Synthesis is Nothing New: Getting Creative with DSP
Archive · 20089 old posts
From the original SynthMaker forum (2004–2014), with its members’ names and dates.
chironex ·
Hello.
We've all heard of subtractive synthesis before. I mean, we can barely make a subtractive synth in SM that can't be done elsewhere. There's FM, PM, AM, RM, and Phase Distortion. These are all said and done. They're nothing new. Additive synthesis is fairly common. Etc.
In my opinion, a synthesizer is truly creative when it uses more, newer methods for synthesis. Take Tone2's HCM, for example. I'm tried FireBird's demo, and it sounds really cool. There are tons of new and odd forms of synthesis, waiting to be taken into the mainstream?
So why can't we do that in SM?
Well, who said we can't? SM is a powerful system, and with time, it will grow into a tool we can use to do whatever we wish. At this point in time, though, there's still a lot to be done...array support, for example, could be better. Yet, there are still probably tons of methods of synthesis which would be simple to implement in SM waiting to be discovered.
Thus, the purpose of this thread: To put our ideas together, and see if we can come up with some new techniques of sound synthesis. Don't doubt that we can.
To get this rolling, I'll contribute an idea I've been considering...
Start out with a database of mathematical functions, preferably "bumps". The user would then give the x-value of the peak, an be able to stretch and condense the function. Then, the values would be taken at X=1, 2, 3, 4, 5, etc. This would then be stored in an array as raw harmonic data. There would also be array where, perhaps, the values would either be 1 or 0 for each harmonic, such as 1 for all evens, 0 for all odds. These would act as "comb filters" of a sort. Then, there might be harmonic data for other common waveforms, and whatnot. At the end, the final tone would be achieved by adding the arrays and multiplying the arrays and whatnot. Sure, it sounds like additive synthesis...but the sound you get out of it depends just as much as the algorithm you put into it. Sure, you can get any kind of Phase Distortion sound with additive synthesis, but that doesn't mean PD is useless.
So, I look forward to any input I receive.
infuzion ·
Sounds interesting, why don't you just post a demo?
SM is really good with making quick mockups easy.
Acrobat ·
I agree chironex, nice spot on the situation, but I learnt (indeed programming with SM, before it was not so clear to me what I'm about to say) that practice is much more difficult than theory when it comes to synthesis whether it's traditional or innovative method. To get things working it's not so easy as it seems and requires many months of work.
It's true that you can often end up re-inventing the wheel, but sometimes it's not just a waste of time because you might learn more about, say, subractive synthesis that you never figured so well and that's not bad, really. You can also tune things as you like and you can't do that with pre-made softwares.
As for very innovative softwares for audio, I didn't see anything so impressive apart from new "methods" like MAX/MSP (that actually introduced a new approach to synthesis, anyway very similar to analogue circuitry mmangling) and Metasynth (that uses images but doesn't offer very significant results).
The rest cannot be considered innovative, at least to my eyes, I mean from a wide, musician point of view hardware instruments are still far ahead of software.
Harmonic content morphing is basically a smart use of tables.
I can give you many examples to support my opinion but just 2:
1. In the eigthies (20 years ago!) you had, in an extremely simple and affordable machine like TX81Z, a much more complex implementation of Phase Modulation than you have now in FM8.
2. In the 90s the market offered things like Z-plane 3D flitering (EMU Morpheus, still uncovered in software) or various vector synthesis incarnations that worked well and neat, now there's just Korg reinventing the WS...
Music is moving from concerts and live to home studios and the market place is wider too, everything so accellarated that's very a few chances you find professional stuff, that there is still, but it's harder to find and recognize.
Most VSTplugins are also completely unware of usability in live situation and that reflects also in the synthesis situation when most of them are difficult to program and uninspiring, that's probably why a lot people are still searching for emulations.
🙂
stefan ·
As far as I can tell, HCM is basically wavetable synthesis where the user is presented with ways to actually manipulate the wavetable, but I'm not entirely sure. That's what it seems like to me though 🙂
chironex ·
Whatever it is, though, it offers an easy approach to additive synthesis...which is helpful.
stefan ·
chironex wrote: Whatever it is, though, it offers an easy approach to additive synthesis...which is helpful.
Like I said, I don't think it's additive synthesis 🙂
chironex ·
Hmmm...I always thought it just had the Fourier harmonies for all sorts of instruments, which you could then alter using several transformations...
I guess the world'll never know X)
stefan ·
chironex wrote: Hmmm...I always thought it just had the Fourier harmonies for all sorts of instruments, which you could then alter using several transformations...
I guess the world'll never know X)
That might be the case but that'd most likely just be a mean to generate wavetables. When using the wavetable create in SM fourier series are used too, it's not really related to additive synthesis though 🙂
chironex ·
Yeah, I get what you mean. Whenever I hear "Harmonics" or "Fourier", my brain immediately jumps to "Additive". 🙂
Wrote something here? To have a post removed or your name hidden, write to [email protected].
SynthMaker