FM for modern sounds?

[Markyboard]

Aside from FM"s unique character having dedicated envelopes for each operator invites sounds that other synths can"t do. I developed many of my patches around this concept. You can create a series of attacks done in succession that almost sound arpeggiated or sequenced. But the sound from each 'step' can be completely different which is really cool.

 

I remember discussing independent envelopes for typical 3 oscillator synths. IIRC Mike Martin said the Casio XW-p1 can do this. I know you can effectively do this also with multi- timbral synths.

[Markyboard]

Modulating a modulator is where fm gets really hairy. I"m thinking of the 3 stacked elements case as opposed to a feedback loop. Trying to get a playable harmonic sound is extremely challenging, especial over any length of the keyboard. I"d be interested if anyone has tips on doing this successfully. Again I get that you can set the envelope on that 'modulator modulator' for cool percussive effects and such.

 

The reason I bring this up is that I"ve been playing with audio rate modulation of the waveshape on an analog VCO including pulse width. I believe this would be the equivalent of the above but in the analog domain. This results in different sounds from that of the more common modulating of the VCO frequency. Unlike Dx-7 type FM however one can still maintain playability from a harmonic standpoint with the right modulation amount and frequency.

 

I think this is easier to achieve on analog because you can modulate the wave shape without also modulating the oscillator fundamental frequency. Not true on the DX-7 type fm synth, unless I"ve been missing something all these years. Of course on an analog you can modulate both frequency and waveshape if desired.

 

Thoughts?

[ProfD]

Modulating a modulator... I think this is easier to achieve on analog because you can modulate the wave shape without also modulating the oscillator fundamental frequency. Not true on the DX-7 type fm synth, unless I"ve been missing something all these years. Of course on an analog you can modulate both frequency and waveshape if desired.

Brotha Mark, if I'm reading your post correctly, the frequency of the modulating modulator can be adjusted so that the harmonic content remains playable across the KB. At least it works that way on the Reface DX.

 

IMO, that ability to manipulate not only waveforms but modulate modulators in various configurations (algorithms) takes FM programming to another level of sound design possibilities.

 

i looked at the FM programming capabilities of the Yamaha MODX and being able to reconfigure the algorithms way beyond the original 32 is sick in a good way.

 

Needless to type, one could spend a programming lifetime digging the depths of FM especially as it has been expanded in some of the newer products.

[Markyboard]

Maybe I"m forgetting but I believe there are only a few 'benign' frequency relations that yield musical playability results when you get into 3 stacked operators. Maybe I"m thinking clean/pure tones as well without noise riding on top. Again I do know there are tons of other uses for that 3rd tier and even 4th tiered modulator.

 

I"m waiting for my converter box to come back from repair this week so I can revisit FM-8. It really has been too long.

[felis]

You guys would probably enjoy the PC4.

The massive list of modulation options is available to use on both carriers and modulators in the typical amplitude usage, like what the envelopes can do.

 

But what I really like is that those same options can be used on the frequencies of each too.

Using them on modulators produces awesome swirling and smearing effects.

 

Setting all the release times to zero in an FM layer and then running it into a VAST layer, I can finally get FM sounds that respond to release velocity too.

[Markyboard]

You guys would probably enjoy the PC4.

The massive list of modulation options is available to use on both carriers and modulators in the typical amplitude usage, like what the envelopes can do.

 

But what I really like is that those same options can be used on the frequencies of each too.

Using them on modulators produces awesome swirling and smearing effects.

 

 

That does sound intriguing. FM-8 let's you do some of that but with a much more limited set of modulators and real time controllers. And they only modulate the amount of each operator, not the frequency.

 

OK - I'm jealous.

[Tusker]

Modulating a modulator is where fm gets really hairy. I"m thinking of the 3 stacked elements case as opposed to a feedback loop. Trying to get a playable harmonic sound is extremely challenging, especial over any length of the keyboard. I"d be interested if anyone has tips on doing this successfully. Again I get that you can set the envelope on that 'modulator modulator' for cool percussive effects and such.

 

Your subject of "serial FM" deserves a more thorough post with than this one. But just to move the ball down the field by a few inches ...

 

[1=>2=>3=> Output] can indeed be noisy. Small changes in 1 can create large changes in 2 which can create unpredictability in 3 as you you point out. Therefore most people use serial structures for a) percussion or b) attack partials, since these types of sounds are complex and inharmonic.

 

But we can simplify a bit ...

 

[1=>2=>3=> Output] is the same as [2=>3=> Output] where 2 is non-sine.

[1=>2] can create both inharmonic and harmonic wave forms depending on pitch relationship and modulation index.

 

So how do we keep 1=2 harmonic? Carrier and modulator in the same octave tend to produce saw waves. Modulator an octave up gives us a square wave, modulator two octaves up gives us a narrow pulse (33%) and so on. That's pretty safe and familiar terrain. This is one of the reasons, I prefer fewer operators and more waveform types in an FM synth. You can get to the same goals faster.

 

As structures become more serial, you will want more precise control of source material. For example, if you have additive oscillators with fm (as in Logic's Alchemy) you can precisely tailor specific partials being FM-ed. You could generate a saw wave and excise partials 2, 5, and anything above 7. You can also modulate specific partials with envelopes. Your sidebands will be unique, hopefully giving your timbre a special personality long before you get to unusable noise. Similar precision can be had if your synth allows you to insert filters between the operators.

 

It's a question of how far we want to go to find the exotic and the evocative right? That's a moving target. John Chowning's Stria sounded really exotic upon release, but sounds pretty safe and familiar in a post DX7 world. Your thoughts?

 

[video:youtube]

[Markyboard]

 

[1=>2=>3=> Output] is the same as [2=>3=> Output] where 2 is non-sine.

[1=>2] can create both inharmonic and harmonic wave forms depending on pitch relationship and modulation index.

 

So how do we keep 1=2 harmonic? Carrier and modulator in the same octave tend to produce saw waves. Modulator an octave up gives us a square wave, modulator two octaves up gives us a narrow pulse (33%) and so on. That's pretty safe and familiar terrain. This is one of the reasons, I prefer fewer operators and more waveform types in an FM synth. You can get to the same goals faster.

...

Similar precision can be had if your synth allows you to insert filters between the operators.

 

 

You reminded me that it's not so much the range over the keyboard but the frequency range of the modulators I was struggling with. It seems like for all that flexibility there's a lot of "wasted" frequency terrain. I always feel like I'm missing something or not understanding how to use these upper ranges effectively.

 

But the one thing I haven't experimented with other than in mostly random fashion is the initial waveshape. I will definitely be attacking this very shortly and let you know what I think.

 

Great post Jerry - thanks!

[Stokely]

I forgot how pristine and clear FM can sound. I'm on a project that called for vintage analog mostly, but on one I decided let's try some FM (from FM8). Man, should have been doing more of that all along!

[ProfD]

I forgot how pristine and clear FM can sound. I'm on a project that called for vintage analog mostly, but on one I decided let's try some FM (from FM8). Man, should have been doing more of that all along!

FM is another spice (sound) in the rack. It is not Frank's Hot Sauce. Don't go putting that sh8t on everything.

[DrSynth]

Your subject of "serial FM" deserves a more thorough post with than this one. But just to move the ball down the field by a few inches ...

 

[1=>2=>3=> Output] can indeed be noisy. Small changes in 1 can create large changes in 2 which can create unpredictability in 3 as you you point out. Therefore most people use serial structures for a) percussion or b) attack partials, since these types of sounds are complex and inharmonic.

 

But we can simplify a bit ...

 

[1=>2=>3=> Output] is the same as [2=>3=> Output] where 2 is non-sine.

[1=>2] can create both inharmonic and harmonic wave forms depending on pitch relationship and modulation index.

 

So how do we keep 1=2 harmonic? Carrier and modulator in the same octave tend to produce saw waves. Modulator an octave up gives us a square wave, modulator two octaves up gives us a narrow pulse (33%) and so on. That's pretty safe and familiar terrain. This is one of the reasons, I prefer fewer operators and more waveform types in an FM synth. You can get to the same goals faster.

 

As structures become more serial, you will want more precise control of source material. For example, if you have additive oscillators with fm (as in Logic's Alchemy) you can precisely tailor specific partials being FM-ed. You could generate a saw wave and excise partials 2, 5, and anything above 7. You can also modulate specific partials with envelopes. Your sidebands will be unique, hopefully giving your timbre a special personality long before you get to unusable noise. Similar precision can be had if your synth allows you to insert filters between the operators.

 

 

Yes, the more Operators in a serial (Stacked) configuration, the more ''amplifed" so to speak the contribution of the 'top' operators become in the final output timbre. Sidebands beget sidebands beget sidebands beget sidebands etc.

 

So the key is to be mindful of the mod index (output) of the Ops as you go higher in the Stack. In complex waveform FM (anything other than sines) for that same reason you don't want complex waves with a lot of harmonics higher in the Stack - you basically just get quickly to noise/aliasing. Also the reason why acoustic instrument samples make lousy modulators.

 

The frequency relationships are much less important once you keep that in mind. A 20:1 ratio can be extremely pleasant & useful at low index at the top of a 4 high Stack; a 3:1 ratio can be annoying and impractical at high index as a single modulator in a 2 Op stack. It's kind of a wholistic balance between the Ratio/Index.

 

Your comment about fewer Ops, more waveforms definitely keeps things more simple, yet with still a wide timbral playground. Lately Bass aside, IMHO that's the main reason TX81Z/DX100 style 4 Op synth platform kept widespread use in more genres over time than the original DX7 6 Op sine platform.

 

Manny

[Stokely]

I forgot how pristine and clear FM can sound. I'm on a project that called for vintage analog mostly, but on one I decided let's try some FM (from FM8). Man, should have been doing more of that all along!

FM is another spice (sound) in the rack. It is not Frank's Hot Sauce. Don't go putting that sh8t on everything.

 

Heh, no indeed.

 

In fact I really liked how you get that nice clear sound (on the patches I was using) to go with the more analog-sounding Diva and Repro. Especially on one of the tracks that has a lot of synth sounds in early/mid 80s style, too much phatness equals a muddy mix.

[felis]

The Korg opsix looks like it should be able to make some new flavors of FM sounds.

[Dr Mike Metlay]

I'm always fascinated by the tools that players are given to interact with the sounds in FM. (Big duh, coming from the Editor of PUSH TURN MOVE)

 

People who had a chance to work with the DX1 said that not only was it a lot more playable than the DX7 because of its keyboard, but its front panel layout made figuring out how operators went together a lot more clear. The big win on the Korg DS-8 was that its simple 4-op FM was given new controls that felt more instinctive than Yamaha's, even though it was harder to figure out what they were actually doing.

 

Every once in a while, though, we hit a trifecta: a synth that costs way too much, can't be programmed without bloodshed, and sounds like ass. In the FM world, my vote would be for the Simmons SDE, an FM-based rack unit meant to trigger tonal and bell-like sounds as part of an electronic drum kit (it also had MIDI). It was already totally out of date, hated, and dirt cheap when I got one in 1994.. I used it prominently on one album session, and then out it went. I miss it every once in a while, but then I take my meds and I'm fine.

[Al Coda]

 

... the DX1 ... was ... a lot more playable than the DX7 because of its keyboard, ...

 

possibly piano player purist´s point of view.

 

The original gen. 1 DX7´s "synth" action is still the very best synth action created ever.

I also wonder if the old Clavinova´s or KX 88´s action is (very) different from DX1´s.

 

DX1 and DX5 engines are just only 2x DX7.

 

Combine a KX-88 w/ 2 TX-1 or 2 TF-1 modules from/in TX816 and you get the same.

 

And, the DX7mkII FD surpassed at smaller formfactor and w/ less weight.

 

Anyway,-

I think you can get out a LOT of "modern" sounds w/ FM/PM because it´s so flexible when implemented right.

 

Even there are some prefering the so called "grid" of the old DX7 mkI, TX-1 or TF-1 modules from TX-816, IMO the most advanced Yammi FM is still in FS1R rack module.

 

 

but its front panel layout made figuring out how operators went together a lot more clear.

 

Most users didn´t edit at all at that time.

Buying commercial cardridges was it at that time until computers appeared,- like Commodore (SX) 64 and "bankmanager" programs, allowing preset bank dump/load and to make copies of cardridge content.

 

Yammi FM synths offered MIDI not only for layering w/ other in future midified synths.

In fact, it was intentional sacrificing haptics for affordable price and leave in depth editing to software applications to come for the public.

It needed some time until these were created and still think Hubertus Maas´ DX Inventor was the best for Commodore followed by Steinberg Synthworks DX/TX and C-Lab X-Alyzer for Atari ST until Opcode Galaxy (+) editors and Emagic Sounddiver for the MAC became kings of the hill.

 

When there were no DX1, DX5, DX7 and the appearence of MIDI incl. SysEx format, we had never seen ANY editor programs at all and possibly no DAW application too.

 

The big win on the Korg DS-8 was that its simple 4-op FM was given new controls that felt more instinctive than Yamaha's, even though it was harder to figure out what they were actually doing.

 

But then, it´s just only 4-OP FM ... which still has a lot of parameters already.

For perfect control of complex FM, you NEED a computer application and at least several clever macros for the onboard UI.

 

Even I still own KX76-, KX5- controllers and TX816, DX7mkII FD and TG77,- I´m amazed how close it comes soundwise w/ p.ex. a freeware VST like DEXED,- while I think today´s most advanced Yammi FM is FM-X in Montage and MODX models !

 

I´d only wish they´d come up w/ a MODX module.

 

 

 

A.C.

[Tusker]

+1000

 

A hearty agreement on this issue of how to design complex FM to be easy to use. Mark alluded to a problem particular to FM, when the "timbral sweet spots" are not concentrated but widely dispersed in little puddles across a complex option-set? I would suggest two analogies.

 

The first analogy is to Amazon, where through the use of learning algorithms, seemingly unrelated product options can be brought to the imagination. You might be searching for a snow shovel and miraculously learn that people who searched for snow shovels also searched for snow blowers and Christmas sweaters. Currently digital synthesizers allow you to catalog your sounds, morph between patches, create offsprings from patches and randomize certain aspects of a patch. Learning algorithms can take it a step further. Imagine an FM synth where you could morph between a pad with one structure and a plucked sound a completely different structure, but the algorithm is predicting which intermediary structures will provide the smoothest timbral path. The software may even warn you when a discontiguity is about to happen. The software would have to "know" synthesis in the same way that Izotope "knows" some of the adjustments which can make my crappy mix just a bit more like that professional mix from a particular song.

 

A second analogy I would use is auto accompaniment. An example of auto accompaniment is where you might play a set of notes or chords and Karma creates an entire orchestra using musical tropes, like "bass", "obstinato" "arpeggiation" and so on. A similar approach can be utilized in sound design, using tropes like "inharmonicity" "envelope tightness," "brightness" and so forth. The algorithm would analyze your playing and adjust timbres in real time to suit the musical purpose of the expression. If I am playing 16th notes, the envelopes could be slightly tightened, the brightness could be reduced and a tiny bit of inharmonicity in the attack may provide the "bite" that the 16th notes are calling for. One might need to analyze scores and audio recordings to develop the taste.... a musical equivalent of google ngrams. Some of this type of work has been done to create services like Pandora, so as computing power increases, why not expect similar advances to be implemented into the design of sound, so that musicians can focus on the music?

 

Improving UI will increase the market size, so I imagine manufacturers will increasingly focus on user experience as the problems of creating good sound become less herculean.

[DrSynth]

+1000

 

A hearty agreement on this issue of how to design complex FM to be easy to use. Mark alluded to a problem particular to FM, when the "timbral sweet spots" are not concentrated but widely dispersed in little puddles across a complex option-set? I would suggest two analogies.

 

The first analogy is to Amazon, where through the use of learning algorithms, seemingly unrelated product options can be brought to the imagination. You might be searching for a snow shovel and miraculously learn that people who searched for snow shovels also searched for snow blowers and Christmas sweaters. Currently digital synthesizers allow you to catalog your sounds, morph between patches, create offsprings from patches and randomize certain aspects of a patch. Learning algorithms can take it a step further. Imagine an FM synth where you could morph between a pad with one structure and a plucked sound a completely different structure, but the algorithm is predicting which intermediary structures will provide the smoothest timbral path. The software may even warn you when a discontiguity is about to happen. The software would have to "know" synthesis in the same way that Izotope "knows" some of the adjustments which can make my crappy mix just a bit more like that professional mix from a particular song.

 

The SmartMorph in the Montage/MODX actually can does this quite well, but there are mathematical constraints you have to keep in mind that are the equivalent to repatching cords on a modular synth in real time. Specifically, changing Algorithms and Ratios cannot smoothly interpolate in real time, they're 'hard switching'. But if you set all the SmartMorph parents to use the same algorithm and ratio structure, fully seamless and smooth morphs and realtime control is the result.

 

If you have a Montage/MODX make a patch in Part 1 using Algorithm 68, setting up Ops 1-7 with uniquely different ratios, waveforms, envelope shapes and Levels. Copy it to Parts 9-12 to be future parents. Then go into each of those parents and reset the envelope parameters to something completely different for each OP, as well as the change all the levels, including turning off some (level =0) i.e. Level 0 for Ops 1, 2 in Part 9; maybe Ops 3 & 4 in Part 10. etc so each parent has different combinations of 'active' modulators of different levels and envelope contours. Then change the envelope for the Carrier Op 8 in each Part so one is percussive, one is a slow pad, another a attack with swell etc. Save them all.

 

Then run SmartMorph (sometime try running it 2, 3 or 4 times in a row with various Morph settings) and a whole lot of cool stuff will result out, and all smoothly continuous across the entire touchscreen. You'll have both subtle and extreme timbral and behavioral results depending how you set your parent parts.

 

Manny

[Markyboard]

All these algorithmic aided creation tools are fascinating and make these complex synths useful to so many who otherwise wouldn"t bother. But it"s an early exit for me as I completely lose interest. If it helps in understanding fm or whatever synth technology we"re talking about than yeah, I"m all for it. But otherwise it"s like traveling using one of these guided group tours instead of winging it yourself.

 

More efficient and way more relaxing, but where"s the adventure?

[DrSynth]

All these algorithmic aided creation tools are fascinating and make these complex synths useful to so many who otherwise wouldn"t bother. But it"s an early exit for me as I completely lose interest. If it helps in understanding fm or whatever synth technology we"re talking about than yeah, I"m all for it. But otherwise it"s like traveling using one of these guided group tours instead of winging it yourself.

 

More efficient and way more relaxing, but where"s the adventure?

 

A lot of truth to that. It will be interesting how the Korg opsix is received, and if they found the right balance between ease of use and complexity/versatility to entice enough people to start out on an adventure without being scared about getting lost in the Amazon.

 

Manny

[Tusker]

I think the tour guide analogy is very good. You've got some people who just want a guided tour, others who want the real experience (but only for a week) and yet others who live as expatriates for years, deeply learning the ecology, the languages, the cultures. In FM, the "tour guide interfaces" haven't been there, especially in the beginning. A lot of us (guilty!) bought cartridges of pre-programmed sounds because DX7 FM was too tough a nut to crack with just one slider. Or it just took too long. One reason I haven't downloaded the very capable free synth Dexed, is that it unfortunately reminds me of my DX and Nord Modular experiences. It's hard to see the complete picture of a sound from the parameter level.

 

Opsix is cool. So are FM8 and and the new toy in my laptop, Flow Motion. Here's to all the attempts to make FM easy to use so that more of us can learn more of it's subtleties.

[felis]

Both the opsix and the Wavestate from Korg have release velocity listed as a mod source.

While I haven't had a chance to try either one yet, I hope this is the beginning of a trend in that direction.

 

My first FM and some of my early computer experiences were one and the same - a Yamaha CX5m music computer.

https://en.wikipedia.org/wiki/Yamaha_CX5M

[Markyboard]

I finally spent a good couple of days with FM-8. Yesterday I just took one of my old patches as a starting point, disabled all but 3 operators and went to town making it something different. In about 15 minutes I had this shimmering tinkly thing that sounded good. I made a couple more versions adding a couple more operators to fill out the lower frequencies.

 

Today I spent hours and... got nowhere until the very end. I was determined to make use of those higher frequency ratios which have a range of 0-64. I also wanted to use 3 stacked operators. This is an experimental geek exercise. Based on a previous discussion (above) I was determined to come up with a playable sound over as many octaves as possible. I started with the carrier at about 32. Any higher and the upper octaves are inaudible. Now the modulator needs to be fairly high up there as well but still lower than the carrier frequency to generate sub harmonics. Once I stacked a 3rd operator this is where it got real messy.

 

Also mentioned above the mod index aka operator level has to be fairly low on the modulators as this is basically a multiplier. And I did use sine waves. I tweaked and tweaked. Got the garbage can lid sound a number of times and eventually threw it away only to start all over several times. Fun stuff.

 

 

Eventually I stumbled on something I liked by having the carrier modulate what was the upper modulator. And there are some high frequencies testing my ears. Such is fm programming. I still stand by my previous comment that a significant amount of 'frequency range' is unusable for playable harmonic sounds. I struggled with the highest operator ratio at 32. There"s still half of the range left. Also It seems starting with a high carrier frequency and generating lower harmonics would yield the same sound as the reverse, all else being equal.

 

 

I still feel like I"m missing something here. Still listening if you guys have further suggestions or better yet examples.

[Tusker]

You are very brave brother Mark. FM is as you say, mostly trash can.

 

But what a trash can! As a teenager, I loved to go to concerts with our local youth orchestra. A lot of my friends were playing and it was free! The star of show was always the main percussionist. He had a blonde mullet and he had muscles like Chris Hemsworth. When he played, he'd swing those mallets and sticks like Thor swinging Mjollnir to save the world. The audience swooned when he played and he knew it. When he wasn't needed for 10 minutes, he would sprawl off on a chair as though nothing else important was going on. Then, he'd bounce up again and usually the orchestra would be wild and loud and dramatic again.

 

I mention him, because his instruments were the FM of the orchestra. Whether it was a Timpani roll or a cymbal swell or a Glockenspiel the key was inharmonicity. The thirty other instruments were like subtractive and additive synths. Tuneful but lacking in drama. Humans invented the drum (a skin over a pot) and flute ( a bone with holes) tens of thousands of years ago and everything in our musical culture can be considered descendant from these two extremes, inharmonic and harmonic, at war and peace with each other. Even today, West Coast with it's creative sequencers, LPG's and wave shaping/FM is a complement to East Coast with it's warm brass and strings. It's not that one can't do what the other side does, but each has it's musical strength. Thanks for sharing your musical adventures. You've encouraged me to put aside the Christmas Card list and do more patching.

[Tusker]

Still listening if you guys have further suggestions or better yet examples.

 

I don't know if this is what you had in mind, but here's some FM I've been totally enjoying. Matt Celitti is FM-ing a vocal filter with a couple of oscillators. He seems to be employing three techniques to obviate issues we have been discussing:

 

- Use of the vocal filter, (a narrow frequency-range filter) limits the sidebands and makes it easier to sound like hand drums and small percussion. I guess alternatively, you could band-pass filter the FM output, but this approach yields quite an open (airy?) sound.

- Use of the sequencer allows you to cherry-pick the sweet spots, which (as you mentioned) can be quite small relative to the potential range of sounds.

- Putting the fm-ed signal through a delay and reverb helps humanize the sound. Some DAWs have snare or bass drum (or other instrument body) convolution impulses which can help the FM sound more like a physically modeled instrument.

 

What do you think?

 

[video:youtube]

[Markyboard]

What do you think?

 

I think you are exactly right. Very much like modular in this regard. I can't stand listening to the repetitive non-key changing sequences that go on forever. But I hear sounds and think man, I would love to play that sound across the keyboard. Only you cant. It's like you say, these inharmonics by their very nature exist in a small range.

 

That OK, I'm not against inharmonics; they too have the right to exist in nature (envisioning the next big protest movement ).

 

I do tend to focus on harmonic type sounds and for this go-round I'm trying to better understand and exploit fm and it's limitations. I think its much easier to come up with cool inharmonic sounds where those upper ranges will be of more use. I just need to torture myself a bit more before going there. Btw I totally agree effects bring these sounds alive. That shimering tinkly sound I mentioned wasn't shimmering and barely more than a dry tink before I applied the "PshcheDelay" in FM-8.

 

As always I appreciate your perspective.

[ProfD]

I still feel like I"m missing something here. Still listening if you guys have further suggestions or better yet examples.

Brotha Mark, I'm fairly cerrtain that you've for the most part explored/exploited the "flavor" FM. Again, I consider FM a spice in the sonic rack. I use FM sounds to complement my EP.

 

FM is as you say, mostly trash can...inharmonic and harmonic... It's not that one can't do what the other side does, but each has it's musical strength.

I've consolidated brotha Jerry's post into this quote because I think it captures FM from my perspective. The inharmonic aspect of it is my secret weapon. It provides a really nice edginess to the sound(s).

 

I doubt that FM will ever be a sound designer's wet dream because it is more cumbersome and the diminishing returns add up.

 

But, in it's own space, FM is extremely capable of producing a variety of musically ineteresting sounds. It is certainly the only "synth" I need.

[Markyboard]

A short demo of those 2 patches I mentioned. Not so sure about the quality when streaming these higher frequencies but hopefully you get the idea.

 

 

FM-8 Demo

[Tusker]

Nice to hear your adventures in person and I think your description matched what I heard pretty well. I agree that SoundCloud seems to have some issues with higher frequencies. They probably sounded a tad sweeter on your monitors before the SoundCloud algo did it"s thing.

 

We all have our go to sounds with which we first proof-test a synth or a concept. For me it"s usually a filter sweep or a pitch sweep. For one of my friends it"s low frequency stuff like bass and kick. Would it be fair to say that one of things you look for a synth to do is a sound which can carry a melody and yet adds something fresh to the sound design? It"s not my place to say but that sounds like a very valid and sensible musical goal.

 

I find FM by itself to be not as good as analog for that kind of thing. Hopefully somebody will prove me wrong, lol. One exception could be when it"s surrounded by super warm timbres and the 'clarity' provides contrast kinda like a piccolo in a thicket of French horns.

 

Still in combination with other synth techniques as (ProfD said), it seems to be ok to me. I wouldn"t put up an FM lead sound against a Knifonium or a Swarmatron. It"s ... different. You could do the classic detuning with FM of course but you might need to use multiple instances of FM8. I used to run FM through a real analog filter but haven"t felt a need to do that recently as DSP filters have improved. I also use body modeling a bit as I mentioned above. It seems to take the edge off when you need it to.

[Markyboard]

I originally layered the high frequency contributing operators with an Andromeda patch. I liked that. But I wanted a stand-alone patch for the demo so I added a couple of operators just for a lower pad sound. Nothing too spectacular compared to using a good analog.

 

I don"t recall ever attempting a lead sound like you would get from a mini or ARP. Until you mentioned it the term filter sweep never really entered my thoughts with fm. Envelopes are used judiciously to shape the timbre over time, but filter sweep? Not really.

 

Those tinkly highs are one of fm strengths. As my buddy Harry says, a synth"s gotta know it limitations.

[ProfD]

A short demo of those 2 patches I mentioned. Not so sure about the quality when streaming these higher frequencies but hopefully you get the idea.

Thanks for sharing brotha Mark. The idea came through.

 

Taming the high frequencies takes a little massaging but it's possible.

 

Great to hear FM-8 if I ever decide to take the Native Instruments Komplete plunge.