Open303 - open source 303 emulation project - collaborators wanted

[mystran]

thanks guys

btw, i just found out that we've all wasted our time, and there is already "the most accurate software emulation"
http://www.pulsecodeinc.com/DB-303/
they have "scanned" the knobs of a real TB-303 into a 3D modeling program!
just kidding ;]

If you want 303 emu for iPhone, I suggest you at least look at audiorealism technoBox. I don't have an iPhone so can't try it, but at least the videos sound more or less like what it's trying to emulate.

That said, I just gave Phoscyon a test drive (and I don't even remember testing it before, though I could be wrong) and while the interface (especially, but not only, the sequencer) is simply awful to use (and it's easy to get outside the 303 range of sounds if you want), I must say I like the clean sound of it more than I like ABL2 (which sounds too clean to my taste; probably Mike's 303(s?) just happens to have a cleaner sound; component tolerances or something).

[Muon Software Ltd]

I was thinking Muon's Tao. Which is nice, but doesn't seem to nail everything. No offense to Dave & Muon, as that was a very early attempt, and processors were much slower then. I really like Muon and kinda feel sorry for them with all the heat directed at them.

Tau Bassline MkII was conceived as an upgrade of the original Tau Bassline MkI, which was basically free and more a fun little synth than an emulation. We also gave a feature-limited version of it to Computer Music magazine (CM303) so it was hardly what you might call a critical project for the company. As ever, our main work is contract software development and we do that very well for a wide range of clients. The vociferous reaction to Tau Bassline MkII on KVR simply convinced me never to try to do a detailed analogue emulation ever again. My time is better spent elsewhere than trying to please the kind of weenies who have endless opinions on what a 303 should sound like but don't actually have one in front of then. You'd think the 303 was powered by magic acid pixies rather than a simple box of (not all that hi-grade) components.

Just as the quoted poster said, Tau Bassline MkII was developed quite some years ago now when the only proper competition was Rebirth. I'm happy that Tau Bassline MkII wipes the floor with Rebirth and have really paid little attention to efforts from ABL or D16 except to "tut" occasionally at their obvious blatant astroturfing every time the topic ever comes up.

Anyway, I can confirm that we did indeed take apart a 303 and subject it to all kinds of invasive analysis, and in my opinion this is the way it needs to be done. I still have the 303, and it still has our modifications. And no, you can't have it

My 2 cents on your process - you can indeed look at waveforms and samples all you want, but how can you be sure the square shape that you are so keenly trying to emulate is not just the result of phase shifts within the filter and the oscillator? and can you tell what effect this has or doesn't have on the sound? if you're only ever looking at the filtered output you are not really in a position to judge. Either way, don't get me wrong, I respect what you are doing and I look forward to the final open-source effort (though I'm confused by the number of people in this thread the seem to be working on seperate projects and not really much helping Robin it seems).

Kind regards
Dave
Muon Software Ltd
www.muon-software.com

[rola]

Anyway, I can confirm that we did indeed take apart a 303 and subject it to all kinds of invasive analysis, and in my opinion this is the way it needs to be done. I still have the 303, and it still has our modifications. And no, you can't have it

Do you have any samples of the waveforms that were taken before the filter?

rola

[Muon Software Ltd]

Do you have any samples of the waveforms that were taken before the filter?

Yes I do, and I also have the all-important samples of test tones going through the filter as well. However, you must appreciate that whilst I applaud what the open-source project is doing, I have no wish to contribute at this time.

Kind regards
Dave

[rola]

Sure. I'm not contributing myself as I don't have any experience with 303s. But I'm watching this project with great interest. The samples antto has posted sound very promising to my ears. Though it would probably help if some proper measurements could be made on the real thing.

BTW I believe you can get pretty close by examining the output if you know what kind of filter you need etc. you can examinine the slide times for example by just looking at the waveform. the pictures antto has posted already look very close. you can probably figure out where shapers and highpass filters need to be placed by experimenting and comparing the waveforms. it's just a matter of how much effort you're willing to put into this.

[Muon Software Ltd]

Well I'm certain he's doing great stuff and his motivations are honourable. To me at least however he seems to be working on his own project, and not (directly at least) helping Robin. I'm only really interested in supporting the open-source aspect of this project. If there was a genuinely accurate, well-documented, cross-platform, open-source 303 emulation freely available to everyone to improve, extend and enjoy then that would be an indisputably good thing for all (except maybe ABL and D16 ).

Maybe, in time, I might be persuaded to contribute my knowledge and my 303 to such a project, but I'm not interested in facilitating yet more closed-source emulations as there are too many of those already rehashing the same ground.

Kind regards
Dave

[coops2]

Well I'm certain he's doing great stuff and his motivations are honourable. To me at least however he seems to be working on his own project, and not (directly at least) helping Robin. I'm only really interested in supporting the open-source aspect of this project. If there was a genuinely accurate, well-documented, cross-platform, open-source 303 emulation freely available to everyone to improve, extend and enjoy then that would be an indisputably good thing for all (except maybe ABL and D16 ).

This is a very good point I have been following this thread and would love to see an open source emulation eventually released but at the moment the project's goals seem confused to someone such as me watching from the sidelines.
-which is the official project? A number of contributors to the thread seem to be working on different projects
-Where is the project being hosted? At the moment it seems to be posted by Robin as a link but what if someone modifies the code etc what do they do with it?
-If someone wants to get involved where can they get the latest source?
-Has a framework been decided on? The posted project seems to be standard VSTGUI but there was a demo with a gui posted a while back which looked like JUCE?
-What are the specific goals etc that the project is aiming for?

I think this project is a great idea and there are lots of intelligent people at KVR who would potentially het invovled so how about making it all more organised/official and hosting it somewhere, such as github? At the moment, if Robin's freetime runs out the whole project could grind to a halt.

[rola]

I don't think it really matters that antto is working on a different project. I think he started his project before the open source project and has been sharing the details of his implementantion and the findings of his research on this board. Discussing different approaches will ultimately lead to a better open source project and everyone will benefit.

[Muon Software Ltd]

I don't think it really matters that antto is working on a different project. I think he started his project before the open source project and has been sharing the details of his implementantion and the findings of his research on this board.

Sure, but the discussion a) seems a bit one-sided and b) nothing seems to have flowed back into the actual source code of the project as far as I can tell.

Perhaps, over time, the project will organise itself onto sourceforge for the code, start a wiki for its documents, and open up its library of samples - and then it will maybe start to attract some quality, concrete, contributions. Otherwise the thread is barely worth following unless you want to follow what Antto is up to - and surely he could start a blog for that.

Kind regards
Dave

[antto]

you mean i talk too much, don't you?

Muon: i'm not helping here? i don't think so i could (like you) keep all my findings to me, but why?

this thread (and the open-source 303 emulation project (AciDevil) was started by Robin, but i believe it was fueled much or less be the previous thread here, about the 4pole filter (which i started before that, which i started after the releasing of my synth (beta1))
Robin (and the other guys) helped me with the filter, it got clear that i wanted a filter for a 303-like synth, and the whole discussion slowly mutated, that's where Robin started this thread here (which is wonderful, we were all going offtopic talking about the TB-303 there)

Robin himself is kinda quiet the past few days, either hes working hard, or maybe offline/having a brake, not sure
this is mainly his project here, you can find the source code in his posts...
if you modify it, you could always tell us here, Robin will do his thing
currently i think no one has "modified" the source except Robin, that's kinda sad, i don't know about the others, but i am a pretty "messy" coder, and i already got my own project started you know
anyway, i share all i know, and anyone can ask me if he doesn't understand what i mean (my poor english) or disagree with what i say (and correct me)

Muon: i see what you mean about looking at the output waveforms that were filtered by a bunch of filters, what if i have the same filters in my synth matched closely? won't the waveform look the same?
blah, i don't want to convince anyone that i am doing the best thing, no, but _i am_ getting close results some times

you know how the waveform looks before being processed by the filter? but you won't share that? ok, why did you mentioned it, blah..

i think Robin has a lot stuff to code now (he is doing a VST, while i was doing a SE-VST, which is simpler)
.. and probably no one else has the time to do the analyzis that i do right now
i might be going in wrong directions, but you could always come here and tell me i'm wrong (but give a good argument)

i have corrected myself a bunch of times already, i'm trying my best not to write bullcrap "facts" that will lead the others to nowhere

i'm not doing an opensource synth? yeah, but it is freeware
and what if i publish my SE-project (with the GUI stuff) and the source code and everything?
what? you think someone will use it? it's a complete mess
what Robin is doing looks like propper C++ to me, he has all the classes and structures, methods, everything OOP, and all the .cpp and .h files
that would be perfect for others to look at, or modify
..i am a very bad C++ coder, really!

i hope you understand me..
ok, said enough
any contribution is welcome

[Muon Software Ltd]

you mean i talk too much, don't you?

Not just that, but its clear that you are effectively talking to yourself

Muon: i'm not helping here? i don't think so i could (like you) keep all my findings to me, but why?

An open-source project survives on contributions to the *code*. If you have other contributions to make, such as analysis data that other people could use to write code, then for the project to have the best chances of survival I'd say put it in a wiki or at least document what you are doing so others can follow and participate.

anyway, i share all i know, and anyone can ask me if he doesn't understand what i mean (my poor english) or disagree with what i say (and correct me)

I'm sure Robin appreciates what you are doing. But I think it'd be a greater help to properly document and describe what you are analysing, and how best you think to synthesise it. Then the project can start to take shape.

Muon: i see what you mean about looking at the output waveforms that were filtered by a bunch of filters, what if i have the same filters in my synth matched closely? won't the waveform look the same?
blah, i don't want to convince anyone that i am doing the best thing, no, but _i am_ getting close results some times

you know how the waveform looks before being processed by the filter? but you won't share that? ok, why did you mentioned it, blah..

I'd say that you don't have much chance of matching the filter's frequency and phase response. The problem is that in a sample, the filter is always opening/closing - and quite often quickly, over a wide range.

If you take an FFT of the response of a closing filter, in order to get the resolution you need (say 65536 samples for a detailed FFT), unless you're sampling at some indecent rate you will be looking at a slice in time of a changing filter over a second or more. That isn't an LTI system and its very hard to model. Besides, to draw any valid conclusions about the filter's response also requires knowledge of its input.

In short, I don't think you have much chance of getting (especially the phase responses) exact. You might get it good enough, but not exact. Certainly not exact enough to derive any conclusions about the shape of the waveform. Besides, you have to consider that the human ear isn't even very sensible to phase differences - so maybe better to concentrate on what it sounds like, rather than what it looks like.

As for "why mention it" - well, I'm prepared to contribute to a good quality open source project - if such a thing exists. At the moment, it doesn't. Personally, I don't want to see another closed-source 303 emulation - I'd rather see something built from a documented, credible process with the analysis data in the public domain.

i'm not doing an opensource synth? yeah, but it is freeware
and what if i publish my SE-project (with the GUI stuff) and the source code and everything?
what? you think someone will use it? it's a complete mess
what Robin is doing looks like propper C++ to me, he has all the classes and structures, methods, everything OOP, and all the .cpp and .h files
that would be perfect for others to look at, or modify
..i am a very bad C++ coder, really!

Well, I would say that for a young guy like yourself, with time on his hands and the evident drive that you have, there is a win/win situation waiting for you. Tidy up your code and integrate with Robin's sources. In the process you will learn alot, and the project will move forward. If you continue to work in your own little world you won't learn as much as you would from working with the kind of people this project might attract if it had legs.

Making another closed-source 303 emulation just isn't interesting enough for me to contribute, and I do, indeed, have a lot to give - but to the right people, for the right reasons and in the right environment. Otherwise its just wasted time, I might as well just write a book.

Kind regards
Dave

[mistertoast]

"An open-source project survives on contributions to the *code*."

That's true, but there's often a visionary who cares enough about it to get it to alpha first. This project is still young.

[rlahalla]

That's true, but there's often a visionary who cares enough about it to get it to alpha first. This project is still young.

I can make bleeping sounds with my mouth that very much resemble a dying 303

[Muon Software Ltd]

That's true, but there's often a visionary who cares enough about it to get it to alpha first. This project is still young.

Well all the more reason to try and stimulate some organisation (a necessary evil) and some actual concrete contributions (the original aim of Robin's post as evidenced by the title). I don't think a scrappy 30 page forum thread of ramblings and musings is the kind of searchable, organised resource a project such as this needs.

I've been reading "We-Think" by Charles Leadbeater recently and I can see his point about how you can get a motivated bunch of individuals to self-organise and work together if you have some interesting nucleus of a problem for them to accrete around. He cites open source software development as a particular example of this amongst other things.

Now if Robin and Antto *really* started to work together, you have something that might just kick off and might just get interesting. If they both work in seperate silos, then it's going to be more like "nothing to see here, move along".

Kind regards
Dave

[antto]

ok, i get your point, i'll shut up untill Robin shows up
i did write a hole pile of posts here, and i'm just a "guest"
ok..

btw, i think disagree with some of the things you said

the filter "opening/closing" you mean that on the samples i got, it's modulated by the envelopes and because of that, i can't tell what is it's frequency/phase response?
hm.. i hope you didn't mean that

As for "why mention it" - well, I'm prepared to contribute to a good quality open source project - if such a thing exists. At the moment, it doesn't.

if everyone thinks like you, there will never be one single person to start such a project
as you said, this here has the spark to become one, but you think it's not yet there, maybe if someone with better knowlage like you should start contributing
or you want to wait till it's almost there, to come along, say what you have to say (which would probably be known by then) and walk away like a hero or somethin..

or you just don't want to share the "secrets" with people like me (closed-source) huh..
whatever!

Now if Robin and Antto *really* started to work together, you have something that might just kick off and might just get interesting. If they both work in seperate silos, then it's going to be more like "nothing to see here, move along".

you're probably right here
but i really don't have much spare time, i've already sacrificed my personal life, during the past few months, all i do is this
also, me and Robin kinda seek a little different aspects of the TB-303
tho, we could build some of the "vital" sections of the synth together
i still want to continue my own synth, i've promised to myself, and a hole lot of guys now, i can't just leave it..

An open-source project survives on contributions to the *code*.

yeah, so i haven't contributed at all, and i probably won't if i keep doing what i'm doing
so, that's bad? ok

i've never worked with another person together on a project (except for my real job, which is kinda different) and will probably fail soon if i try

Robin helped me in the begining, with the filter (i'm not forgetting the other guys there!) and i owe him a lot, i would do whatever i can to help him, even that when i listen to guys like you, it seems i can't?

you killed my motivation for today, damn, gonna drink my coffee

EDIT: just to give you a clue of how my weeks go:
monday to friday: wake up at 06:30, work from 08:00 to 17:00, home at 18:00, eat, check forums, 303 stuff, bed time 01:00
saturday: the same, but work is from 08:00 to 13:00
sunday: off
i'm not complaining..
but i just don't have the time for other projects, doing blogs, writing books or whatever