Synthesis and Sound Design Techniques

Bleed controls pleeease

Any/all of the above! :slight_smile:

1 Like

One thing I’ve always struggled with is a good source oscillator(s) for long-ringing metallic things. E.g. crash, ride, splash etc. cymbals. Have you done any exploring there?

1 Like

Everybody has questions, but for those of you who have insight, knowledge and are prepared to bless us lesser mortals, please, share don’t shy with … With … With…

Learning here is giving, it’s unconditional knowledge whilst being powerful ! When Milton Wrote the poem of paradise lost he pleaded with his audience, understand human frustration… embrace the weak, don’t expose them for being lesser than thou. Thou art great, greatness out of shared ownership, no strings attached, you are the light…light my path that I might elevate you

Sorry, it’s not a religious thang, solomente humanidad, y nada mas

@Joe: are you trying to encourage me here :rofl:

well, as for the “source oscillator(s)” fm synthesis is the first thing that comes to mind…
(with physical modelling being the second thing)
i remember two chaps from the forum who presented quite a beast of a dual oscillator
which they called “Evil Twin” for a reason:
and there was also this creature from @Evs for this “This Land”:

there are even more, just search for ‘fm synth’ in the forum…

i’ve read in several books and forums that fm synthesis is strongly recommended especially for
ringing and metallic things. now, why is that?

Historically speaking, wait, maybe i should rather say: hysterically speaking
in a land before our time there seemed to be exactly two kinds of creatures.
One type of creatures seemed to be exploring the possibilities of imitating “traditional instruments” by the means of electronic circuitry.
And the other type? Some say they just don’t cared about… basically anything. (which i believe is flatout wrong to say about them) But especially when it came to electronic circuits they didn’t seem
to honour our western musical tradition…
some say that it was the first type of creatures that had invented fm synthesis some even spun out a conspiracy theory that the main reason was that they saw more profit in imitating instruments their customers were already familiar with.
“back to life, back to reality”: in history you can observe for each newly introduced technology that there is a tendency (not a law) for imitating their corresponding predecessors (e.g. radio imitating theatre, especially when stereo came up). so when the first electronic instruments were presented to a broader public it must have been obvious that those inventions would have to live up to the most popular traditional instruments of that time. of course, piano came to mind and believe it or not: bells were also kept in the hall of fame of (western!) musical tradition. and i believe that in the so called western tradition those instruments became the predecessors of what we nowadays refer to as
“long ringing metallic things”. BTW, some of this is deliberate and total BS! but some of it is actually more than just educated guesses.

technically speaking, i’ll better hide behind this quote from the wikipedia entry mentioned above:
“FM synthesis can create both harmonic and inharmonic sounds. For synthesizing harmonic sounds, the modulating signal must have a harmonic relationship to the original carrier signal. As the amount of frequency modulation increases, the sound grows progressively more complex. Through the use of modulators with frequencies that are non-integer multiples of the carrier signal (i.e. inharmonic), inharmonic bell-like and percussive spectra can be created.”

anecdotically speaking
a used Casio VZ 1 was the second synth i owned (it’s also one of the few things i still regret to have sold…) though that thing was great in many respects,
it especially spoke to me when i went percussive and/or metallic with it…and i didn’t understand at that time that its fm synth engine was one of the main contributors to that effect…

as for the “long” in long-ringing metallic…
well, a “bleeding VCA” is one, though limited, tool to let it rinnnnnnnnnnnnng…
possibly next time.

@joncharliefeathers: sorry, initially i thought Joe’s question about cymbal-types of sound was an easy one. and it is, but i totally went essayistic there. somehow he pulled a string on me :cold_face:
which again reminds me of physical modelling, which was the second…rinnnnng


Now you’re talking x

That suggests an orthodox attitude y’all

1 Like

Yes, absolutely. :slight_smile:

I agree, in a world where I can download countless cymbal sounds from the internet for free (and buy even more), it would seem the only point in learning to synthesize them would be to get one step closer to finding the question to the answer of life, the universe, and everything [which we all know is 42]. I mean, I could play back a sample cheaper (from an ER-301 CPU perspective) than I could synthesize it, and it would obviously sound very realistic.

But I guess part of the point is that understanding how to get very close to the real thing would also lend an understanding toward starting to deviate from the real, and create something more surreal that still hearkens back to the real.

I haven’t really found anything in FM that really satisfies. I’m not saying it can’t be found; just that I haven’t found it.

One thing I haven’t really explored is using the granular players to try to extract the fundamental components of metallic sounds and then resynthesize them into something new. That might prove interesting. Would it be cheating?

1 Like

@mopoco god darn it if I didn’t know better you just a hill Billie with a big gun!

You realise more than 700 cowboys and cowgirls swing off the tip o your tongue for sharp knowledge! You are the sheriff? Ain’t that so?

Not at all! I remember experimenting with it back in the day when I had the Waldorf NW-1 module. It came with a little utility that could take an arbitrary WAV file and resynthesize it with sine waves for use in the NW-1. Honestly, as quirky and kind of unwieldy that module is, I regret getting rid of it.

1 Like

i disagree! most of us still think it would be “42”. but the true answer actually is “43 with garlic”.
which also brings us closer to the question this awefull planet is still computing on to find out.
seriously you should try both and compare, you’ll see i’m absolutely right. hold on, what was the phone number of that girl again?

for a moment, let us lock away those rather traditional concepts of ringing and metallic. and let me ask you this: do you also like percussive sounds that go with long decays and somehow feel ‘wooden’ at the same time?
koto? 18" floor tom? marimba?

that’s why i love physical modelling: on the one hand it gives us a far better understanding as well as a far better simulation of what is actually going on physically in certain classic instruments. And on the other hand you’ll lay your hands on such an equation and mess with its parameters. blown guitars, struck horns, bowed bass drums…

yes, absolutely! :smiling_imp:
but first of all, in this case it seems “mostly harmless” to me.
and most importantly i’d love to seefeel hear your findings!

1 Like

My 42 cents is that FM marvels tend to be obscured by traditional (DX7 et al.) approach to control, making editing sounds in many cases sort of not worth the time. I think we have to be “thankful” solely to DX7 for what can be called “presets takeover” and “preset mentality” of many “producers”, if i understand the word correctly (sorry, not a native English speaker here). However, for example, in the ER-301 with its transparency and customizeable approach to control FM becomes an endless source of fun, at least for me. Also, FM can be a part of more complex things (as in Buchla and post-Buchla land). Also, FM doesn’t always have to be drastic, sometimes a touch of it is enough to animate otherwise dull things. Just my opinion, however. And thanks for all the fish.


Something that comes to mind for metallic percussion is the method the 808 used for cymbal sounds – stacking 6 square waves, and running those through a bandpass filter. (A little simplified – the actual 808 had parallel filters apparently. But this totally works.)

Like this:

808noise.unit (20.4 KB)

This is just the 6 squares and a bandpass in a mixer channel container, no VCA or envelope. I used some reference frequencies I found and then tweaked them because they sounded a bit too harmonic for my tastes, but didn’t spend a lot of time on it, so it could possibly be improved or just adjusted to taste.

Probably sampling this and using the sample would be the most CPU-efficient way to go.

I know Erbe-Verb is pretty good for gongs and cymbals etc.-- modulating the tilt EQ to emphasize the strike – and that makes me think that on the ER-301, something with Schroeder Allpass filters in a feedback loop might work?


thanks for the unit @Starthief!

FYI @Joe: since this was the second unit preset contributed through this thread i took the liberty to add the custom unit tag so we can better track them down when we’ll add them to our List of User Contributed Creations

and you can sue @joe_biomassa for this because he drove me to revisit a scene in Adams’ Holistic Detective Agency and now i have to share it here. i can’t help it, i just have to:

"St Paul’s glowed in the dark distance and he stared at it for a moment or two but it didn’t do anything special. After the events of the evening he found this came as a pleasant surprise.

At the other end of the room were a couple of long tables smothered in, at the last count, six Macintosh computers. In the middle was the Mac II on which a red wire-frame model of his sofa was lazily revolving within a blue wire-frame model of his narrow staircase, complete with banister rail, radiator and fuse-box details, and of course the awkward turn halfway up.

The sofa would start out spinning in one direction, hit an obstruction, twist itself in another plane, hit another obstruction, revolve round a third axis until it was stopped again, then cycle through the moves again in a different order. You didn’t have to watch the sequence for very long before you saw it repeat itself.

The sofa was clearly stuck.

Three other Macs were connected up via long tangles of cable to an untidy agglomeration of synthesisers - an Emulator II+ HD sampler, a rack of TX modules, a Prophet VS, a Roland JX 10, a Korg DW8000, an Octapad, a left-handed Synth-Axe MIDI guitar controller, and even an old drum machine stacked up and gathering dust in the corner - pretty much the works. There was also a small and rarely used cassette tape recorder: all the music was stored in sequencer files on the computers rather than on tape.

He dumped himself into a seat in front of one of the Macs to see what, if anything, it was doing. It was displaying an ‘Untitled’ /Excel/ spreadsheet and he wondered why.

He saved it and looked to see if he’d left himself any notes and quickly discovered that the spreadsheet contained some of the data he had previously downloaded after searching the /World Reporter/ and /Knowledge/ on-line databases for facts about swallows.

He now had figures which detailed their migratory habits, their wing shapes, their aerodynamic profile and turbulence characteristics, and some sort of rudimentary figures concerning the patterns that a flock would adopt in flight, but as yet he had only the faintest idea as to how he was going to synthesise them all together.

Because he was too tired to think particularly constructively tonight he savagely selected and copied a whole swathe of figures from the spreadsheet at random, pasted them into his own conversion program, which scaled and filtered and manipulated the figures according to his own experimental algorithms, loaded the converted file into /Performer/, a powerful sequencer program, and played the result through random MIDI channels to whichever synthesisers happened to be on at the moment.

The result was a short burst of the most hideous cacophony, and he stopped it.

He ran the conversion program again, this time instructing it to force-map the pitch values into G minor. This was a utility he was determined in the end to get rid of because he regarded it as cheating.

If there was any basis to his firmly held belief that the rhythms and harmonies of music which he found most satisfying could be found in, or at least derived from, the rhythms and harmonies of naturally occurring phenomena, then satisfying forms of modality and intonation should emerge naturally as well, rather than being forced.

For the moment, though, he forced it.

The result was a short burst of the most hideous cacophony in G minor.

So much for random shortcuts.

The first task was a relatively simple one, which would be simply to plot the waveform described by the tip of a swallow’s wing as it flies, then synthesise that waveform. That way he would end up with a single note, which would be a good start, and it shouldn’t take more than the weekend to do.

Except, of course, that he didn’t have a weekend available to do it in because he had somehow to get Version 2 of /Anthem/ out of the door sometime during the course of the next year, or ‘month’ as Gordon called it.

Which brought Richard inexorably to the third thing he was shaking about."


whoa. i love the book but it’s been a while and I’ve completely forgotten that it has that very precise itinerary of gear STILL highly relevant to us! how awesome. thanks for the reminder! and compositional process — might remind of something disquiet to fellow lines members. Re-reading the book now.

p.s. “Performer”, as it turns out to be, is the midi only, pre-audio MOTU Digital Performer.


@Joe: i didn’t try @Starthief’s noise unit yet. but his description reminded me of a technique that i actually try to avoid (most of the time). i.e. tones that some filter types add to the source when you turn up resonance at upper mid cutoff frequencies (somewhere around the main frequencies annoying heavy metal guitars work with). i believe no matter which source you choose a resonant filter could add a metallic feeling to it. add water to that cutoff frequency (i.e. joe_biomassa’s advice that “FM doesn’t always have to be drastic, sometimes a touch of it is enough to animate otherwise dull things”) and while constantly and carefully stirring that soup: pour it carefully down the drain. et voila!


That inspired me to play around a bit more.

Velvet noise -> Joe’s Ladder BPF -> Linear Unipolar VCA

It can help to use the same envelope to control the velvet noise frequency as the VCA. The lower that frequency the more rough and “shattered” the sound is (something a bit like WMD Fracture at times).

The dual peaks on the BPF (as you tweak the base frequency and bandwidth) interact with each other in interesting ways and create some FM-like artifacts, and increasing Q lowers the noisiness and tends more toward pure resonance.

Then you can stick a sample scanner after it (maybe in a mixer channel) with any single cycle waveform, if you want to make it noisier

:raising_hand_man: Go on… :drooling_face:

those weren’t meant to be rhetorical questions. i was actually trying to get at something that i was suspecting…

neither was it meant to be “percussive pron”…
wait a minute… maybe i should… go on! :smirk: