defined shortcut, it works 100% the same as the modulation matrix.

Ah, thanks. Good to know.

[Chaparral]

I did list all of the possible P12 FM combinations per layer elsewhere in the P12 section of the forum under 'Algorithms (operator combinations)'

[jdt9517]

Both is different from the kind of frequency modulation Yamaha utilized: phase modulation. In phase modulation not the frequency of a carrier is linearly modulated, but rather its playback speed, or – to be more precise – the phase value of the carrier waveform. If you rise the modulation signal, the carrier will be accelerated, if you lower the modulation signal, the carrier will be slowed down. This was done to avoid phase shifting, which would lead to dissonances and not desired tunings, particularly in complex FM situations where you use more than a pair of oscillators.

Do you think the P-12 has linear but not DX phase modulation?  I have A/B'd both the DX and the P-12 using sine waves on the P-12 and an oscilloscope to look at the waves, and the P-12 does not come close to resembling the DX using equal ratios.  I thought it was lack of processing power on the P-12, but maybe the form of FM is actually different from the DX.  Thoughts?

[Paul Dither]

Both is different from the kind of frequency modulation Yamaha utilized: phase modulation. In phase modulation not the frequency of a carrier is linearly modulated, but rather its playback speed, or – to be more precise – the phase value of the carrier waveform. If you rise the modulation signal, the carrier will be accelerated, if you lower the modulation signal, the carrier will be slowed down. This was done to avoid phase shifting, which would lead to dissonances and not desired tunings, particularly in complex FM situations where you use more than a pair of oscillators.

Do you think the P-12 has linear but not DX phase modulation?  I have A/B'd both the DX and the P-12 using sine waves on the P-12 and an oscilloscope to look at the waves, and the P-12 does not come close to resembling the DX using equal ratios.  I thought it was lack of processing power on the P-12, but maybe the form of FM is actually different from the DX.  Thoughts?

Yes, the Prophet 12 and Pro 2 use linear FM but no phase modulation like any Yamaha DX synth, the SY77/99, or the Montage. One can only approximate the outcome of phase modulation with linear FM as long as one uses sine waves. With linear FM, all other waveforms except sine waves will lead to different results than with phase modulation. With phase modulation a triangle wave, for example, will result in a square-like modulation while a square wave modulator will hack the carrier waveform (the result of phase jumps = abruptly changing phases of a wave).

[jdt9517]

@Paul-  Thanks for the info.  I wasn't aware of all the flavors of FM in your posts.  You provided a great education!

[Razmo]

The cool thing about the P12 is that it is exactly NOT like the DX7 in my opinion... it may sound different, but to me, the FM tones is much more about being a source of "oscillator" that you can manipulate in the filter... a filter that the DX never had... you can create complex FM tones that can constantly change in timbre, and then throw that thru the filters and amps... it's a very different way of utilizing FM synthesis in contrast to a DX... if I wanted that DX way of doing it, and that sound, I'd simply get a DX instead... they are not that expensive.

What makes the P12 so special is that it has audio rate modulation going on... in previous threads it was also mentioned that this modulation is bandlimited at about 11Khz, which may be part of the reason you cannot push those high sidebands out of it (The DX had a custom FM chip in it that worked in the megahertz of operation to achieve that kind of complexity back then... no processor back then would be able to handle this task, and Yamaha made several itterations of their FM chips after the DX7 appeared)... but when you think about it, FM is not at all the limit... you can do A LOT with audio rate modulation, but i simply think that not many users dare to venture into sculpting using this semi-modular smorgasboard of possibilities... you could mix and route FM with AM... take one oscillator, FM it with another, and then take this and AM it with a third oscillator... i mean... the options are pretty much endless, but probably just as hard to predict as with creating an FM patch on a DX.... this is certainly why there is so few demo's of the P12 FM out there... FM is hard to predict, and the way you set up the algorithms on a P12 is vastly different to how you do it on a DX... probably even harder, so it's not hard to understand why there are so few demos out there.

In fact... with todays synthesizers that have the depth of a DSI engine or equivalent, few ever reach down deep enough to give these synths the praise they deserve.... they have a huge amount of potentially untapped power in them that very few knows how to dig up, or even want to dig up because of the amount of time it takes to fully utilize them... the P12 is by all means, such a synth, and I'd say that no one really knows what it's capable off... not even Dave himself... in theory maybe, but the potential still remains to be heard I'd say

[Paul Dither]

The cool thing about the P12 is that it is exactly NOT like the DX7 in my opinion... it may sound different, but to me, the FM tones is much more about being a source of "oscillator" that you can manipulate in the filter... a filter that the DX never had... you can create complex FM tones that can constantly change in timbre, and then throw that thru the filters and amps... it's a very different way of utilizing FM synthesis in contrast to a DX... if I wanted that DX way of doing it, and that sound, I'd simply get a DX instead... they are not that expensive.

What makes the P12 so special is that it has audio rate modulation going on... in previous threads it was also mentioned that this modulation is bandlimited at about 11Khz, which may be part of the reason you cannot push those high sidebands out of it (The DX had a custom FM chip in it that worked in the megahertz of operation to achieve that kind of complexity back then... no processor back then would be able to handle this task, and Yamaha made several itterations of their FM chips after the DX7 appeared)... but when you think about it, FM is not at all the limit... you can do A LOT with audio rate modulation, but i simply think that not many users dare to venture into sculpting using this semi-modular smorgasboard of possibilities... you could mix and route FM with AM... take one oscillator, FM it with another, and then take this and AM it with a third oscillator... i mean... the options are pretty much endless, but probably just as hard to predict as with creating an FM patch on a DX.... this is certainly why there is so few demo's of the P12 FM out there... FM is hard to predict, and the way you set up the algorithms on a P12 is vastly different to how you do it on a DX... probably even harder, so it's not hard to understand why there are so few demos out there.

In fact... with todays synthesizers that have the depth of a DSI engine or equivalent, few ever reach down deep enough to give these synths the praise they deserve.... they have a huge amount of potentially untapped power in them that very few knows how to dig up, or even want to dig up because of the amount of time it takes to fully utilize them... the P12 is by all means, such a synth, and I'd say that no one really knows what it's capable off... not even Dave himself... in theory maybe, but the potential still remains to be heard I'd say

Agreed. I also just wanted to mark the differences between DX-style FM and the kind of FM utilized in the P12/P2, that's all. There was no judging involved.

Like you said there are lots of modulation options available on the P12/P2 and the linear FM feature adds a nice way to modify the instant timbre of a sound in basic but also rather complex ways before you even make use of anything beyond the oscillator section. It's worth exploring what can be done from within this module alone, i.e. with only oscillator-specific parameters.

A general note on FM, be it linear or PM: Once it gets complex, i.e. beyond 2 oscillators, this is not really a type of synthesis made for spontaneous tweaking. FM synthesis usually requires a rather analytical approach that includes the definition of a goal and figuring out what techniques are necessary to get there. That's pretty much the only way to make the use of FM predictable: the understanding of sound and why certain sounds sound the way they do. Of course one can just tweak some FM parameters as one moves along, but that typically only leads to rather arbitrary and largely forgettable sounds if all your tools consist of FM parameters only.

[Chaparral]

I too have appreciated the additional insights into Linear FM posted here. I agree with Razmos comment above - although I got a P12 to 'replace' 2 broken DX7s, I didn't ever think I would be cloning my DX7 voices, only using 30 years of FM experience in new ways. If I had another 30 years of music making ahead of me I might have explored morphing more deeply instead, but as it is I get to use the P12s expression capabilities to slide in and out of FM sounds and to change the FM in ways that were too complicated to achieve quickly with a DX7. If I had any money for a new synth I would buy another P12 so I could gig without the paranoia of being left keyboardless.