FM

FM
« on: August 07, 2018, 08:02:37 AM »
I don't own the Prophet 12 yet. My first real synths was the DX27, then the DX7IID. I found this thread on the 12: http://dsiforum.com/viewtopic.php?t=7067 Beside that I can't find much examples of LFM sound demos on the net for the 12. Maybe some of you could share demos similar or close to what I hear in the video? Nothing fancy. LFM possibilities or limits on the Prophet 12 compared to 2,4 and 6 op.

https://youtu.be/fxSTjiE_5V0

Then going further with FM on the Prophet 12 would be cool too.
 :)

Re: FM
« Reply #1 on: August 07, 2018, 01:54:50 PM »
I posted this link somewhere else here. I had 30 years of DX7 programming, so when I got Cassandra (my P12) I expanded on the Yamaha FM world, so my voices are not like the classic ones shown on your DX5 video, but for what it's worth - https://soundcloud.com/andy-15-3/fm-examples

Shaw

  • ***
  • 1185
Re: FM
« Reply #2 on: August 07, 2018, 03:50:32 PM »
...so when I got Cassandra (my P12)...
You name you’re synths? All of them?  Are they all female?
"Classical musicians go to the conservatories, rock´n roll musicians go to the garages." --- Frank Zappa
| Linnstrument | Suhr Custom Modern | Mayones Jaba Custom | Godin Multiac Nylon | Roland TD-50 | Synergy Guitar Amps | Eventide Effects Galore |

Re: FM
« Reply #3 on: August 07, 2018, 06:20:16 PM »
I posted this link somewhere else here. I had 30 years of DX7 programming, so when I got Cassandra (my P12) I expanded on the Yamaha FM world, so my voices are not like the classic ones shown on your DX5 video, but for what it's worth - https://soundcloud.com/andy-15-3/fm-examples
Thank you, I'll check it out. How would you compare "Cassandra" with the DX7? Was there a hm.. sound barrier as far as your creativity with LFM, places where you wish you could go further?

Re: FM
« Reply #4 on: August 08, 2018, 01:48:21 AM »
My DX7 (1984 - 2014) was 'Marsha', I don't remember why (something to do with a Zappa Album). Somehow my Bass became 'Tree' (probably the polished wooden body) and Guitar 'Peg' (body shaped like a clothespeg). Cassandra was an ancient Greek Prophetess. I've not had any other proper synths.
 I am an 'outsider artist' - all my music follows non traditional rules, I started with redefining all possible chords and keys (for equally tempered instruments) then Timbre for which LFM was a godsend and most recently 8712 rhythm families and over 40 million polyrhythms.
 I have written over 240 quadraphonic voices for Cassandra, originally the P12 had (unknown to me) the 'wrong' kind of FM but I went with it anyway and found as many useful sounds as I could, so when the LFM update arrived I was comfortable with all the elements that could be integrated to expand my FM knowledge. It is just different to the DX7, instead of being 'restricted' (I never was and made 256 great DX7 voices) by the limitations of fixed algorithms and sine waves, I get to mess with the shapes of the waves before FM occurs, and assemble any of the possible permutations of 4 oscillators including loops in voice A, and then double it up (or not) with something in part B.
 I guess I am a pragmatist, I don't wish for what equipment will do for me, I wish to push what it can do beyond the obvious to achieve a satisfying result for my music. Whilst I have a very limited budget (a £2000 synth works out at £2 per week over 20 years) and little music gear compared to many, I can make original music, find the timbres and flavours of sounds I create to match the unusual musical structures I employ, or just throw away the rules and freely improvise - Cassandra is really Great for Improvisation, all those performance controls (and I use 4 pedals, Stereo output A and B plus 2 expression pedals) dedicated to bringing out the subtleties (or not so subtle parts) of each voice. Nothing else I have used has been so good for free Improv - I have been 'Headhunted' for a duet with a major Improv player from 100 miles away (a long distance in the UK) because of my use of timbre.
 I pretty much stick with Stack mode for outputs A and B, the two volume pedals mean I get to blend the parts of the voices and If I feel the need to go further there are all the lovely knobs on top - I often edit both parts A and B simultaneously (hold the button down til it flashes) to alter the attack or decay etc - maybe in another 5 years I will be good enough to tweak the right part of the voice (A or B) but with so many voices I am not adept yet and crude is best.

Re: FM
« Reply #5 on: August 08, 2018, 07:17:36 AM »
My DX7 (1984 - 2014) was 'Marsha', I don't remember why (something to do with a Zappa Album). Somehow my Bass became 'Tree' (probably the polished wooden body) and Guitar 'Peg' (body shaped like a clothespeg). Cassandra was an ancient Greek Prophetess. I've not had any other proper synths.
 I am an 'outsider artist' - all my music follows non traditional rules, I started with redefining all possible chords and keys (for equally tempered instruments) then Timbre for which LFM was a godsend and most recently 8712 rhythm families and over 40 million polyrhythms.
 I have written over 240 quadraphonic voices for Cassandra, originally the P12 had (unknown to me) the 'wrong' kind of FM but I went with it anyway and found as many useful sounds as I could, so when the LFM update arrived I was comfortable with all the elements that could be integrated to expand my FM knowledge. It is just different to the DX7, instead of being 'restricted' (I never was and made 256 great DX7 voices) by the limitations of fixed algorithms and sine waves, I get to mess with the shapes of the waves before FM occurs, and assemble any of the possible permutations of 4 oscillators including loops in voice A, and then double it up (or not) with something in part B.
 I guess I am a pragmatist, I don't wish for what equipment will do for me, I wish to push what it can do beyond the obvious to achieve a satisfying result for my music. Whilst I have a very limited budget (a £2000 synth works out at £2 per week over 20 years) and little music gear compared to many, I can make original music, find the timbres and flavours of sounds I create to match the unusual musical structures I employ, or just throw away the rules and freely improvise - Cassandra is really Great for Improvisation, all those performance controls (and I use 4 pedals, Stereo output A and B plus 2 expression pedals) dedicated to bringing out the subtleties (or not so subtle parts) of each voice. Nothing else I have used has been so good for free Improv - I have been 'Headhunted' for a duet with a major Improv player from 100 miles away (a long distance in the UK) because of my use of timbre.
 I pretty much stick with Stack mode for outputs A and B, the two volume pedals mean I get to blend the parts of the voices and If I feel the need to go further there are all the lovely knobs on top - I often edit both parts A and B simultaneously (hold the button down til it flashes) to alter the attack or decay etc - maybe in another 5 years I will be good enough to tweak the right part of the voice (A or B) but with so many voices I am not adept yet and crude is best.
Great feedback (pun intended). Thank you :)

Re: FM
« Reply #6 on: November 19, 2018, 06:15:14 PM »
Can anyone describe or point to a video regarding how the sub oscillators function when running the main oscs in linear FM?  I'm assuming you can't make them a 5th "operator"?  Since they track Osc1 (I believe?), does it just add extra beef to whatever Osc1 is doing, does it get modulated by the other Oscs, or is it taken out of the equation somehow?

Oddly enough, the first thing I'm going to do when I get my P12 is play with it's linear FM...    ;D

Re: FM
« Reply #7 on: November 19, 2018, 08:38:45 PM »
The linear FM on the P-12 is not really a DX style FM.  I was hoping that could mimic the DX FM but it doesn't really happen. 

I found doing some testing that the P-12 FM does not have the high frequency response like the DX.  The maximum frequencies generated in the P-12 FM is about 4khz.  The DX generates north of 10khz.  So really crisp  bell sounds are not achievable on the FM of the P-12.

I am really happy with the P-12.  So, please don't take this as a slap at the P-12, but the P-12 FM will not replace a DX.
Jim Thorburn .  Toys-  Dave Smith: Prophet 5, Rev 4; Prophet 08; Pro 2; Prophet 12 module; EastWest Orchestral soft synths; Yamaha S-90; Yamaha Montage 8, Yamaha DX-7; KARP Odyssey; Ensoniq ESQ-1.  All run through a Cubase DAW with a Tascam DM-24 board.

Razmo

  • ***
  • 2168
  • I am shadow...
    • Kaleidoscopic Artworks
Re: FM
« Reply #8 on: November 20, 2018, 08:31:45 AM »
There is no real point in getting a P12 for FM, if the FM you want is DX FM sound... the way they are programed are quite different even if the technique is the same... if one want real Yamaha FM sound, get a Yamaha FM synth instead, it's a much more immediate result for what you want.

The P12 FM (both linear and nonlinear) i see more like yet another sound sculpting feature... it's there to give you FM tonality, not to be a DX replica... when linear FM was introduced the intend was the same... to give you FM tonal characteristics, but this time just in a more melodically friendly fashion.

The FM of the P12 should be used as a tool to get more tonal variety into the rest of the synthesis engine... at least that is my opinion. That is where the unique magic lies with the P12 FM features... the capability to combine FM synthesis with analog filters and VCAs, and further mangle it up in the audio rate modulation engine.
« Last Edit: November 20, 2018, 08:33:24 AM by Razmo »
If you need me, follow the shadows...

Re: FM
« Reply #9 on: November 20, 2018, 12:47:07 PM »
The linear FM on the P-12 is not really a DX style FM.  I was hoping that could mimic the DX FM but it doesn't really happen. 

I found doing some testing that the P-12 FM does not have the high frequency response like the DX.  The maximum frequencies generated in the P-12 FM is about 4khz.  The DX generates north of 10khz.  So really crisp  bell sounds are not achievable on the FM of the P-12.

I am really happy with the P-12.  So, please don't take this as a slap at the P-12, but the P-12 FM will not replace a DX.

Well I already have a DX7iid, so I definitely wasn't looking for a replacement.  :D

I just am very curious to see/hear how the FM will work couched within the rest of the P12 architecture.  And actually I really like 4op FM anyway; it's not as rich or complex, but same with monos, sometimes you just want a simple sound that 4op does just as good as 6 would anyway.

LoboLives

Re: FM
« Reply #10 on: November 20, 2018, 01:02:10 PM »
Actually the new update for the Roland System 8 has FM and it sounds so phenomenal that I’ve actually considered getting the System 8 specifically for that.

Re: FM
« Reply #11 on: November 20, 2018, 02:05:12 PM »
Actually the new update for the Roland System 8 has FM and it sounds so phenomenal that I’ve actually considered getting the System 8 specifically for that.

Yes I know, I just got one of those a few weeks ago.   ;D ;D 8)   I currently have an SH-101 loaded in it tho, because I've always wanted one of those and just couldn't justify the prices they have been going for the last few years.

But the System 8, despite being "fake analog", sounds amazing, and utterly different than the P12.  Between the 2 of them, especially if you're using Roland Cloud for all the plug-outs, offer a ridiculous amount of sonic variety and quality.  It will be fun to compare their different approaches to 4op linear FM, given that neither of them were designed to be FM synths.

Re: FM
« Reply #12 on: November 20, 2018, 08:05:01 PM »
The linear FM on the P-12 is not really a DX style FM.  I was hoping that could mimic the DX FM but it doesn't really happen. 

I found doing some testing that the P-12 FM does not have the high frequency response like the DX.  The maximum frequencies generated in the P-12 FM is about 4khz.  The DX generates north of 10khz.  So really crisp  bell sounds are not achievable on the FM of the P-12.

I am really happy with the P-12.  So, please don't take this as a slap at the P-12, but the P-12 FM will not replace a DX.

Well I already have a DX7iid, so I definitely wasn't looking for a replacement.  :D

I just am very curious to see/hear how the FM will work couched within the rest of the P12 architecture.  And actually I really like 4op FM anyway; it's not as rich or complex, but same with monos, sometimes you just want a simple sound that 4op does just as good as 6 would anyway.

The P-12 FM is a lot of fun to use in conjunction with the rest of the instrument.  You will find a lot of interesting sonic qualitis.
Jim Thorburn .  Toys-  Dave Smith: Prophet 5, Rev 4; Prophet 08; Pro 2; Prophet 12 module; EastWest Orchestral soft synths; Yamaha S-90; Yamaha Montage 8, Yamaha DX-7; KARP Odyssey; Ensoniq ESQ-1.  All run through a Cubase DAW with a Tascam DM-24 board.

LoboLives

Re: FM
« Reply #13 on: November 20, 2018, 09:50:22 PM »

Re: FM
« Reply #14 on: November 21, 2018, 03:13:58 AM »
There is no real point in getting a P12 for FM, if the FM you want is DX FM sound... the way they are programed are quite different even if the technique is the same... if one want real Yamaha FM sound, get a Yamaha FM synth instead, it's a much more immediate result for what you want.

The P12 FM (both linear and nonlinear) i see more like yet another sound sculpting feature... it's there to give you FM tonality, not to be a DX replica... when linear FM was introduced the intend was the same... to give you FM tonal characteristics, but this time just in a more melodically friendly fashion.

The FM of the P12 should be used as a tool to get more tonal variety into the rest of the synthesis engine... at least that is my opinion. That is where the unique magic lies with the P12 FM features... the capability to combine FM synthesis with analog filters and VCAs, and further mangle it up in the audio rate modulation engine.

Agreed. You can get into DX-type territory with a Prophet 12/Pro 2, though, if you stick to sine waves. The main difference lies in the kind of frequency modulation utilized, though. Yamaha used phase modulation, not linear FM in the traditional sense. Both can produce similar outcomes if sine waves are used as modulators and carriers, but they will start to differ a lot, once you use other waveforms, which include all the classic ones like triangle, square, and sawtooth. The reason is that while phase modulation can be considered a variation of linear FM, no frequencies are being modulated, but only the phase values of the carrier waveform.

One more note: You write, "when linear FM was introduced the intend was the same... to give you FM tonal characteristics, but this time just in a more melodically friendly fashion." Linear FM was available from the get-go on all modular systems of the 1960s. It's the way you implement linear FM and how it affects the sidebands that will determine how pleasant or clangy it sounds. There is not just one version of linear FM, but at least three. And there are versions of linear FM that sound way more dissonant than any type of exponential FM.
« Last Edit: November 21, 2018, 03:20:41 AM by Paul Dither »

Re: FM
« Reply #15 on: November 21, 2018, 07:20:59 AM »
Linear FM was available from the get-go on all modular systems of the 1960s. It's the way you implement linear FM and how it affects the sidebands that will determine how pleasant or clangy it sounds. There is not just one version of linear FM, but at least three. And there are versions of linear FM that sound way more dissonant than any type of exponential FM.

This is something I've been wondering about since hearing the first System-8 FM examples, how the sidebands seem quite different than what I hear on P12 FM examples.  That said I'm still not really sure how the System-8 FM works... you dial up "FM oscillators" with variations (like on all the Sys8 oscillators), but I'm not sure you're really FMing or even PDing them.  It's almost like you are using two "FM-sounding" oscillators and cross-modding between them or something.  Which I guess could be a kind of FM, but with relatively limited control compared to actual FM or PD.

Could you elaborate on or point a link to what you would consider the (at least) 3 types of FM?

Re: FM
« Reply #16 on: November 21, 2018, 08:17:22 AM »
Could you elaborate on or point a link to what you would consider the (at least) 3 types of FM?

Sure. Three types of linear FM: positive linear FM, positive and negative linear FM, and logarithmic or exponential positive and negative FM.

Now first about the positive and negative thing. That is related to the question of whether upper and lower sidebands are being produced (positive and negative) or only the upper one (positive) – these are the harmonics produced by FM that shape the timbre of a note. Why's that a thing? Usually, an oscillator signal is bipolar, i.e. it oscillates between a positive and a negative value during one duty cycle. In FM, the applied depth of that oscillation is determined by the modulation depth value. Let's say you choose a modulation depth that causes an amplitude of +/-200 Hz. Now you apply that to a frequency of 440 Hz. The resulting upper and lower frequencies are going to be 640 and 240 Hz. Now you increase the modulation depth to an amplitude of +/- 600 Hz while still applying it to 440 Hz. Mathematically, the result would be upper and lower frequencies at 1040 and -160 Hz. There's only one problem: There's no such thing as negative frequencies in the real world. They can only drop to 0, not below. So what happens to the lower sideband once 0 Hz are hit? The applied modulation depth will be inverted once it hits the bottom. In this case, the -160 Hz turn into 160 Hz. That's usually what happens in the case of +/- linear FM. One can even hear the outcome if one turns the modulator's frequency down into LFO range and chooses an extreme mod depth. The inverted lower sideband will become audible as an "upwards bump" that goes against the grain of the modulation wobble you'd expect.

+ linear FM only results in upper sidebands being produced, while the lower sidebands remain the same. In practice, that produces sounds that are not better behaved, but rather resemble the outcome of ring modulation. It's one of the most unruly types of FM.

A typical byproduct of standard linear FM is that the modulation is more pronounced in the lower registers and less so in the upper octave range. This is because the logarithmic nature of scales is not taken into account. One can compensate for that, though, if one adds a logarithmic function that ensures that the modulation intensity is multiplied by the factor of the relation between two semitones with each increasing note. That resulted in the most sophisticated version of +/- linear FM, the one that can be called logarithmic or exponential +/- linear FM. However, there's still an obstacle that makes it less perfect than Yamaha's PM. The relation between two semitones is 1 : the twelfth root of two, which is an irrational number going on forever. Rounded, it's about 1.06. And in digital technology you have to cut off that number at some point, otherwise the processors wouldn't stop calculating. That ultimately results in rounding errors, which become more or less pronounced depending on the notes you play and of course the FM depth.
« Last Edit: November 21, 2018, 08:25:21 AM by Paul Dither »

Razmo

  • ***
  • 2168
  • I am shadow...
    • Kaleidoscopic Artworks
Re: FM
« Reply #17 on: November 21, 2018, 08:51:23 AM »
There is no real point in getting a P12 for FM, if the FM you want is DX FM sound... the way they are programed are quite different even if the technique is the same... if one want real Yamaha FM sound, get a Yamaha FM synth instead, it's a much more immediate result for what you want.

The P12 FM (both linear and nonlinear) i see more like yet another sound sculpting feature... it's there to give you FM tonality, not to be a DX replica... when linear FM was introduced the intend was the same... to give you FM tonal characteristics, but this time just in a more melodically friendly fashion.

The FM of the P12 should be used as a tool to get more tonal variety into the rest of the synthesis engine... at least that is my opinion. That is where the unique magic lies with the P12 FM features... the capability to combine FM synthesis with analog filters and VCAs, and further mangle it up in the audio rate modulation engine.

Agreed. You can get into DX-type territory with a Prophet 12/Pro 2, though, if you stick to sine waves. The main difference lies in the kind of frequency modulation utilized, though. Yamaha used phase modulation, not linear FM in the traditional sense. Both can produce similar outcomes if sine waves are used as modulators and carriers, but they will start to differ a lot, once you use other waveforms, which include all the classic ones like triangle, square, and sawtooth. The reason is that while phase modulation can be considered a variation of linear FM, no frequencies are being modulated, but only the phase values of the carrier waveform.

One more note: You write, "when linear FM was introduced the intend was the same... to give you FM tonal characteristics, but this time just in a more melodically friendly fashion." Linear FM was available from the get-go on all modular systems of the 1960s. It's the way you implement linear FM and how it affects the sidebands that will determine how pleasant or clangy it sounds. There is not just one version of linear FM, but at least three. And there are versions of linear FM that sound way more dissonant than any type of exponential FM.

I did not explain myself clearly I can see :) with FM introduction I meant when DSI decided to implement linear FM on the P12... before that, the FM capabilities was not of much use in melodic presets, more like for sound FX etc... linear FM is what makes the pitch stable... but you know all that :)
If you need me, follow the shadows...

Re: FM
« Reply #18 on: November 21, 2018, 09:05:04 AM »
I did not explain myself clearly I can see :) with FM introduction I meant when DSI decided to implement linear FM on the P12... before that, the FM capabilities was not of much use in melodic presets, more like for sound FX etc... linear FM is what makes the pitch stable... but you know all that :)

Ah sorry, that was clearly me misreading your comment.

Re: FM
« Reply #19 on: November 21, 2018, 01:09:20 PM »
Could you elaborate on or point a link to what you would consider the (at least) 3 types of FM?

Sure. Three types of linear FM: positive linear FM, positive and negative linear FM, and logarithmic or exponential positive and negative FM.

Now first about the positive and negative thing. That is related to the question of whether upper and lower sidebands are being produced (positive and negative) or only the upper one (positive) – these are the harmonics produced by FM that shape the timbre of a note. Why's that a thing? Usually, an oscillator signal is bipolar, i.e. it oscillates between a positive and a negative value during one duty cycle. In FM, the applied depth of that oscillation is determined by the modulation depth value. Let's say you choose a modulation depth that causes an amplitude of +/-200 Hz. Now you apply that to a frequency of 440 Hz. The resulting upper and lower frequencies are going to be 640 and 240 Hz. Now you increase the modulation depth to an amplitude of +/- 600 Hz while still applying it to 440 Hz. Mathematically, the result would be upper and lower frequencies at 1040 and -160 Hz. There's only one problem: There's no such thing as negative frequencies in the real world. They can only drop to 0, not below. So what happens to the lower sideband once 0 Hz are hit? The applied modulation depth will be inverted once it hits the bottom. In this case, the -160 Hz turn into 160 Hz. That's usually what happens in the case of +/- linear FM. One can even hear the outcome if one turns the modulator's frequency down into LFO range and chooses an extreme mod depth. The inverted lower sideband will become audible as an "upwards bump" that goes against the grain of the modulation wobble you'd expect.

+ linear FM only results in upper sidebands being produced, while the lower sidebands remain the same. In practice, that produces sounds that are not better behaved, but rather resemble the outcome of ring modulation. It's one of the most unruly types of FM.

A typical byproduct of standard linear FM is that the modulation is more pronounced in the lower registers and less so in the upper octave range. This is because the logarithmic nature of scales is not taken into account. One can compensate for that, though, if one adds a logarithmic function that ensures that the modulation intensity is multiplied by the factor of the relation between two semitones with each increasing note. That resulted in the most sophisticated version of +/- linear FM, the one that can be called logarithmic or exponential +/- linear FM. However, there's still an obstacle that makes it less perfect than Yamaha's PM. The relation between two semitones is 1 : the twelfth root of two, which is an irrational number going on forever. Rounded, it's about 1.06. And in digital technology you have to cut off that number at some point, otherwise the processors wouldn't stop calculating. That ultimately results in rounding errors, which become more or less pronounced depending on the notes you play and of course the FM depth.

Awesome, that is very helpful!  Do you know what if any synths provide good examples of each?  I'm assuming the P12 would not be doing + only, and +/- also seems problematic if it's creating "upward bumps", but then if that means it's doing log/exp +/-, Dave would have needed to somehow introduce the log function for mod intensity?