pouët.net

i'm not a musician and have only limited synth knowledge, but this one sounds like one of those typical dubstep sounds to me. and yeah, what i've learned is, that many dubstep producers use a random super short snippet of one track and play this in a loop (e.g. a 1/20s sample) while applying tons of effects and filters with LFOs - so you could basically create one random sample (but not noise) and loop this as your base oscillator. dunno about the sounding though, but if you have your synth framework open, it might be worth the try.

DUBSTEP!!!

That's true. However, our synth has no sample-support, so the retrig-like effect is achieved assigning af saw-down LFO synced to 1/32 or 1/64 beats of the tune to the gain of the oscillators. That part sounds close enough. But I agree, that whenever Gloom replies, it'll probably be the recipe.

But then I'll just have to keep hunting it down until it gets right. There's unervingly long "distance" between the low- and highfrequent parts of the sound and the filterenvelope/LFO seems to affect each "part" of it differently.

I need to go on vacation in New York more often - all kinds of interesting threads start popping up when I do. :)

Quick and dirty answer: it's very simple OSC modulation. Here is an MP3 of the sweep isolated from the rest of the track: http://cl.ly/4JTD (no EQ, straight from the single channel).

Here you can see the setup as done in Massive:



With regards to how I did it, I started with a plain FM template and tried different OSC waveforms until I found one I liked and then added a OSC modulation to it. To reproduce the sweep, simply pick the clean "Carbon" OSC waveform and add the OSC modulation afterwards.

Great ;-) Then Torus was right after all! Pure phase-modulation of an oscillator.

I'll give this a shot when I get home tonight.

Just two additional questions:
1 - The little, white number below the pitch-setting, is that Hz?
2 - What the heck is a "Carbon" waveform?!

Anyways - thanks for lifting the veil on this one, Gloom.

1) It's just a leftover from an automation experiment and can be safely ignored. :) Basically, it let's you set the range of the modulation over time, but I don't actually use it in that sweep.

2) "Carbon" is just one of the many wavetable oscillator templates in NI Massive (along with saw, square etc. except more complex). You would need to take a look at what it looks like clean to understand how it builds the sound from there. Here is a clean WAV of it: http://cl.ly/4IWk

FUUUU MAN. I nearly told you the same except that I involved sync additive methods and pulse width but ITS THE FUCKING SAME :(

(trying my best)

I also hear a vowel filter in there, for the aouie character of the sound.

This extract puzzle me since a while.

http://hotfile.com/dl/100807315/e182af8/BYPASS_UNIT_Distorted_Frames_extract.mp3.html

What filter or synth could possibly do that? The producer is
This sounds like... undoable in software or at least not .

The producer are bypass unit*
at least not with all synths*

Gloom: Thanks ;-) I've got wavetable oscillator as well - I'll try something with a sawtooth with a sine representing harmonics and a lowcut somewhere.

#ponce: I'm not completely sure, which of the sounds you mean, but I gues it's the rapidly wably high-pitched sound.

One way to produce such a sound would be to have a simple saw-oscillator, that has an LFO (probably synced to 1/32 or somewhere in the vicinity) assigned to pitch. Run it through a bandpass-filter with a rather narrow bandwidth and have an LFO with a rather low frequency assigned to filter-cutoff.

Also, if the synth of your choise supports velocity-assigned controllers, try assigning velocity to the frequency of the LFO that is assigned to oscillator pitch.

But either way, I'd say it's definately doable in a software synth.

As punqtred said - and with quite some resonance om the filter.

There is also a slight (but noticable) phasing going on over time as well, which helps bringing character to the sound.

Punqtured & Gloom: thanks, it must be that indeed (tried in Synth1). The problem is I don't know of a digital filter that would have this type of resonance (no saturation of the feedback, support for very fast modulation without aliasing). I should try with ACE.

#ponce: Most synths have Q/Emphasis on filter. It causes a "peak" in resonance just on or around the cutoff frequency.

Instead of plain cutoff like: ¯¯\__
adding Q/Emphasis would create: ---´\__

Ok, that's a really poor attempt of ascii, but I gues you get the point ;-)

In order du produce the sound you need, a high value of Q/Emphasis would be required along with a rather high resonance-setting.

For rather detailed info on various aspects of filters, This article on filters is a really good place to start and has an emphasis on trying to explain filters using plain english instead of math. Some nice pictures of various frequencyresponses of the filters would have been nice, though.

MAN I'm struggling with that "carbon"-waveform.

The isolated waveform, you sent me looks like this (blue line is the carbon-waveform).


Now it's obviously a triangle wave as indicated by the red circles, but I'm struggling a bit to find out, what the heck is going on at every other harmonic there (indicated by the green circles)

It seems to me as if it's another triangle or a sine but at higher frequency. I've got no problem getting "something" appear on every other harmonic, but I don't have the option to specify frequency of a basic waveform used in the wavetable for the oscillator.

Does anyone have a clue as to how such modulation can be achieved?

A sawtooth is the obvious choise, but I'm afraid it's not that either.

From green circle to green circle is just a single oscillation of the waveform. Looks like a phase/frequency modulated sinewave, possible modulated with another sine. Let me see if I can recreate something like it real quick.

Ah, sorry, I'm off base here, didn't see the variations between different cycles. A sine frequency modulated with a saw?

Yes, I've found out, that adding a saw as first operation to the wavetable, then removing every other harmonic as second and have the triangle waveform as third (thereby apearing on every harmonic) does most of the trick. But from the sound of it (the sample Gloom posted further above) there's some of the high-pitched harmonics, I'm missing. I tried multiplying whitenoice into it, which at some settings comes closer.

But it's definately not a sine. Perhaps it's not that apparent in the above image, but definately in the sound of it ;-)

I've had a look at the original carbon wave too. But this pic looks like a basic sine perturbed by a higher-frequency triangle on a sharp positive sine lfo modulating the "wet mix" of the triangle.

Sure, maybe the lfo is a filtered pulse, which is anyway close to a fake sin.

I'll stick the thread. But I'll "viruz II" it tomorrowr. Sure possible. :)

Mind you, Massives oscillators are not "standard" oscs. You can modulate each with the other (they call it wave scanning), on top of that there are many custom waveforms which i.e. carbon belongs to.

You still wanna get that first sound you posted in V2?

The waveform is from a wavetable oscillator. Not a "standard" oscillator. An example from any synth would be fine as long as it's obvious what's going on for wavetable generation. But does V2 have that at all?

My point being, that the above waveform is not the actual resulting sound I wish to achieve, but simply an attempt to recreate a custom oscillator used to create the sound that this thread was started because of ;-)

The sound I'm trying to recreate is the one Gloom posted in the same post as the clean oscillator output. Hopefully there's no modulation or anything other than one, clean wavetable-based oscillator.

How ironic is it that generating that waveform (though not the following modulation) would be SO EASY in Fast Tracker II? :)

Punqtured: If you look at the waveform screenshot you posted above, the green circles approximately indicate start (and ends) of each cycle in the waveform oscillation (e.g. 440 Hz => 440 waveform cycles per second). There is quite a bit of difference in the look of each cycle, which tells us that there is SOME kind of modulation going on, and that it is not simply a static single-cycle wavetable.

You cannot match the way the waveform evolves using a single-cycle wavetable oscillator. Now, that does not mean that you cannot match the sound in some other way, of course :-)