From map@cs.washington.eduTue Apr 25 16:08:33 1995 Date: Tue, 25 Apr 1995 16:00:04 -0700 (PDT) From: Mike Perkowitz To: analogue heaven Subject: Re: oh no not again > You know, I asked these questions on r.m.m.s last year and got all kinds > of why the hell would you wanna do that, the 303 seq is so boring why not > duplicate the arp seq instead. My understanding is that the 303 seq is a > 16 step pattern sequencer where each step can be set to normal, silence, > accent or slide. Is this correct? If you set two consective steps to > slide, ie 14 and 15, does it take two step times to slide accross these > two notes? Where chromatically does the slide start from and finish to, > does it go from whatever is programmed in step 13 to whatever is > programmed in step 16, or do the notes programmed in steps 14 and 15 > effect the slide. the 303 is a 16-step sequencer. you can have a sequence play fewer than 16 notes and change their timing, but the default is 16 16th notes. in step-recording, sequences are entered in two steps. first you enter in all your notes, pushing the appropriate buttons on the little one-octave pushbutton keyboard. you can also push octave up/down buttons as you do this. once the notes are entered in, you can step back through them and play with the modifiers. there are four of these: octave up, octave down, accent, and slide. the octave keys do the obvious thing. accent increases the level of the note (not transmitted via cv/gate btw) as well as doing some screwy things with the envelope which also effects the filter. for example, if you have accents in your pattern, you can hear a difference, even with the accent knob turned all the way down. slide means that portamento will be performed between this note (the one marked slide) and the next one. this note, then, will play, but the next one wont retrigger the envelope, and the pitch will slide from this one to the next. the other step is entering in timing. you step through your 16 steps, at each one entering either play note, tie, or rest. note will just play the next note that you entered in above. tie holds the previous note, and rest plays nothing. pitches and timing can be done in either order and redone separately. the whole thing is so weird and such a pain in the ass, it's usually easiest to just punch in a bunch of notes, then step through some semi-random timings, and then go back through the notes and diddle with octaves and accent and slide. the results: random noise. you just have to iterate this process until suddenly the patch sounds right. not that this is the only way to compose. i've definitely put in specific things and worked at them until they were right (the buttons often think you hit it multiple times when you didnt mean to), but lately i take a much more zen approach to it. :) i would love to write a midi/cv sequencer with these capabilities. in the midi realm, you'd probably have to settle for velocity instead of accent. with slide, you could do one or both of two things. first, there is a portamento on/off midi controller. your midi303 could turn this controller off and on for a slide effect. also, many synths have a legato mode -- when you play legato, portamento happens, but otherwise you get normal play. your midi303 sequencer could, when there's slide, hit the note on of the next note before the note off of the current, triggering legato, or something. i have no idea how well these things would work. it would vary from synth to synth, and you'd have to twiddle a bit with both the midi303 and the sound source to get the cool things happening (not that it's necessarily bad if something un-303-ish happens, as long as the results are cool and wonky). if i were building a sequencer with 303-like features, i think i'd want to make it both programmable in the traditional 303 style (that is, separate note and timing) as well as in a more conventional step style, where they're done together. in fact, there's a nanophonix design for an SQ-303 i've been working on with precisely these features. too bad nanophonix designs are entirely fantasy. (anyone want to build some things to my functional specifications? i've got a *lot* of designs, and imho they're pretty cool :) m ----------------------------------------------------------------------- "With repeated accents, a standard Bassline's filter goes up at the first accent, and gets a bit higher with subsequent ones. This gives an emotional response like something is being poked or hit and it just can't cope -- its cries gets higher and higher." -- Robin Whittle ----------------------------------------------------------------------- map@cs.washington.edu http://www.cs.washington.edu/homes/map/ mike perkowitz http://www.hyperreal.com/~tint