The No.1 Website for Pro Audio
 Search This Thread  Search This Forum  Search Reviews  Search Gear Database  Search Synths for sale  Search Gearslutz Go Advanced
Roland MKS-70 Upgrade/Modification beta testers wanted
Old 13th March 2013
  #31
Gear Maniac
 

good news

Good news... I bought a JX10. Paid this morning, hopefully I'll receive it in good shape and soon! I am already porting the new MKS70 code for it.
Old 13th March 2013
  #32
Quote:
Originally Posted by grumphh View Post
Not that i am going to be whiney or anything, i think it is a fantastic project, and it will add a lot of value to these synths - but if you add extra envelopes, could you please look into the shapes of those too, and maybe model them on some classic envelope shapes rather than "just" making them faster?

And of course, if possible, let the extra LFO's go way up into audio range (several Khz) as well as having both up and down ramps in addition to the already available shapes?
+1

envelopes on MKS70/JX10 are too linear. thy need to have exponential slope shapes.

also, they are so damn slow that they should be at least four times as fast, to get into 2.5ms range, well.. ideally 10 times faster for 1ms range. if you want punch and snappy basses and sequences.


doing the change from 10ms to 5ms, imho just isn't as worthwhile. sorry to be a partypupper. but as a 25yr+ long mks/jx user i am painfully aware of the shortcomings. i actually sold my mks70 and pg800 this last summer. but might get one back if there is a substantial speed upgrade.

everything else, like more lfos etc would be nice, but not crucial. with speed and expo slopes, the synth would be a beast.
Old 13th March 2013
  #33
Lives for gear
 
thermos's Avatar
This is really cool! Any way to add noise as an lfo option?
Old 14th March 2013
  #34
Gear Maniac
 

The envelopes are already exponential (computed in software). I can provide a few screenshots from the 8051 simulator. A new sample is computed every 10ms, but the CV is updated every 5ms (the 2nd update is a linear interpolation). Using the new CPU, it should be possible to run twice faster, and with a crystal change, 4 or 5 times faster, which should provide what people are looking for.

The change is unfortunately not simple, but I'll try to prototype something, once I get the assigner code stable (on MKS70 *and* JX10). There is a lot of work there to clean all the mess written by Roland. Even if the code is rewritten from scratch, there are many things to support in order to keep the same set of functionalities. The midi implementation is a pain, but is required if we want to keep backwards compatibility.
Old 14th March 2013
  #35
Lives for gear
 
Starspawn's Avatar
 

Im very excited about this
Take your time and let us know when you need testing or any other help.
Old 14th March 2013
  #36
Lives for gear
 
Signifier's Avatar
Colin Fraser of Sequentix P3 and Cirklon fame did an OS update for the Jx-10 and (I think) the MKS-70 years back. It tried to fix a few aspects of the unfinished MIDI Sysex implementation. He would be worth emailing over this and is a good correspondent. Contact him via the Sequentix website...
Old 14th March 2013
  #37
Gear Maniac
 

I talked to Colin already. He added a few things to the JX10 assigner code, but here it's a complete rewrite. I need to add new menus, etc.. so hacking the existing code would have been impossible. The sound board code has also been rewritten, from the Roland code this time (that code is much better than the assigner code).

I have written emulators, etc.. so on that front, I am very good. All I need now is to complete the code (I want the same code base for MKS70 and JX10). I should have a JX10 decent prototype within 2-3 weeks (and should have received my JX10 by then).
Old 14th March 2013
  #38
Lives for gear
 
Signifier's Avatar
This will be amazing! Massive Props to ya!
Old 14th March 2013
  #39
Gear Nut
 
hatzalam's Avatar
JX10 owner here, with Colin's mod installed. This is very exciting to read about, guys. I am definitely bookmarking this thread, and would obviously be interested in the practical implications of this.
Old 15th March 2013
  #40
I would be all over this if you get it out there.
Free running LFO would be great. Actually, if it's not a hardware limitation, having the lfos shared by both tones as an option would be on my wish list. Layered patches lead to train wrecking LFO tempos as it stands. Of course, separate LFOs for pitch and freq would not be a bad thing, either.
Also, being able to get both boards to respond to a parameter change simultaneously (without a pg800 or a controller sending 2 different sysex messages) would also make the synth infinitely more useable.
Thanks for putting your time into the project.
Old 15th March 2013
  #41
Gear Maniac
 

Quote:
Originally Posted by coolveco View Post
Hi.

I am the Fred mentioned earlier in this thread. Happy to see that there is interest in my project.

Here are my current plans :
- get the code tested (I already tested many things) on MKS70
- since I got a lot of requests for JX10, I am working on a JX10 version, and I already have a beta image for JX10, although it is less mature than the MKS70 (one obvious reason is that I don't own a JX10... unfortunately).

- once the code is stable, start adding the new features. To be fair, I already have new features running on sound boards, but with a faked assigner (just a PIC micro controller, used to test the sound board code). Adding the new feature support in the assigner (JX) is now easy with the new code. There is not much room in the current sound board (based on 8031 CPU), because the CPU is used at 100%. That's why I changed that old chip with a modern Dallas equivalent (80C320). It runs 2-3x faster (with the same crystal), so new functionalities can be added. I have 4 envelopes, 3-4 LFOs (free running optional for each)... I am now looking, due to popular demand, to have envelopes 3 and 4 running twice faster.

Just remember that it's a hobby (even if I design hardware/software for a living), so my availability to work on this will vary !

-Fred

Do you have an estimate of when it'll be released? I'll gladly donate or purchase from you when you're finished. Thanks for all your effort.
Old 15th March 2013
  #42
Lives for gear
 
depulse's Avatar
Very interesting. While you are rewriting the oS, could you remove the confusing patches implementations? From a pool of patches two are combined into a program. Change one of the patches and every program using this patch are also changed, very annoying. Just make each program have its own set of patches.
Old 15th March 2013
  #43
Gear Maniac
 

Quote:
Originally Posted by depulse View Post
Very interesting. While you are rewriting the oS, could you remove the confusing patches implementations? From a pool of patches two are combined into a program. Change one of the patches and every program using this patch are also changed, very annoying. Just make each program have its own set of patches.
You are talking about a patch using 2 tones, in Roland vocabulary.

That is not doable because it would take too much memory. Patches are stored compressed (48 bytes) and a tone is 59 bytes. Make 64 x 48 + 50 x 59 and you are using almost all the protected ram. The same reasoning goes for the cartridge.
Old 15th March 2013
  #44
Gear Maniac
 

Sorry, no estimate. At this point, a few more weeks to get the assigner code stable. Then a few more to get all new features implemented and tested... I use all my spare time on this, so things are moving regularly.
Old 15th March 2013
  #45
Gear Maniac
 

Quote:
Originally Posted by major7th View Post
I would be all over this if you get it out there.
Free running LFO would be great. Actually, if it's not a hardware limitation, having the lfos shared by both tones as an option would be on my wish list. Layered patches lead to train wrecking LFO tempos as it stands. Of course, separate LFOs for pitch and freq would not be a bad thing, either.
Also, being able to get both boards to respond to a parameter change simultaneously (without a pg800 or a controller sending 2 different sysex messages) would also make the synth infinitely more useable.
Thanks for putting your time into the project.

Free running LFO is really easy. If folks are interested, I can release that today (eprom change for sound board). Of course, not user settable (that will require the new assigner code).
Separate LFO for pitch and freq => new code required, but will be there.

I don't understand your need to program the same parameter of 2 tones at the same time. Can you explain me why this would be useful ? (You can easily switch from one tone to the other by using the left/right arrows)
Old 15th March 2013
  #46
Lives for gear
 
kirkelein's Avatar
 

It's like christmas in here!
Old 15th March 2013
  #47
Gear Maniac
 

Yes. The ROMs are slightly different (2 bytes, if my memory deserves me well), because they respond to commands from the assigner. The patch is at the same offset.

For the story, the sound board code also responds to a specific command given by the assigner to reset the LFO, but the MKS70 doesn't use that. I haven't seen it on the JX10 yet, but I haven't done much testing with it yet.
Old 15th March 2013
  #48
Lives for gear
 
depulse's Avatar
Quote:
Originally Posted by coolveco View Post
You are talking about a patch using 2 tones, in Roland vocabulary.

That is not doable because it would take too much memory. Patches are stored compressed (48 bytes) and a tone is 59 bytes. Make 64 x 48 + 50 x 59 and you are using almost all the protected ram. The same reasoning goes for the cartridge.
Yes, 1 or 2 tones makes up a patch, you have 50 spaces for your own tones and 64 patches to combine them, very confusing way of working, the Wavestation has the same stupid implementation. Maybe worth looking into doing an upgrade board with more RAM, etc? Change the patch handling, add more memory for patches, add an arpeggiator, better sysex etc.
Old 15th March 2013
  #49
Gear Maniac
 

Quote:
Originally Posted by depulse View Post
Yes, 1 or 2 tones makes up a patch, you have 50 spaces for your own tones and 64 patches to combine them, very confusing way of working, the Wavestation has the same stupid implementation. Maybe worth looking into doing an upgrade board with more RAM, etc? Change the patch handling, add more memory for patches, add an arpeggiator, better sysex etc.
It's not a stupid implementation, but a limitation of the past. Changing the CPU, more RAM, etc.. is basically designing a new assigner board. Too much hassle.

One thing that I may do, once this is "all done" (if that happens!), is to modify M64C or M16C cartridge into something smarter. That thing could contain more RAM to store many more patches/tones. But right now, I have no intent of doing this. (Another option would be to add a RAM chip, on top of the eprom, and do bank swapping (forgot to mention that the whole 64KB of address space is used)). Then you have the issue of powering the RAM from the battery, etc... Again a lot of hassle for little benefit (specifically if you have a huge cartridge).
Old 15th March 2013
  #50
Gear Nut
 
wavesequence's Avatar
Instead of covering coolveco of requests, primary goal should be having a JX-10 that's really MIDI and sys-ex compliant, a JX-10 that can communicate with a simple librarian running under Windows and OSX, without hassles
IMHO adding new envelopes, changing their velocities means also adding features that weren't present in the original synthesizer.
JX-10 is well known for its highs and lows.
Once JX-10 works over MIDI and sys-ex, we can start asking for new features - if coolveco wants to implement them.
Old 15th March 2013
  #51
Gear Nut
 
wavesequence's Avatar
Quote:
Originally Posted by Signifier View Post
Colin Fraser of Sequentix P3 and Cirklon fame did an OS update for the Jx-10 and (I think) the MKS-70 years back. It tried to fix a few aspects of the unfinished MIDI Sysex implementation. He would be worth emailing over this and is a good correspondent. Contact him via the Sequentix website...
BTW I sent him a copule of email to colin@colinfraser.com but he never replied
Old 15th March 2013
  #52
Lives for gear
 
grumphh's Avatar
 

Quote:
Originally Posted by wavesequence View Post
Instead of covering coolveco of requests, primary goal should be having a JX-10 that's really MIDI and sys-ex compliant, a JX-10 that can communicate with a simple librarian running under Windows and OSX, without hassles
IMHO adding new envelopes, changing their velocities means also adding features that weren't present in the original synthesizer.
JX-10 is well known for its highs and lows.
Once JX-10 works over MIDI and sys-ex, we can start asking for new features - if coolveco wants to implement them.
Sorry, but you got that all wrong imo.

It is easy to live with a synth without (proper) midi - look at all the vintage stuff that doesn't have it, and at least the JX does have basic midi for sequencing and program changes.
While i agree that it not the easiest synth to program from the front panel, it is not all that hard to do either - as it doesn't have the ridiculous amounts of parameters that newer (digital) synths have.

...what is really exciting about the upgrade is that this synth with its sloooooow envelopes and just one rather boring LFO could truly become a monster with added mod options.

I for one am drooling at the thought of having two snappy envelopes for filter and amp, and then being able to assign the slow envs with those really long attack and decay stages to e.g. pitch on the modulating oscillator, or oscillator mix, or what have you.
And then add some 60hz LFO (if possible) to that mix in small doses...

Slurrrrp!
Old 16th March 2013
  #53
Gear Maniac
 

Quote:
Originally Posted by coolveco View Post
Sorry, no estimate. At this point, a few more weeks to get the assigner code stable. Then a few more to get all new features implemented and tested... I use all my spare time on this, so things are moving regularly.
Take all the time you need...No rush at all. Getting it right is the most important thing.


Will your operating system fix the midi problems associated with the mks70?? And can you list any other fixes (not feature updates) to the mks70 that your operating system will bring in this thread, or on your own site in the future...or if you have already written up a list of fixes to the mks70 somewhere online please share the link. Thank you.
Old 16th March 2013
  #54
Quote:
Originally Posted by coolveco View Post
Free running LFO is really easy. If folks are interested, I can release that today (eprom change for sound board). Of course, not user settable (that will require the new assigner code).
Separate LFO for pitch and freq => new code required, but will be there.

I don't understand your need to program the same parameter of 2 tones at the same time. Can you explain me why this would be useful ? (You can easily switch from one tone to the other by using the left/right arrows)
Thankfully, I use a midi controller to avoid left and right arrows for the most part. I have posted MKS 70 templates for the Novation Remote elsewhere on GS, and there are BCR 2000 ones out there somewhere too. Life is too short for analog synths without knobs and sliders!

Not being able to adjust a parameter on both tones is a limitation of the MKS 70 in the same way that independent, out of sync LFOs affecting each board is a problem.

Often tones are layered in a patch to create huge pads or doubled, slightly-detuned, unison and/or mono basses and leads. There are less predictable uses as well, of course.

As it stands, only tone A or tone B can be edited at one time. So a filter sweep, for example, can only affect one half of a patch. Pretty weak. Adjusting envelopes is an equally half-hearted experience.

I said the PG 800 could tweak both tones at once, but I don't have one, so I don't know. The KiwiTechnics hardware controller does allow simultaneous editing/control of both tones.

Not sure of what kind of magic you can work in there. I'm guessing that KiwiTechnics handled the problem by sending sysex for tones A and B simultaneously from the same control as an option. My Novation Remote only sends one sysex string per knob. It would be nice to have an option within the MKS itself to "trick" both sound boards into responding to each other's sysex. The code for each parameter is identical except the byte identifying which board is supposed to respond.

Any idea whether sending a single LFO to both tones at once is feasible without soldering? Big LFO filter sweeps across a layered dual tone pad would really let the MKS 70 live up to its potential. Out of sync LFOs on the 2 tones just don't cut it. Setting LFOs to the same value on each tone doesn't sync them, either.

Unless someone knows some trick I'm missing?
Old 16th March 2013
  #55
Gear Maniac
 

Quote:
Originally Posted by Darkstar2010 View Post
Take all the time you need...No rush at all. Getting it right is the most important thing.


Will your operating system fix the midi problems associated with the mks70?? And can you list any other fixes (not feature updates) to the mks70 that your operating system will bring in this thread, or on your own site in the future...or if you have already written up a list of fixes to the mks70 somewhere online please share the link. Thank you.
Can you describe *all* the midi flaws that you know about ? I can really do anything, and I'd like to get midi right.

At the moment, I am adapting the code to the JX10. I want the JX10 to be a MKS70 when local is set to off. Speaking about the local settings, I wonder if there is any utility in having a "UPPER LOCAL" and "LOWER LOCAL". Do people care about splitting the keyboard, with one side local and the other remote ?
(It's not that it's difficult to implement, but if nobody uses that...)
Old 16th March 2013
  #56
Gear Maniac
 

Quote:
Originally Posted by major7th View Post
Thankfully, I use a midi controller to avoid left and right arrows for the most part. I have posted MKS 70 templates for the Novation Remote elsewhere on GS, and there are BCR 2000 ones out there somewhere too. Life is too short for analog synths without knobs and sliders!

Not being able to adjust a parameter on both tones is a limitation of the MKS 70 in the same way that independent, out of sync LFOs affecting each board is a problem.

Often tones are layered in a patch to create huge pads or doubled, slightly-detuned, unison and/or mono basses and leads. There are less predictable uses as well, of course.

As it stands, only tone A or tone B can be edited at one time. So a filter sweep, for example, can only affect one half of a patch. Pretty weak. Adjusting envelopes is an equally half-hearted experience.

I said the PG 800 could tweak both tones at once, but I don't have one, so I don't know. The KiwiTechnics hardware controller does allow simultaneous editing/control of both tones.

Not sure of what kind of magic you can work in there. I'm guessing that KiwiTechnics handled the problem by sending sysex for tones A and B simultaneously from the same control as an option. My Novation Remote only sends one sysex string per knob. It would be nice to have an option within the MKS itself to "trick" both sound boards into responding to each other's sysex. The code for each parameter is identical except the byte identifying which board is supposed to respond.

Any idea whether sending a single LFO to both tones at once is feasible without soldering? Big LFO filter sweeps across a layered dual tone pad would really let the MKS 70 live up to its potential. Out of sync LFOs on the 2 tones just don't cut it. Setting LFOs to the same value on each tone doesn't sync them, either.

Unless someone knows some trick I'm missing?

I see your point. I'll think about it.

About the in-sync LFO on both tones, that one is hard to do. The LFO is software based, and tone is run from its CPU, so you are asking about syncing 2 cpus, running with 2 different clocks. Sorry for that one... not doable. (Even with the soldering iron, that would require connecting the 2 cpus together, one generating a time sync to the other. And then we have the problem that both CPUs are running out of I/O pins...)

About the PG800, as far as I know, it edits only one tone at a time. That's how I have implemented it on my code.
Old 16th March 2013
  #57
Lives for gear
 
Starspawn's Avatar
 

Quote:
Originally Posted by coolveco View Post
Can you describe *all* the midi flaws that you know about ? I can really do anything, and I'd like to get midi right.

At the moment, I am adapting the code to the JX10. I want the JX10 to be a MKS70 when local is set to off. Speaking about the local settings, I wonder if there is any utility in having a "UPPER LOCAL" and "LOWER LOCAL". Do people care about splitting the keyboard, with one side local and the other remote ?
(It's not that it's difficult to implement, but if nobody uses that...)
There is notable delay when using midi to control the JX10, not so much when using it as a midi master, someone mentioned there was some unneccessary process the midi went through going in.
Also there is a problem when splitting the keyboard and using midi, its simply ignored as far as I can tell.
Old 16th March 2013
  #58
Lives for gear
 
grumphh's Avatar
 

Quote:
Originally Posted by Starspawn View Post
There is notable delay when using midi to control the JX10, not so much when using it as a midi master, someone mentioned there was some unneccessary process the midi went through going in.
The guy with the eeprom upgrade (Colin Fraser?) mentions that in his writing.
When just using a sequencer to play notes (at reasonable tempi) along with a drummachine i have not noticed that "delay" - but it might of course show up if you analyze the tracks in a DAW.

(However, might add that all this ms-accurate timing stuff is not really all that important to me, so i might just have missed it... )

Quote:
Also there is a problem when splitting the keyboard and using midi, its simply ignored as far as I can tell.
What problem would that be?
If i set up two sequencer tracks to two midi channels and set the JX up to respond to the same two midi channels i get two independent synths sounding?
Old 16th March 2013
  #59
Lives for gear
 
Starspawn's Avatar
 

Damnit, maybe its user error?
When I set up a split on 2 midi channels and try to control for instance my VP550 hoping to get strings on lower and vocoder on upper, I get either one or both with the full keyboard range. The split is not there at all.

The latency is noticable when I layer both the JX and something else using the same midi info, and it would be nice to have it fixed as its another keyboard Id control it from, not a DAW so I cant shift it.
Its ok when I play just the JX remotely, I play live so I dont even notice any latency then, just adjust to it probably, but it would be nice to layer without workarounds.
Old 16th March 2013
  #60
Gear Maniac
 

Quote:
Originally Posted by Starspawn View Post
There is notable delay when using midi to control the JX10, not so much when using it as a midi master, someone mentioned there was some unneccessary process the midi went through going in.
Also there is a problem when splitting the keyboard and using midi, its simply ignored as far as I can tell.
That one is another hardware restriction. The assigner (HD6303 cpu) has only one serial interface, used for midi. The serial out is going to both sound boards. Each sound board has a 8031 cpu, also with a single serial interface. The output of one sound board is going to the MIDI OUT. When the assigner decides to send midi, it sends F5 to the sound board (means "forward the following data to midi out") and the data. Serial out data is put in a buffer, so depending on what you did before (sending tone data, bender, mod, etc. to the sound board) will influence the latency before the midi out data is sent.

Another hardware restriction...
New Reply Submit Thread to Facebook Facebook  Submit Thread to Twitter Twitter  Submit Thread to LinkedIn LinkedIn  Submit Thread to Google+ Google+  Submit Thread to Reddit Reddit 
 
Topic:
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Similar Threads
Thread
Thread Starter / Forum
Replies
Meriphew / Electronic Music Instruments & Electronic Music Production
20
FlamFive5 / Music Computers
1
jdjustice / Music Computers
2
sbmayer99 / So much gear, so little time!
2
racemize / Music Computers
9

Forum Jump