@kaffimusic -- this is explained in video #6, basically SHAPER is a sine wave lookup table, a sawtooth wave on input therefore "reads" back the table from beginning to end and immediately jumps back to the beginning, thus producing a sine wave output. SAW+ mixes in another signal, with the result being that the signal coming into SAW+ modulates the pitch of the sine wave.
Thanks about the 11:24 tip about hitting edit when FUN is highlited! That kind of little tidbits of information are always great bonus. It works on K2500 as well! I just had not realized that, not even after owning kurz three years or so. :D Saves great bunch of time and effort knowing all those shortcuts.
Thank you for posting this tutorial, it was really helpful. Now I know how to make FM sounds but I don't have parameters for some Yamaha dx11 sounds which I need. Can I find them somewhere on the internet or you could be helpful to send me? Thanks.
Thanks for the terrific tutorials. Is there a way in VAST to use a sawtooth wave to modulate a sine wave, such that the output is of a similar nature to kick drums, basically a fast and large ramping of the pitch down? I have found, using envelopes and LFO's applied to the pitch are limited by a delay of 20ms before they update, and especially odd and annoying, they induce a ramp up in pitch of around 10-15ms which is audible, even if you set an attack to 0 and only what to ramp down from a beginning pitch. After watching all your FM synthesis tutorials and then mucking about in my PC3 and watching the output in a scope, it does not behave as I would have expected, which may be a lack of my understanding of FM. However, in other VST instruments, I do find the FM to produce output as I would expect. What I expect is that the modulating source amplitude at any given moment combined with the gain should determine the current frequency of an output sine wave. However, with VAST shapers, the output being a lookup of a mapped amplitude, I don't see a way that it can evolve the output from a beginning frequency progressing lower or higher smoothly and predictably. Another way of describing what I expect, is for example, in one cycle of a sawtooth wave a number of cycles of a sine wave progressively ramps down in pitch, and that sawtooth cycle should be 20ms or less. I appreciate any take on solutions to the problem or explanations that might demonstrate what I have misunderstood. Thanks.
VAST pre-Forte does "FM" in a different way than many other synths, as near as I can tell it's essentially using a sine-wave lookup table (that's SHAPER). Anyways, on the Forte and related, there is "real" FM and this is possible there. I'm not entirely sure how to do it otherwise, but I will think on it.
@@poserp I figured it out finally. The key is that one needs to put a wrap before a shaper (not saw+shp) and use a waveshape (oscillator through dsp functions or a portion of an existing sample) that has an exponential curve in time(x) vs. amplitude(y), flip what you see on an oscilloscope 90 degrees counterclockwise. The PC3 Musician's Guide does not explain all the DSP functions, do you know a resource/manual that does contain them? The manuals for older Kurzweil synths include explanations of them, but the PC3 has some newer ones.
PC3k was and still is an incredible tool! I bought it recently because I was lucky it lay unpackaged in a music store for a long time, and I received the PC3K8 brand new and for relatively little money. Now I was also able to buy KORE64, German Grand D Extensions + RIP controller. Now I'm overjoyed. The only dilemma now will be to study the thick 460-page "black book". Although the most important thing for me would be to simply understand how to save the music being played to an USB flash drive or to a PC. And I still can't do it. And the other thing is why the stupid metronome is always ticking when recording (song+record), which bothers me. And how can I turn it off? 😢 Tanks!
I'll have to see if I have the files around someplace (I haven't used those in a few years), if I find them I'll look for a spot to post them that's more durable.
good evening to you. sorry, I don't understand English and I follow your tutorials. specifically I don't understand how the 1041 fm parts series program can play if all the volumes are at -96db? did my problem with English make me miss a few passages? thanks if you want to answer.
These are meant to be "cascaded" into another Layer that handles output volume. Generally speaking, the last Layer in a chain that applies amplitude envelopes and overall level control.
@@poserp of course. seeing your video I noticed that the last layer was still at -96 db and I can't understand how it could sound. another thing I would like to ask you, how come, at the beginning of the video the keymap is set to NONE and when you open the program 1041 is set to 174 SAWTOOTH WAVE? I would try to understand but this FM is really difficult, forgive me.
The next tutorial, #8 showing an FM bass sound, give an end-to-end example of how the programming works. These tutorials are more for immediate/advanced programmers, so I leave out some things that, if you're new to VAST and Kurzweil programming, might make things more confusing. This one in particular isn't about _how_ to program an FM sound, but rather how to create some Programs that can be useful when you're programming an FM sound on the PC3. These can then be re-used via "Import Layer" in new Programs that you may make. Sometimes, for that sort of "template" Program, it makes sense to have a Keymap (even though you don't ever hear it) for various reasons. Sometimes, it makes sense to have the Keymap set to "None". It really depends on what you're trying to accomplish and how you want to go about doing it.
I thought the harmonic series was multiplications of the fundamental frequency by an integer, that represent the harmonic in the series. So if the fundamental is 440hz, the second harmonic would be 880hz, the third 1320hz etc. hmm I need to dig deeper, it would seem to me a multiplication done in the "FUN" would do the trick.
One thing to bear in mind -- all multiplications will happen in cents, not hertz. This is why I have to do the things I did here. If I could control pitch directly via hz, it would be easier.
He means 32 Algorithms on the Yamaha DX7, not harmonics. There are 6 Operators/Oscillators on the Yamaha DX7. The number of harmonics is a function of how the modulation is set up, you aren't limited to 32.
As I explain below, it's related to "ratio". The PC3 has no pre-set FM "algorithms" so you can roll your own, and aside from feedback you can use these Programs to create any of the 32 algorithms in the DX7, plus many many more that go well beyond 6 operators. The 32 Layers in each Program are for ease of picking "nice" "ratio" values without having to futz around with figuring out the tuning of each harmonic in the harmonic series yourself, and to kinda-sorta allow you to program a Yamaha FM sound by matching the "ratio" values to a Layer number.
"the 32 harmonics usually available in various Yamaha FM-based synthesizers" Sorry, but what are you talking about? I'm afraid you don't seem to know, as you seem to have confused FM with additive synthesis. FM synths used small numbers of operators ("partials") to produce more harmonics, resulting from the set modulator:carrier pitch and amplitude combinations. Yamaha did this with only 4, 6, or 8 operators (which can be crude "partials" if each is an unmodulated sine to audio out). The DX7 and related synths had 32 ALGORITHMS - combinations of operatorss - but never 32 ops, which seems to be what you're trying here. Your task of replicating this would've been far, far easier if you had known what you were replicating...
Nope, we're talking "ratio", as in the frequency ratio between op a and op b. So, the idea here is that by having those pre-programmed in Layers we don't have to guess/look up what the frequency values are for the harmonic series (roughly speaking, my understanding is that the "Ratio" on a DX_ synth when not set to "fixed" will allow you to select the first 32 harmonics in the harmonic series). I can't remember off-hand where I picked that up, but it certainly could be wrong...
Sorry, duh, that's what I get for watching while half-asleep. So, you only create the "partial" layers as shortcuts for a desired ratio, for a given operator connection/'segment' type, and load them as needed - to speed up setting the frequencies of each op. That makes sense! Ignore me. Fwiw, IIRC most Yamaha FM synths could go up to something like 63 x ratio, but obviously it's of very low usefulness; anything beyond about 20 is sharply diminishing returns. Well, arguably such ratios can contribute to producing interesting noise sources, etc - albeit dependent on lucking into the right combination of frequency, amplitude, sampling rate of the given synth (as it tends to depend on fold-back), etc... it can take a long time to find the perfect hi-hat base. :D
Yeah, aliasing is what I was thinking of, since it's usually a case of picking a crazy ratio and finding the one key that produces exactly the right pitch that the aliasing fold-back has the desired effect. It's both an exact science, and complete trial-and-error!
@kaffimusic -- this is explained in video #6, basically SHAPER is a sine wave lookup table, a sawtooth wave on input therefore "reads" back the table from beginning to end and immediately jumps back to the beginning, thus producing a sine wave output. SAW+ mixes in another signal, with the result being that the signal coming into SAW+ modulates the pitch of the sine wave.
Thanks about the 11:24 tip about hitting edit when FUN is highlited! That kind of little tidbits of information are always great bonus. It works on K2500 as well!
I just had not realized that, not even after owning kurz three years or so. :D
Saves great bunch of time and effort knowing all those shortcuts.
Thank you for posting this tutorial, it was really helpful. Now I know how to make FM sounds but I don't have parameters for some Yamaha dx11 sounds which I need. Can I find them somewhere on the internet or you could be helpful to send me? Thanks.
14:21 . Hi, how did you create that algorithm 1036? I can't manage to create it....
Thanks for the terrific tutorials. Is there a way in VAST to use a sawtooth wave to modulate a sine wave, such that the output is of a similar nature to kick drums, basically a fast and large ramping of the pitch down? I have found, using envelopes and LFO's applied to the pitch are limited by a delay of 20ms before they update, and especially odd and annoying, they induce a ramp up in pitch of around 10-15ms which is audible, even if you set an attack to 0 and only what to ramp down from a beginning pitch. After watching all your FM synthesis tutorials and then mucking about in my PC3 and watching the output in a scope, it does not behave as I would have expected, which may be a lack of my understanding of FM. However, in other VST instruments, I do find the FM to produce output as I would expect. What I expect is that the modulating source amplitude at any given moment combined with the gain should determine the current frequency of an output sine wave. However, with VAST shapers, the output being a lookup of a mapped amplitude, I don't see a way that it can evolve the output from a beginning frequency progressing lower or higher smoothly and predictably. Another way of describing what I expect, is for example, in one cycle of a sawtooth wave a number of cycles of a sine wave progressively ramps down in pitch, and that sawtooth cycle should be 20ms or less. I appreciate any take on solutions to the problem or explanations that might demonstrate what I have misunderstood. Thanks.
VAST pre-Forte does "FM" in a different way than many other synths, as near as I can tell it's essentially using a sine-wave lookup table (that's SHAPER). Anyways, on the Forte and related, there is "real" FM and this is possible there. I'm not entirely sure how to do it otherwise, but I will think on it.
@@poserp I figured it out finally. The key is that one needs to put a wrap before a shaper (not saw+shp) and use a waveshape (oscillator through dsp functions or a portion of an existing sample) that has an exponential curve in time(x) vs. amplitude(y), flip what you see on an oscilloscope 90 degrees counterclockwise. The PC3 Musician's Guide does not explain all the DSP functions, do you know a resource/manual that does contain them? The manuals for older Kurzweil synths include explanations of them, but the PC3 has some newer ones.
PC3k was and still is an incredible tool! I bought it recently because I was lucky it lay unpackaged in a music store for a long time, and I received the PC3K8 brand new and for relatively little money. Now I was also able to buy KORE64, German Grand D Extensions + RIP controller. Now I'm overjoyed. The only dilemma now will be to study the thick 460-page "black book". Although the most important thing for me would be to simply understand how to save the music being played to an USB flash drive or to a PC. And I still can't do it. And the other thing is why the stupid metronome is always ticking when recording (song+record), which bothers me. And how can I turn it off? 😢 Tanks!
Good blass you for that work 🎉
I wonder if you could do this stuff with the PC3LE series keyboards?
Tuomas: Yes, it is possible with newer firmware. It is hidden under VAST item. :-)
Hi, looks like the link to your PC3 FM templates is not working anymore.
Is there any way to get them?
Thank you for all your amazing work!
RM
I'll have to see if I have the files around someplace (I haven't used those in a few years), if I find them I'll look for a spot to post them that's more durable.
@@poserp Thank you, that would be amazing!
Hi PoserP, any update on this? 🤞🏻
good evening to you.
sorry, I don't understand English and I follow your tutorials.
specifically I don't understand how the 1041 fm parts series program can play if all the volumes are at -96db?
did my problem with English make me miss a few passages?
thanks if you want to answer.
These are meant to be "cascaded" into another Layer that handles output volume. Generally speaking, the last Layer in a chain that applies amplitude envelopes and overall level control.
@@poserp
of course.
seeing your video I noticed that the last layer was still at -96 db and I can't understand how it could sound.
another thing I would like to ask you, how come, at the beginning of the video the keymap is set to NONE and when you open the program 1041 is set to 174 SAWTOOTH WAVE?
I would try to understand but this FM is really difficult, forgive me.
The next tutorial, #8 showing an FM bass sound, give an end-to-end example of how the programming works. These tutorials are more for immediate/advanced programmers, so I leave out some things that, if you're new to VAST and Kurzweil programming, might make things more confusing. This one in particular isn't about _how_ to program an FM sound, but rather how to create some Programs that can be useful when you're programming an FM sound on the PC3. These can then be re-used via "Import Layer" in new Programs that you may make. Sometimes, for that sort of "template" Program, it makes sense to have a Keymap (even though you don't ever hear it) for various reasons. Sometimes, it makes sense to have the Keymap set to "None". It really depends on what you're trying to accomplish and how you want to go about doing it.
@@poserp
thank you very much for your availability, I follow you for a long time and the little I can do on PC3K I owe it to you ... good day.
I thought the harmonic series was multiplications of the fundamental frequency by an integer, that represent the harmonic in the series. So if the fundamental is 440hz, the second harmonic would be 880hz, the third 1320hz etc. hmm I need to dig deeper, it would seem to me a multiplication done in the "FUN" would do the trick.
One thing to bear in mind -- all multiplications will happen in cents, not hertz. This is why I have to do the things I did here. If I could control pitch directly via hz, it would be easier.
poserp Ah. I get it. Thank You.
He means 32 Algorithms on the Yamaha DX7, not harmonics. There are 6 Operators/Oscillators on the Yamaha DX7. The number of harmonics is a function of how the modulation is set up, you aren't limited to 32.
As I explain below, it's related to "ratio". The PC3 has no pre-set FM "algorithms" so you can roll your own, and aside from feedback you can use these Programs to create any of the 32 algorithms in the DX7, plus many many more that go well beyond 6 operators. The 32 Layers in each Program are for ease of picking "nice" "ratio" values without having to futz around with figuring out the tuning of each harmonic in the harmonic series yourself, and to kinda-sorta allow you to program a Yamaha FM sound by matching the "ratio" values to a Layer number.
More difficult than CALCULUS!!! No wonder why the vast majority of FM sounds used on records were presets!!
"the 32 harmonics usually available in various Yamaha FM-based synthesizers" Sorry, but what are you talking about? I'm afraid you don't seem to know, as you seem to have confused FM with additive synthesis. FM synths used small numbers of operators ("partials") to produce more harmonics, resulting from the set modulator:carrier pitch and amplitude combinations. Yamaha did this with only 4, 6, or 8 operators (which can be crude "partials" if each is an unmodulated sine to audio out). The DX7 and related synths had 32 ALGORITHMS - combinations of operatorss - but never 32 ops, which seems to be what you're trying here. Your task of replicating this would've been far, far easier if you had known what you were replicating...
Nope, we're talking "ratio", as in the frequency ratio between op a and op b. So, the idea here is that by having those pre-programmed in Layers we don't have to guess/look up what the frequency values are for the harmonic series (roughly speaking, my understanding is that the "Ratio" on a DX_ synth when not set to "fixed" will allow you to select the first 32 harmonics in the harmonic series). I can't remember off-hand where I picked that up, but it certainly could be wrong...
Sorry, duh, that's what I get for watching while half-asleep. So, you only create the "partial" layers as shortcuts for a desired ratio, for a given operator connection/'segment' type, and load them as needed - to speed up setting the frequencies of each op. That makes sense! Ignore me. Fwiw, IIRC most Yamaha FM synths could go up to something like 63 x ratio, but obviously it's of very low usefulness; anything beyond about 20 is sharply diminishing returns. Well, arguably such ratios can contribute to producing interesting noise sources, etc - albeit dependent on lucking into the right combination of frequency, amplitude, sampling rate of the given synth (as it tends to depend on fold-back), etc... it can take a long time to find the perfect hi-hat base. :D
Indeed. For hi-hats though I've found it's best to skip straight to aliasing (aliasing + FM is magical for metallic percussion).
Yeah, aliasing is what I was thinking of, since it's usually a case of picking a crazy ratio and finding the one key that produces exactly the right pitch that the aliasing fold-back has the desired effect. It's both an exact science, and complete trial-and-error!