FM synthesis / Analog synthesis sound/technology difference?

[Borogove]

it's easy to get confused by we hair-splitting, pedantic and prone-to-obfuscation KVR members.

I'm not pedantic, I'm precise.

[emdot_ambient]

it's easy to get confused by we hair-splitting, pedantic and prone-to-obfuscation KVR members.

I'm not pedantic, I'm precise.

How pedantic of you.

BTW, I'm not a snob either, I'm just highly diacritical.

[aciddose]

i'd like to point out as usual the differences between fm, phm, and other types of modulations.

fm, typically, is a linear modulation of frequency. what you normally have on subtractive/analog synthesizers is an exponential frequency modulation input, which means you get different results. some provide connection points for both, all transistor exponential convertors allow for a linear input if you're capable of very basic modding.

the lfo (as used for a vibrato) will always be an exponential modulation, and not linear meaning there will be a vast difference in the behaviours you experiance.

phm though, is something completely different. phm or phase modulation, is used in all of the dx series and later synthesizers claiming to be 'fm' synthesizers. phm gives yet another type of behaviour for the same parameters.

even once you understand the basics of different synthesis techniques, you will probably still not understand this difference. once you get to the point where you need to know, you should learn more about the differences between the systems.

you may find that using an lfo at audio rate on a subtractive synthesizer does not produce the same results you'd expect from something like the DX-7. what i've tried to explain will be the reason for that.

but to answer your original question: they're completely different systems which have very little in common. like has been said so far by others many modern synthesizers are actually hybrid designs implementing many types of synthesis. you can not make any even remote comparison between something like DX-7 and minimoog, they're completely different.

[funkadil]

Soundwise I believe they can both create similar sounds, but some do certain sounds much more efficiantly.

[Rajiv]

The original question is one I am interested as well -- is there any difference between the types of sounds that can be produced by the different types of synthesis?

I understand that subtractive/analog vs. additive vs. FM are different ways of producing sounds. However, isn't this about the how and not the what? Are there sounds (or types of sounds) that cannot be produced from one type of synthesis vs. another?

If not, why are there so many different types of synths out there? Does it all come down to the workflow?

Here's a corollary question -- even between the same type of synths (say... FM) are there sounds one FM synth can produce that another cannot? If not, why are there so many FM synths?

Or is it about the presets? The marketing?

[aciddose]

there are so many variables involved in the design of a synthesizer that no two synthesizers will really be able to sound exactly the same, even if you spend a lot of time making a patch as close between the two as possible. for example if i compare xhip to <some va vsti>, there is no way either synth can mimic the other perfectly. generally xhip can do most if not all of the sounds other synths can, but there are certain things i choose not to implement. xhip for example doesnt have three oscillators. it does have pcm.. but it isnt the same. very basic differences like this, with number of oscillators, type of filter, type of modulation.. those things add up and make it so that it is nearly if not completely impossible to mimic one synth perfectly with another.

its like "can an apple be like an orange?".. well, they're 'alike' to begin with. can one completely emulate the other? the answer SHOULD be obvious! exactly the same thing applies to "can a subtractive mimic an fm?".. it just seems that to most people, the answer is not obvious in this case.

[Borogove]

I understand that subtractive/analog vs. additive vs. FM are different ways of producing sounds. However, isn't this about the how and not the what? Are there sounds (or types of sounds) that cannot be produced from one type of synthesis vs. another?

The answer is: it's complicated.

A 3 osc analog subtractive without FM routings can't do rich enharmonic sounds like an FM synth can.

A 3 osc analog subtractive with FM routings is an FM synth.

Most analog subtractives include saw, square, and triangle waves as basic wave types, and so in the presence of FM routing capability can actually make some sounds that an FM synth that only offers sine waves can't (given the same number of oscs).

Most of the Yamaha FM synths offered rectified-sine, half-sine, and other waveforms that aren't often seen on analog synths, and so they can make some sounds that an analog subtractive with FM can't.

An analog subtractive (or an FM synth) with enough oscillators becomes an additive synth if you switch each oscillator into sine and turn off the filters.

Any synth which has oscillators that can produce more than a sine, and has a filter, is a subtractive -- that's most FM synths right there.

If not, why are there so many different types of synths out there? Does it all come down to the workflow?

Here's a corollary question -- even between the same type of synths (say... FM) are there sounds one FM synth can produce that another cannot? If not, why are there so many FM synths?

A 2-osc subtractive VA synth can make sounds a 1-osc synth can't. A 3-osc synth can make sounds a 2-osc synth can't. A synth with separate envelopes for amplitude, pitch, and filter cutoff can make sounds that a synth with one envelope which can modulate amplitude, pitch, and filter cutoff by variable amounts can't. A synth with 12 sine/square/saw/tri LFOs can make sounds that a synth with 1 LFO can't. A synth with one Lorenz/Rossler chaotic LFO can make sounds that a synth with 3 sine-only LFOs can't.

A synth might have 50 knobs to control its parameters. You could put a different UI on the same synth that consisted of a single knob which interpolated linearly through 100 different random points in that 50-dimensional space. One of those two would reward blind exploration by ear. The other would reward understanding exactly what each parameter affected.

[VicDiesel]

The answer is: it's complicated.

Quite.

The oscillators in subtractive synthesis all conform to the fact that they generate f,2f,3f,4f,.... with diminishing amplitudes. All you can do with filters is further diminish certain partials (ok, and hitch up some of them), but the basic overtone series stays intact.

FM does something completely different and I'm not sure how to describe it. The "overtone" series does not necessarily consist of multiples of the fundamental, and what left sidebands mean in acoustic terms is completely beyond me.

Maybe I should read some of the old Chown papers.

Victor.

[Rock Hardbuns]

Unless you count osciallator sync.

[VicDiesel]

Unless you count osciallator sync.

Ok, that's more complicated than just filtering. However, I still don't think that gives you hte weird inharmonious overtones of FM synthesis. You get multiples of the frequency of the syncing oscillator.

Victor.

[Borogove]

The oscillators in subtractive synthesis all conform to the fact that they generate f,2f,3f,4f,.... with diminishing amplitudes. All you can do with filters is further diminish certain partials (ok, and hitch up some of them), but the basic overtone series stays intact.

FM does something completely different and I'm not sure how to describe it. The "overtone" series does not necessarily consist of multiples of the fundamental, and what left sidebands mean in acoustic terms is completely beyond me.

FM partials appear at integer multiples of the modulator frequency (or of each partial of the modulator if it's more complex than sine) above and below the carrier. If the modulator is an integer multiple of the carrier frequency, then as with a simple oscillator, all the partials are harmonic; if not, the partials are inharmonic.

But in a 2-osc subtractive, you can detune your oscs so that their frequencies are no longer in a simple harmonic relationship, just as with a 2-operator FM synth ("operator" being Yamaha-FM-ese for "oscillator"). More complex FM setups rapidly increase the complexity of the relationships among the partials, true, but even so, this isn't as fundamental a difference from analog subtractive as it might seem at first.

(What I'm trying to say here is that it's not fair to match a multiple-operator FM voice against a single subtractive osc; it's more fair to match operators against oscillators.)

[aciddose]

"A 3 osc analog subtractive with FM routings is an FM synth."

but not fm in the way it is used when you're talking about yamaha "fm" :P

[Borogove]

"A 3 osc analog subtractive with FM routings is an FM synth."

but not fm in the way it is used when you're talking about yamaha "fm"

I don't consider the difference between linear FM and PM to be all that significant.

Exponential FM is a capybara of a different color, but even so, for "special effect" sounds, it's no more or less valuable than PM.

[afreshcupofjoe]

The original question is one I am interested as well -- is there any difference between the types of sounds that can be produced by the different types of synthesis?

For all practical purposes YES, the sounds they make can be altogether VERY different. If you want to get really technical, then no, you should in theory be able to reproduce any sound with either type of synthesis. However, the complexity of the synthesis that would be needed to produce the same sounds using each technique would be unfeasable. Most synths are somewhat limited by their number of oscillators and routings. So for all practical purposes, you can create wildly different sounds using each synthesis technique.

What makes the issue confusing, and what most people are griping about, has to do with the fact that most synths on the market are not limited to one type of synthesis. Most good VA (subtractive) synths have some rudimentary fm capabilities, and most fm synths can do decent subtracive techniques. This is because both types of synths use the same basic sound generating and signal modulating components. The general difference between VA and FM synths are that FM synths are designed with FM in mind and have all the necessary components and routings to do complex FM easily, but they often lack in the grit and character of the filters and ocillators used in VA. VA synths which are primarily designed for subtractive synthesis really lack in the number of oscillators and routings necessary to do FM well. Yes, you can do FM on an analog synth, but the number components you would have to wire together for a good FM synth is just ridiculous, and there are also issues with the tolerance of electrical components, so it is rarely done.

I guess the short answer is: Most synths that are labeled VA (or subtractive) synths can do some FM synthesis, and most so-called FM synths can do subtractive, but for all practical purposes the sounds you can make on them are wildly different.

If not, why are there so many different types of synths out there? Does it all come down to the workflow?

Here's a corollary question -- even between the same type of synths (say... FM) are there sounds one FM synth can produce that another cannot? If not, why are there so many FM synths?

Or is it about the presets? The marketing?

A lot of it is workflow, which is very important. You don't want to be flipping through pages of operators when you are just trying to make a squelchy single osc bass. You just can't have everything at once. Although some synths come close to being able to do everything, it usually comes down to a balance of features and useability, and there is no "perfect solution. Yes, the market is a bit flooded, but just be happy having so many great choices. For a synth with a good balance between fm and subtractive capabilities that's dead easy to use, I highly recommend Toxic III.

[afreshcupofjoe]

Looks like I'm the threadkiller again!