Analogue-sequencer

colin- I've got the second-latest build (19) running, & finally got the 
studio rewired to test it just now.
here's the thing. well, two things, but this first is the biggie for me.

the 8 tracks are set to midi channels 1 to 8. on these channels, & only 
on these channels, I can't get the p3 to pass controllers unless it's 
in record. 
a nearby maq, wth it's own controller box, is passing though the p3 
without any problems (except the second thing). it gets f'd-to-s & it's 
controllers are passed by the p3. (the second thing is that changing 
the FTS settings causes hanging notes).

but my main problem is that no matter what combination of midi-through 
options I dial up, the controllers for the 8 channels that the p3 is 
actually set to talk to, won't pass through unless it's in record.
it used to, & now it doesn't. it's fallen off.

am I doing summat awry?

d.

> colin- I've got the second-latest build (19) running, & 
> finally got the 
> studio rewired to test it just now.
> here's the thing. well, two things, but this first is the 
> biggie for me.

It seems that some of the changing and optimising of code I've done has
introduced a few oddities.
I've uploaded tonights build, which fixes these.

> (the second thing is that changing 
> the FTS settings causes hanging notes).

That is an inherent problem that's always been there.
Because the soft-thru doesn't track which notes are held, if the FTS
settings change, a note number may be mapped differently under the new
scale, so the note-off for the held note has the wrong number, and the note
is left hanging.
It was to address this (at least for the record channel) that I've been
working on the soft-thru code.
The record channel had more a problem with this after I included PXPos in
the thru notes, but this limitation has always existed, though I hadn't been
aware of it. My MIDI setup is so simple, it can take me a while to spot this
sort of thing.
I think the only solution to this issue will be to track all held notes on
all channels passing thru under FTS...

Best regards,
Colin Fraser
Sequentix Music Systems Ltd
http://www.sequentix.com

>>It seems that some of the changing and optimising of code I've done 
has introduced a few oddities.
I've uploaded tonights build, which fixes these.<<

aye, that's done the trick. also, the midi-clock-only output on this 
machine (#1003) is working as advertised too, ta.
I have quite a complex little set-up here now (this is in the studio, 
though the live rig has a few things in common). & because of this, 
I'm getting some false clock-starts (F8, is it?)

here's what's going on: the clock-only output of the p3 goes to a 
doepfer pocket controller, which does things to the maq & whatever 
it's talking to. this goes to a midi through-box, & then to my 
oldest, most knackered, but up-to-date software-wise maq. 
the maq's output is merged with another output of the through-box & 
with my novation x-station controller keyboard. (this last bit of 
merging happens inside the keyboard). 
the whole lot then goes back into the p3, while everything else 
(repeater, supernova, samplers, proteus boxes, ms6's & so forth) is 
fed from the p3.

what's happening is that the repeater &/or the maq will suddenly 
start up on their own... I have to start & then stop the p3 to stop 
them, & then they do it again a few seconds later.

while the maq & the repeater are having their moments of madness, the 
tempo control on the p3 is locked out, almost as if it's receiving an 
external clock. which it is, I suppose

what I can't figure out is where the spurious F8's are coming from. a 
normal clock-loop would just chase itself- this is running at the 
right speed, it's just not starting & stopping when I want it to.

I think I either need to use the maq as the master clock (ugh...) or 
use something else as the master clock. in the live-rig, this would 
probably be the korg es-1 anyway.

but everything else is fine.... :-)

duncan.

> here's what's going on: the clock-only output of the p3 goes to a 
> doepfer pocket controller, which does things to the maq & whatever 
> it's talking to. this goes to a midi through-box, & then to my 
> oldest, most knackered, but up-to-date software-wise maq. 
> the maq's output is merged with another output of the through-box & 
> with my novation x-station controller keyboard. (this last bit of 
> merging happens inside the keyboard). 
> the whole lot then goes back into the p3, while everything else 
> (repeater, supernova, samplers, proteus boxes, ms6's & so forth) is 
> fed from the p3.

Wow.
I'm not entirely sure my brain didn't just go into a feedback loop...

> what I can't figure out is where the spurious F8's are coming from. a 
> normal clock-loop would just chase itself- this is running at the 
> right speed, it's just not starting & stopping when I want it to.

I had some initial problems with the PIC chip doing the sync to MIDI
conversion in the production P3.
There was some false triggering of the PIC inputs, even though there was no
signal at all to be seen on my scope or logic probe.
I fixed this initially by adding capacitors to the input pins.
It's possible this isn't enough on the chip in your unit.
My later fix was to add soft noise reduction on the PIC.
I'll send you a new PIC to try.

Best regards,
Colin Fraser
Sequentix Music Systems Ltd
http://www.sequentix.com

>>Wow. I'm not entirely sure my brain didn't just go into a feedback 
loop....... I'll send you a new PIC to try.<<

ta. 
it looks more complicated than it really is, when you try to write it 
down. I drew it out in five seconds... there /is/ a loop, but the p3 
seems to be able to ignore the fact that there's clock coming at it 
from outside... most of the time... 
I /think/ the spurious starts are being issued by the p3- it doesn't 
set off itself, but the maq & the repeater both start running when 
they feel like it. the p3's tempo control is locked out when this is 
going on, I've noticed, same as when it's receiving external clock.

the problem with the design of this rig is that the clocks arriving 
at the maq have to be passed round the outside of it aswell, since it 
doesn't have any sort of "through" & I'm using a single small 
controller (the little doepfer pocket control) to alter settings on 
the maq (row length) & also send controllers to what the maq's 
driving.
in the live rig, the p3 isn't the master clock, so this isn't an 
issue.

I've dealt with it for now by rearranging things & including my old 
mmt8 in the scheme: keys> pocket controller> thru box> maq & merger 
(with maq output)> p3> stuff including mmt8> keys again.

so now the mmt8 can capture everything that's going on, all F'd-T-S 
of course, & it's the master clock. it can be set to "internal clock 
only" though it doesn't seem to matter in this configuration, & the 
through-filtering is pretty comprehensive. & it might as well do 
something useful.

in the live rig, there is a five-input merge box holding everything 
together. I think I can get away wth only using three inputs though. 
it's a similar sort of arrangement, except that the clock (from my ES-
1 this time) & the keys are fed to the same merge box as the pocket 
controller & the maq. 
(ES-1> controller> thru'> maq, merge box w/ maq output. keys also to 
merge box, whole lot to p3 &c)

the clock-only output from the p3 in both rigs is now being used to 
run gary's jamman, since it seems to prefer a feed without too much 
else going on.

does anyone else have this sort of thing going on or is it just me?
8-^

I also run a regulated 14V dc around the spare wires in my midi 
cabling, to cut down the number (& weight!) of power supplies & their 
associated cabling. I use easily-available laptop supplies which work 
anywhere & are smooth enough that the midi doesn't get interfered 
with. no problems so far....

duncan/r.m.i.

p.s. hi ken l... when are you going to get a p3? :-)