Open303 - open source 303 emulation project - collaborators wanted

[rv0]

how slow/fast can the tempo be with the external sync?

dont have a clue..
lets just say, very slow, and 300BPM to simplify?

how many 303s are over there?

2 in my studio

and IIRC you have always given me samples from one them, right?

no, but whenever i took samples from both i mentioned it, either in this topic or with a txt i provided


Anyway, I'll record your patterns tonight.

[antto]

thanks ;]
where would we be without your assistance ;]

[rv0]

Yesterday computer monitor in studio didn't work..

So I just started now,

this is the slides:
slide_time.zip

its stereo, 44Khz. L is tb one, R is tb two


More after dinner!

[rv0]

here's the envelope tests:

env_test.zip

88.2 khz, same "stereo" setup as before

the other stuff will have to wait cos this filled up my webspace..

[rv0]

ok, random filedump site, sorry for the ads,

this is the accent test (3 x 88.2kHz stereo wave files in archive):
http://www.filebox.com/9jeapou8vhlo

[antto]

got the first two, i'm downloading the third one now..
pretty cool way you sync'ed them together! i can see the diffs right away ;]

EDIT: got em all now

wow, huge difference between the peak cutoff frequency, when CUTOFF and ENVMOD are 100%
the one on the left is at 27.5KHz (this is the one i have analyzed before too)
while the other is at 23.5KHz
acc_test2: highest cutoff peak of the one on the left is 36.4KHz
amazing ;]

[antto]

report:
minimum decay time = 0.385 (was 0.341)
the last rc-decay shaper is pretty accurate

i'll make another testcase to checkout the VCA response (Gate attack/release times, Accent "tail" and Accent switch-off during a note)

[rv0]

hehe funny
i was talking with the dude who owns 6 303's earlier this week, about my plans to take samples from all of them. he was kinda amused because he knows very well all 6 of them sound different

also about the accent-gimmick circuit (the one where accent amount increases in longer rows of accents), he told me that the devilfish allows you to control that very intuitively.

I can't remember: have you tested for any accent increase with successive notes?


And yes, the way I recorded them now should be my best samples I did so far: sample-accurate tempo, totally synced up, recorded straight into the motu. Only thing I had to do was lower the internal volume in the accent tests (because otherwise it would clip). As they were recorded simultaneously, so were all parameter changes (thank god for giving us 2 hands)
I also tuned them the same when you asked for center position (tune in C)
keep in mind that for other parameters, the center position isn't accurate.

[antto]

.. i was thinking of something, but i lost it
this happens every time you mention that guy with the 6 303s ;]

yes, the gimmick is a freakin RC circuit, and it's all about figuring how many are there, how are they wired up, and the coefficients..
that's the purpose of the accent001 test at different tempos
1. i make a model of the envelopes (with the gimmick and everything)
2. then i plot a sonogram of a specific section of your samples
3. i map the noteON/OFFs and ACC signals according to the sonogram plot
4. i run these signals thru my model and plot the result (which is cutoff frequency in Hz) right over the sonogram
i've shown such pictures before
the thing is, i actually plot 2 versions of the same section at different tempos (slow and fast)
since, you can fairly easy approximate the curve by runing the envelope thru a LPF, but when you compare to the same thing at different tempo - the approximitation fails..
it's more than just a LPF ;]

another clue about why different clones don't ever sound the same, because the character of the TB-303 varies between units too much, it ain't no digital synth where the sawtooth is all 11010011000101001 ;]

another thing, about your recording device
it's pretty brilliant, it seems like it almost doesn't have a DC blocker
i mean, the waveforms look so good
when you switch to Sawtooth, you can trace the amplification curve of the VCA by looking at the upper "flat" line of the sawtooth, it goes away from zero on accents and it "draws" the curve of the vca-gimmick ;]
you can't really observe this in other samples recorded with a strong DC killer
i never found other samples with such great quality
and especialy the ~36KHz filter peak, i would never knew
i'm pretty sure you have no mices near the house?

the tuning of your first 303 (the one on the left channel, that i love the most) hasn't changed since the first time i measured, it is 1.027 (scale)
but that gives it a certain character

btw, i've rewritten all my data structure formats now, my sequencer is in a state where it almost doesn't work ;]

[rv0]

yeah, soundcard is DC coupled
which would mean there's no coupling capacitor in it, afaik.

so you can send +5 volts signals with it. (which is what I use to sync the 303, i send clock thru the interface's headphone out)

[antto]

it doesn't have a DC blocker?
you can input a 5V DC and it won't fall to zero on the recording?
this means that the last HP filter on the recording is from the TB-303 itself, pretty cool ;]

[audiojunkie]

I really admire all of the work you guys are doing. I watch (lurk) with continued anticipation. I have very little time, and less talent (I think), but am really excited to see how this thing is taking shape. You guys are all doing wonderful work! Thank you!

--Sean

[antto]

oops, i kinda deleted 90% of my sequencer code ;]
the big sh!t is coming.. wish me luck..
oh, and the GUI and everything..
i'll be flat-lined for the next few weeks (months?)

[rv0]

oops, i kinda deleted 90% of my sequencer code ;]
the big sh!t is coming.. wish me luck..

good luck!

I'm curious to see what you will come up with.

i hope you have backups of everything!!!


btw, to confuse you even more, here's another friend of mine:


dont worry, they're not all his.

[antto]

erm, you got 2 - okay, but.. the other guy has 6.. this guy.. 9..
isn't it too much? i mean.. you know.. "If you have 9 TB-303s, give one to the poor."
seriously, i'm starting to think you have TB-303s growing on a tree somewhere??

50% of the excitement is that i don't have backups really
well, the style of the GUI will look more 303ish.. but it'll have a 20x4 green LCD screen somewhere in the middle, with a piano-roll view ;] and lots of buttons again
thanks, unfortunately, today was the last day of my "vacation" i'm back to my daytime job tomorrow, and progress falls to "positive denormal number" again