built-in modules - reverse engineered

i spend some time analyzing impulse responses of reaktors built-in filter modules in matlab. i'd like to know the exact behaviour of all built-in modules to rebuild reaktor ensembles in other modular software synths (synthmaker and so on). this should result in open source modules (core cells, synthmaker modules, c code, pseudocode, whatever) that have the IDENTICAL effect to the very last bit.
that means i don't want any, say "multi/lp 4 pole" filter, but exactly the same one, so a structure using it will sound exactly the same as in reaktor.

i just started, then thought i should ask at this place if somebody did already work on that before.

 

i dont mean to sound grumpy or anything, but why? i mean, why not just use reaktor? im curious, not trying to discourage or anything.

i have been using max/msp alot more lately, and i think i like it and reaktor both because they have such differant sounds and methods of working, and therefore can really compliment each other (i gotta try using the max VST~ obect and see how thqat works out - using Max as a host for Reaktor just sounds too good to be true, so long as Reaktor plugin accepts OSC, i think ill be soft-synth heaven!_)

 

why? i got a bunch of answers to this.

but first let me state that i still prefer reaktor from any other digital synth, the only thing i would criticize is NIs policy of what i see as product disintegration. i think every single of their non reaktor products could and should have a modular representation in reaktor. perhaps not in the standard version because it would be a real monster (increased hardware and dollarware requirements), but something like a 'pro' version.

now my answer: if other modular synth could reproduce reaktor ensembles ...
1. their additional features (as synthmaker's spectral processing or vst-export, max/msp/jitter's visualization techniques, distributing polyphony over multiple machines using that bunch of code i'm currently working at) could enrich these ensembles
2. NI might get a bit more motivated to implement new features into reaktor
3. you might run your ensemble on a less pricy environment
4. we might see our favourite synths on new platforms (linux, game consoles, even low polyphony / low sample rate versions on PDAs and mobile phones)
5. somebody might come with a new, perhaps better GUI approach for editing the structures

 

no disrepect or anything, but do you think you could "possibly" change your screen name. It is really in poor taste.

 

really that bad?
that name came from making fun of english speakers calling anything ****ing ...
it made programming a bit more amusing calling the single modules like ****ing oscillator, ****ing filter and so on.
i guess it depends on your native language how that f word sounds to you, i think it's so overused that its not a strong word at all and it's rather ironic. i hope i didn't tread on someone's toes too much.
but of course i could 'possibly' change that screen name if more people resent it.

 

i personally dont mind it, but i was also thinking before my fellow moderator-in-crime said anyhting that it may not be a good idea, as it can be a very harsh word for some people.

i see your point, i think it may be very hard to come up with an efficient version of all the reaktor modules in even one other platform, and very time-consuming, but it is always worth a shot. i dont use synthmaker (i did try the beta/demo), i am using max/msp now, but im still in the learning process there, and i like using reaktor as a VST inside of max anyways.

still, as i said it isnt a bad idea, but i dont know how many people will be wiling to do alot of work for this, although if you pick the platform and find other users on it, perhaps you can get people to work on jsut one or 2 modules each. i think alot of the event stuff shouldnt be too hard, i think the real hard bit is the filters and samplers/oscilaltors, which do have their own distinct sounds, especially the grain samplers. i tried making a core version of the resynth and beatloop, and although the fuinctionality isnt too hard, getting the sound identical is. then again, i know jack about DSP and software designjk, i am just going on my experience using samplers and synths, and my experience using reaktor over the past few years every day (even more since R5/core came out_).

 

how precise is reaktor's sample rate?

as i am observing reaktor's behaviour in detail i see something that doesn't make that much sense to me:

i observe results that i would interpret as inaccurateness in sample rate. for example when i'm recording the output of a clock osc at F=0.5 Hz, W=0, i get 44122 or 44123 (!) Values of 1.0 followed by 44123 Values of 0.0 and so on. reaktor's and my audio device's sample rate both are set to 44100 Hz, master tuning is set to 0.
same result whether i use the recorderbox or i write into an audio table.
my audio device is a m-audio delta via asio at 6 ms latency.

any idea why?