Reproducing x0x sequences using CV (x0xi0)

[crx091081gb]

I've noticed that sequences I produce using the sokkos sequencer sound different (better, more x0xy) than when I try and reproduce them with the gate/cv/slide/accent inputs on the x0xi0. I'm guessing this is down to small differences in the timing of the control voltages. I'm setting accent and slide voltages at the same time as the gate and pitch voltages and holding them until I trigger a new note.

What's the conventional timing scheme for control voltages coming out of the internal sequencer?

Cheers

[guest]

i dont have my notes here
but i know there is info on timing
up on robin whittles site

http://www.firstpr.com.au/rwi/dfish/303-slide.html

i dont think its complete though

[antto]

erm, these docs are kinda wrong in some aspects (specifically the stuff about the clock high-low ticks and note duration)

basically, the dinsync clock signal is a square wave, 24ppqn
this means 24 cycles of the square per one beat
each cycle has 2 distinctive moments, the lo-hi transition and the hi-lo transition
aka this is equivalent to 48ppqn MIDI TICKS

a normal pattern puts 4 steps into one beat
a triplet pattern puts 3 steps into one beat

thus, for a normal pattern each step is 6 pulses long (triplet - 8 )
inside one step, the sequencer puts a note, and a silent "gap" after the note
so the duration of the note is 3 pulses and the gap is also 3 pulses

a tied note is extended onto the following step, thus the duration is 3+3+3 pulses (3 for the normal note + 3 for the gap + 3 for the next step's note) => thus it takes 2 steps = 9 pulses of "Gate ON" and 3 pulses of gap "Gate OFF"

if the note has a slide the duration is the same as in a tied note
the only difference is that the SLIDE control voltage is now used, and the pitch CV changes on step 2 (it goes thru a RC Lowpass filter (cutoff = 7.23Hz) which smooths it to make the distinctive 303 slide
i don't remember exactly how the SLIDE CV changes but it should be either on step1 after the 3rd pulse or in the beginning of step2 (but then it would have to be changed before the pitch changes)

i can check my notes if you want more info

EDIT: btw on a note which has slide, the SLIDE CV doesn't immediately go low whenthe Gate goes Off
it should go off at the end of the step
because Gate OFF doesn't mean the note immediately shuts down - there is usually a small tail, and there is also another tail when accent is used
so if the SLIDE CV goes low at the note end - during the gap part (where there is a tail) it will be audiable how the pitch immediately changes, which is wrong, it should continue to sweep

[crx091081gb]

Cheers this has given me a lot to think about. I need to sit down and program some test patterns and compare the CV results with the internal sequencer.

[crx091081gb]

Using that info I was able to pretty much exactly reproduce the sequences I had on the x0xb0x's sequencer using the CV inputs. The only real trick to it was to half the length of the midi notes if they were untied. Saved me a whole heap of hassle, I thought I was going to have to get all technical in MaxMSP. Cheers for the notes and pointers guys.

[crx091081gb]

Yeah I can't make the slides and accents sound right when I use the CV inputs. Could you go into any more detail about when the accent and slide CVs should rise and fall in relation to the steps?

Also can you or anyone else tell me where are good patch points to monitor the CVs in the circuit, I'm trying to record how the stock sequencer behaves without CV input.

[guest]

slide = r59
gate = r146
accent = r141

[antto]

so this is a recording from a 303 (the waveform in blue)
the pattern used is also written
i've painted some stuff around the waveform to show the notes, gates and accent
on top, there are Blue "boxes" these represent the GATE_ON portions of the audio clip
they are 3 clock pulses long each
between them there are also 3 clock pulses of "gap"
on some notes, there is a dark-Blue filling - these are "slid" notes
this dark-blue thing is again GATE_ON
above that, there is a yellow trace, where i've approximated the Accent CV based on how the waveform looks

with green text i've written the pattern (notes and Time data)
"G" is "Gate"
"o" is "tie" (remember, it's on a 303, not a x0x)
"-" is "Rest"

so from this, you can see that:
- Accented note slide to normal note - the ACC CV is up for a duration of one whole step - the accented step
- Accented Tied note - the ACC CV is up but note a whole 2 steps, it's 1.5 steps instead (here it is equivalent to the GATE CV)
if that step was double Tied (G o o) the ACC CV duration was going to be 2.5 steps.. etc..
- Accented note - the ACC CV is up for the duration of the whole step (same as accented-slide-to-normal)

this a bit weird looking, i know ;]

[crx091081gb]

Cheers guys, you're both really helpful. I've got slides much closer now. They seem to end a little sharp but I think that might a small difference in internal vs supplied pitch CVs.

EDIT just re read what you said about the pitch going up on step two of a slid note, that tallies with my experiments. Thought I was going crazy. So much to learn about the sound of the silver box.