BlahBlahCafe - Jean-Michel Jarre forum

Great !
It's gonna work soon I guess

With Playback-Function and Clockgen (will be later divided on option):

http://www.phasereffect.com/nopub/test3.wmv

I have problems on reading and displaying row 0.0, but that´s a program mistake...so in the Vid it sttarts at row0, but programmed is row1...

Driving LED´s against Decoder´s VCC output is also not state of the Art, but as long they work so...where´s the problem...

Steffff

Looks good!

PHASEREFFECT wrote:With Playback-Function and Clockgen (will be later divided on option):

http://www.phasereffect.com/nopub/test3.wmv

I have problems on reading and displaying row 0.0, but that´s a program mistake...so in the Vid it sttarts at row0, but programmed is row1...

Driving LED´s against Decoder´s VCC output is also not state of the Art, but as long they work so...where´s the problem...

Steffff

Great work!! Any chance you are going to share the schematics, pcb and bom? That would be nice! Software is fine but the real deal even better!
I think I know a way to fabricate the matrix. It involves plain copper strips and hardboard. The only thing you need to find are the pins. You do need to be a real DIY to be able to fabricate it (my plan is to duplicate the EMS Matrix. It's really easy in makeup)...

J.

Hm. -I- would share it, no problem, but you have to consider the amount of work which is within the software and building prototypes with my little incoming. Not that i would make the fat money with it, but a few € for the Chip including the Software would be -in my opinion- Ok. The thing with a selfmade matrix is the debounce/electrical contact while inserting them. The MatriSequencer doesn´t decode the holes -permanently- like my version does (it stores the entire sequence inside the memory of the Controller); it runs from row to row in play mode. If i would debounce each hole, i would need at least 100 msec to get a absolute value and that´s far too long for decoding a 50*20 matrix. No, latched Pushbuttons (so my design works) with a debounce of approx. 5 msec must be slow enough, normal values are about 25-30 msec. And debouncing is a -must-.
I won´t use the "original" MatriSequencer decoding because this type of decoding, which i use, allows a better sequencer programming, because 1) it jumps automatically to the selected row and 2) it plays after that automatically the keys stored in this row. So you´ll save two buttons for forward/backward. I have months ago another sequencer programmed, which works with potentiometers on each row and jumps automatically (when turning a potentiometer) to the selected row and play back the stored value.

The other questions is, if Mr. Geiss would be happy if i would release a "copy" of his MatriSequencer without his agreement and i´m pretty sure that he holds a sort of patent on this type of sequencing, like Kraftwerk has on their E-Drums (or even JMJ, since it is his -own- machine). I can program the software, building this machine would be a part of one of my friends (he has more experience than i with such large devices), any sort of electrical analysis must be done by another friend, testing the device on extreme circumstances another part of me, since i´d like to include more than one Midi-out. And then it comes up to Michel Geiss or even JMJ itself, what they say. And, -unfortunaly-, i have -no- contacts to them, even i´m pretty sure some of you guys could manage this. But i asked -so often- the past 7 years for different reasons, but nothing happened (except Francis, many thanks again!). I believe they simply won´t have anything to do with me. Don´t ask -me- why.

Greetings
Stef

For the dials I assume "PortL" and "PortR", same goes to "StepL" and "StepR" and "StopL" and "StopR". I believe that the "Port"-dials are for setting the CV-Range or generally the Output-CV-Voltage for controlling different Modules, when using other things instead of Pitch-CV. -Or- simply the Gate.

Stef

PHASEREFFECT wrote:For the dials I assume "PortL" and "PortR", same goes to "StepL" and "StepR" and "StopL" and "StopR". I believe that the "Port"-dials are for setting the CV-Range or generally the Output-CV-Voltage for controlling different Modules, when using other things instead of Pitch-CV. -Or- simply the Gate.

Stef

Hmmm... intresting! At first I thought the PortL&R where jacks capped off by the red plastic but lateron I recognised potentiometer dials in them too

What I see (top2bottom) is:
- PortL jack, COM step switch, PortR jack
- Stop RET(urn)? switch, Tempo LED, Stop RET(urn)?
- 2 BCD switches (could be max lenght), tempo 20-turn dial, 3 lamps of wich the last one has a different colour
- PortL dial, reset pushbutton, PortR dial
- StepL, power, StepR

There must be more outputs and they are presumably at the back. Suppose the portL/R jacks are CV....then where is the gate???

I've created a 50-port shift register but stupid Reaktor won't allow loops.. I really hope next version it's possible to put in some c-style code to speed things up... The GFX is updated (again and will be in the future) but for now it works. It counts from 1 to 50 then sadly stops:



I have to think of something else then logic.... And I really should call it the GMS150 instead of 125.. 21*50 pinholes = 1050

J.

I finally have it working (more or less) when it loops it 'waits' one extra note

The sequence from the photo (studio_3) is as follows:

I've used the RMI emulation as synth because I needed some sound source. First you hear the sequence slow than twice as fast. I've just fiddled with the RMI nothing specific... in the end you can hear the harmonic faders being faded out. I'm sure Geiss would have thought it to be impossible to hook the GMS to the RMI

Anyone any idea where this sequence is from?? I've looped it @ position 32...the only thing that can be wrong is the octave switching. I assume from no pin = 0 to pin4 = +4 counted from the bottom to the top. There is logic behind this. Geiss started with a C on the bottom in the 12 above. If not then the lowest note would be a B. That seems not logical for a sequence...

Get it here. I won't be liable for headaches or ear-damage :] Study...observe and listen. Quality is unfortunately low and I noticed i've been through the 0dB barrier.. sorry for that..

It seems the first row is nothing more then a reference where to loop since you have to use the BCD switches to define the length and there is no label at all... I have the feeling the routing is for external outs independant of the PortL/R. For example filter,volume or adsr triggering or what you want and that's the juice

Now it's zzzz time!

J.

Anyone any idea where this sequence is from??

Maybe Matrisequencer is more complex you imagine.
It's look like "Globe Trotter" or "Chronologie 2" sequences.

VinceBlood wrote:

Anyone any idea where this sequence is from??

or "Chronologie 2" sequences.

Thats what i first thought,looking at that Falling diagonal line,theres a sequence in Chrono 2 that uses a falling gliss that starts just before the main verse

But.... JMJ had the digisequencer in the Chronologie era - so I wonder why he would use the matrisequencer in stead?

Hm. If i would have two cars, i would drive them both, as long i have the money to do so. Y not? Same goes to music devices. I would -prefer- the device which -seems- to be obsolete, but that doesn´t mean that i don´t use my other devices.

Greetings
Stef

All remember this pic?????



It's now copied (just the left side)



and it sounds like this

That is if I got all pins right.. I know the sequence though just not wich track.. I've switched from loop @ 32 to loop @ 48...50... is this a 50+ seq? stay tuned !

ow and I created switchable between midi tempo and free run this is @120bpm and did some more GFX work.

J.

Doesn't sound familiar to me, but this may be due to the sound used...

Anyway - great work!!!

He I thought about trying to make the sequences from the pictures. Have you tried the sequence from the larger picture, the one you used for the software?