DSP Optimized devices

[djmicron]

I'm converting some devices to be loaded straight on specific dsp.
The list will grow as i convert them.

Here the download page:

http://goo.gl/x4qeAQ

The use is simple, the modules have the dsp number on the name, and synthesizers have also a fixed polyphony set on the module.

Once the module is loaded, open up the surface and push the button to open the main device surface as devices are protected and the only way i have found to open the original surface is to use midi cc sent by the button of course.
The default cc for opening surface is 116, but you can change it.

Demo videos are here:

http://youtu.be/038YnIqneI4

http://youtu.be/NKQ1a7fkS98

http://youtu.be/O4zRFTefRwE

Thank you

[dawman]

I just dl'd them and loading them into the XITE-1 now.
This is similar to Modular now.
I'll see if this allows larger projects.

Thanks So much for your time...

[dawman]

Works great. Recall prooves they remember the assignments.

Well Done Boss...

[Bud Weiser]

I'm converting some devices to be loaded straight on specific dsp.
The list will grow as i convert them.
Thank you

We thank YOU !

Watched all the vids now and downloaded.
Maybe you´re able to optimize the ZARGs too ???

DjMicron, maybe you should move to germany and enter the S|C developer team.

thx

Bud

[djmicron]

Maybe you´re able to optimize the ZARGs too ???

I'll do some test and see what can be done on protected devices...

[dante]

How bout a version for the 'D' ? I don't have a 'DSP11' ..... just kidding...don't really need it

Well done though.

[djmicron]

I think you are right Dante,
the xite-1D is 10 dsp, but i have not found in deep details about the dsp's.
Is it like the xite-1, where the first 6 dsp are older sharc for I/O operations, or is it different?
If so, i could add devices optimized for dsp 7 to 10.
Thank you

[borg]

Is it like the xite-1, where the first 6 dsp are older sharc for I/O operations, or is it different?
If so, i could add devices optimized for dsp 7 to 10.

that would be great!

lately, I have been using scope more and more again, and have been wondering about DSP assignment for the smaller D.
Enough to be found about the full Xite, but not the D.
Haven't looked that hard yet though. So, mixers are better on the first 'new' sharcs?

Thanks a lot!

[dante]

I think you are right Dante,
the xite-1D is 10 dsp, but i have not found in deep details about the dsp's.
Is it like the xite-1, where the first 6 dsp are older sharc for I/O operations, or is it different?
If so, i could add devices optimized for dsp 7 to 10.
Thank you

yes the DSP 7 - 10 are the 21369 same as on full XITE with the first 6 being the older DSP for compatibility and IO etc also the same as on full XITE.

The DSP on full XITE are arranged in 4 slots. Can't remember whether the D has 1 slot of 4 or 4 slots of 1 - Ralf once told me but I forgot. Anyway this smaller design lead to simpler architecture for coding purposes, but may also make the gains of manual DSP assignment redundant or minimal. No idea but of course it's open to experimentation.

[jksuperstar]

I think it's four slots of 1, so that means 6 "old" dsps also.

[Bud Weiser]

The DSP on full XITE are arranged in 4 slots. Can't remember whether the D has 1 slot of 4 or 4 slots of 1 - Ralf once told me but I forgot.

My guess is, it´s 4 slots of one,- so generally same architecture as XITE-1 and the same mainboard in use.
Less SAT connection errors because DSPs #3-#6 have only to communicate w/ ONE (1) of the new SHARCs,- not 3.
Higher load (%) possible for the time being.

Bud

P.S.:

sorry JKS,- you posted while I typed ...

[jksuperstar]

I was just thinking of using this scheme to get more voices from a synth.

So: a device is placed on a DSP, say a prodyssey. You get 7 voices on 1 DSP. Then you load another 7 in the DSP next door. Load the same preset into both, and link them with MIDI..so the same preset is loaded, and turning knobs in 1 with effect the other (so same MIDI map is loaded also).

We just need maybe a modular device to split the incoming MIDI notes from 14 voices into two seven voice outputs, and configure the method for dropping notes above 14.

In this way, a John Bowen Solaris, with only two voices on each DSP, could be loaded onto multiple DSPs to get more voices.

Anything else we'd need to consider to make this work?

[jksuperstar]

The DSI evolver and polyevolver kinda work his way.

[Bud Weiser]

I was just thinking of using this scheme to get more voices from a synth.

So: a device is placed on a DSP, say a prodyssey. You get 7 voices on 1 DSP. Then you load another 7 in the DSP next door. Load the same preset into both, and link them with MIDI..so the same preset is loaded, and turning knobs in 1 with effect the other (so same MIDI map is loaded also).

We just need maybe a modular device to split the incoming MIDI notes from 14 voices into two seven voice outputs, and configure the method for dropping notes above 14.

In this way, a John Bowen Solaris, with only two voices on each DSP, could be loaded onto multiple DSPs to get more voices.

Anything else we'd need to consider to make this work?

That kind of "MIDI controller device" is already in my Kurzweil PC361.
When you´re in Setup Mode, you have ZONEs which are in fact,- MIDI Channels.
In Setup Mode there´s a "notemap-parameter" and options which notes play on which MIDI channel when keys are been pressed.
For 2 Zones (or 2 MIDI channels) playing different notes,- p.ex. odd ones on MIDI channel 1 and even ones on MIDI channel 2, set Zone 1 to "1 out of 2" and Zone 2 to "2 out of 2".
There are more combinations possible.
It´s the only feature which makes 2 Oberheim Xpanders a 12-voice poly instrument and would make 2 Prodyssey on 2 different chips and assigned to 2 different MIDI channels the same,- a 14 voice Prodyssey p.ex. when poly limit is 7 voices for each Prodyssey being loaded and they play the same patch.

Must be a relative simple algo, filtering odd and even note numbers from incoming MIDI data stream on a given MIDI channel, then odd notes play on the original MIDI channel an even ones MIDI channel n+1.

That feature was also in other MIDI sound modules for the MIDI output and cascading 2.

Bud

[jksuperstar]

While the kurzweil may support this, a module could support all keyboards. If the special MVC (as I'll call it) was placed on XITE DSP 3, it'd have very low latency to each chip in the slot for dsps 7-11-15.

I'll start thinking how to get this done in the SDK.

[sunmachine]

My guess is, it´s 4 slots of one,- so generally same architecture as XITE-1 and the same mainboard in use.

Yes, they probably just removed DSP 11 to 18 from the chip layout on the XITE-1 D.
http://www.hitfoundry.com/issue_07/xitemast.htm

[shayart]

Wow
This is great!!!!!!!!!!!!!!!!!!!!!!!!!!!

This works in 32 bit?

[djmicron]

Wow
This is great!!!!!!!!!!!!!!!!!!!!!!!!!!!

This works in 32 bit?

yes

[djmicron]

I was just thinking of using this scheme to get more voices from a synth.

So: a device is placed on a DSP, say a prodyssey. You get 7 voices on 1 DSP. Then you load another 7 in the DSP next door. Load the same preset into both, and link them with MIDI..so the same preset is loaded, and turning knobs in 1 with effect the other (so same MIDI map is loaded also).

We just need maybe a modular device to split the incoming MIDI notes from 14 voices into two seven voice outputs, and configure the method for dropping notes above 14.

In this way, a John Bowen Solaris, with only two voices on each DSP, could be loaded onto multiple DSPs to get more voices.

Anything else we'd need to consider to make this work?

i'm also working on something like this

[yayajohn]

I was just thinking of using this scheme to get more voices from a synth.

So: a device is placed on a DSP, say a prodyssey. You get 7 voices on 1 DSP. Then you load another 7 in the DSP next door. Load the same preset into both, and link them with MIDI..so the same preset is loaded, and turning knobs in 1 with effect the other (so same MIDI map is loaded also).

We just need maybe a modular device to split the incoming MIDI notes from 14 voices into two seven voice outputs, and configure the method for dropping notes above 14.

In this way, a John Bowen Solaris, with only two voices on each DSP, could be loaded onto multiple DSPs to get more voices.

Anything else we'd need to consider to make this work?

I did something like that a while ago with Six String and SpaceF key splitter. Not sure if it is exactly what you are talking about but it worked great. Here is the topic. Scroll down a little to the pic of the project with the key splitter.

http://forums.planetz.com/viewtopic.php ... =+max+poly