Roland MKS-70 Upgrade/Modification beta testers wanted
#151
8th October 2013
Old 8th October 2013
  #151
Gear Head
 
Joined: Sep 2013
Posts: 35

At last !!!!!!

I am the Dave that CoolVeco referred to...With an svn number to prove it, Right Fred

For Wavesequence...yep MidiOx works,Send SX works.Midiquest works but you need to speed UP!! the buffers there... even the MC500 Library software works...It sends all the right sysex messages on a patch change so your sysex compatible DAW will send patches to the edit buffer too..ALL WITHOUT AN M64 cart..........bliss.

Also the Tone A/B edit thing has been nailed down so an MKS70 editor will work correctly, there are speed/responsiveness improvements to front panel controls and the keyboard/midi too.....Its like a CPU upgrade....watch this thread.

Dave.

PS will not answer questions about further ENV/LFO/PWM upgrades or updates..that`s CoolVecos domain......PM`s/Pleas will be Ignored..Ta
Quote
1
#152
8th October 2013
Old 8th October 2013
  #152
Gear interested
 
Joined: May 2013
Location: Oxford, England
Posts: 21
My Studio

Great news coolveco.

Just out of interest, in what practical form will the upgrade be? I'm presuming it consists of software download, IC and Eprom changes? Is it something you'll be selling, or just providing the component information and software? Thanks, Louis.
#153
8th October 2013
Old 8th October 2013
  #153
Lives for gear
 
depulse's Avatar
 
Joined: Oct 2008
Location: Europe
Posts: 1,199
My Studio

Quote:
Originally Posted by coolveco View Post
Some update.

The JX10 code is now almost complete, and includes support for the keyboard (duh!), the pedals and the controllers. As a bonus, I already added support to tune tone parameters with C1 or C2 sliders for those without a PG800.

I also improved the midi implementation on the JX10. The parameters for aftertouch, bender, etc.. which were ON/OFF are now OFF/RECV/SEND/ON, so you can decide what gets received and what gets sent (or both). Per tone is still supported...

There are still a few minor issues that I need to fix, but the code is now really usable on the JX10. The MKS70 is of course still supported, and is sharing the same code base.

I must say a big thanks to Dave, who has been beta-testing with me. Without him, the code wouldn't be where it is now.

We are now about to start adding new features. I have yet to decide if more RAM is required on the sound board. Roland's implementation (which is surprisingly good compared to the assigner) uses almost all RAM, and the 80C320 only provides 128 bytes extra. So will be just the 80C320 or maybe more RAM and also a new crystal to get even more cpu cycles...

Finally, to answer a previous question : yes, you can update the sound boards to the 80C320 and still use the old Roland code on the assigner. You will need the new code (same as Roland for now, but with adapted timing to support the faster CPU).

-Fred
Great news. Any way to improve Rolands suboptimal way to handle tones and patches? Could each patch contain their own tones and not just a reference to the tones it is using? As is today changing one tone changes all patches using this tone. If each patch would contain its own tones this would not be a problem anymore.

And one more thing, adding an arpeggiator would be cool.
#154
9th October 2013
Old 9th October 2013
  #154
Gear maniac
 
m4thlab's Avatar
 
Joined: Aug 2013
Posts: 222

This is very exciting, and thanks to Dave for helping Fred on this!

I have money in my fist and I am ready to just throw it at you Fred, take my money... TAKE IT!

I know you were also working on a solution for the dodgy and irreplaceable display coil, in the form of a supplementary/replacement display, and I'd love to know if that's progressed much, or has all your focus been on the os improvements?

One thing is certain, this will elevate the JX and make it far more desirable, and it's going to increase in value, if i didn't have one already I'd be looking to snap one up, word to the wise.
#155
13th October 2013
Old 13th October 2013
  #155
Gear Head
 
Joined: Sep 2013
Posts: 35

Quote:
Originally Posted by depulse View Post
Great news. Any way to improve Rolands suboptimal way to handle tones and patches? Could each patch contain their own tones and not just a reference to the tones it is using? As is today changing one tone changes all patches using this tone. If each patch would contain its own tones this would not be a problem anymore.

And one more thing, adding an arpeggiator would be cool.
I believe that coolveco answered this one a few pages ago, there is simply not enough memory onboard to store 128 user tones, also soldering in a bigger memory chip will not work because the JX/MKS has no way of accessing the extra memory added. (that`s as non technical as I can make it...)

The improved sysex can send and receive individual tones as well as patches including tone data,so for a DAW user you can store your data there for recall and the hardware sequencers MC500 and recently tested Yamaha QX3 will record/send patch/tone/sysex to the synth too...all we will need is a decent MAC/PC librarian and editor for shuffling things around.
#156
17th October 2013
Old 17th October 2013
  #156
Gear nut
 
wavesequence's Avatar
 
Joined: Mar 2013
Location: Mestre - Venezia, ITALY
Posts: 81

Quote:
Originally Posted by Quavermass View Post
I believe that coolveco answered this one a few pages ago, there is simply not enough memory onboard to store 128 user tones, also soldering in a bigger memory chip will not work because the JX/MKS has no way of accessing the extra memory added. (that`s as non technical as I can make it...)

The improved sysex can send and receive individual tones as well as patches including tone data,so for a DAW user you can store your data there for recall and the hardware sequencers MC500 and recently tested Yamaha QX3 will record/send patch/tone/sysex to the synth too...all we will need is a decent MAC/PC librarian and editor for shuffling things around.
Digging in the past, I can suggest the old faithful opcode Galaxy plus Editors that contains also a librarian for the MKS-70.
If the "new" JX-10 EEPROM will deal with sys-ex as MKS-70 does, I think that the same librarian can be used.
The only drawback is that the aforementioned librarian runs only under MacOS9.2.x or OSX 10.4.11 under Classic Mode.

Beware MidiQuest on both Win and Mac.
#157
17th October 2013
Old 17th October 2013
  #157
Gear interested
 
Joined: May 2013
Location: Oxford, England
Posts: 21
My Studio

Quote:
Originally Posted by wavesequence View Post
Digging in the past, I can suggest the old faithful opcode Galaxy plus Editors that contains also a librarian for the MKS-70.
If the "new" JX-10 EEPROM will deal with sys-ex as MKS-70 does, I think that the same librarian can be used.
The only drawback is that the aforementioned librarian runs only under MacOS9.2.x or OSX 10.4.11 under Classic Mode.
Any pointers as to where one might locate a trustworthy keyless version of Galaxy Plus? I've seen a couple on download sites, but I'm quite mistrustful of using such sites.
#158
19th October 2013
Old 19th October 2013
  #158
Gear nut
 
wavesequence's Avatar
 
Joined: Mar 2013
Location: Mestre - Venezia, ITALY
Posts: 81

Quote:
Originally Posted by Louis1960 View Post
Any pointers as to where one might locate a trustworthy keyless version of Galaxy Plus? I've seen a couple on download sites, but I'm quite mistrustful of using such sites.
Subscribe the Opcode Users Group on Yahoo, you'll find it under Files directory.
Remember that you can run Galaxy plus if you have a Mac running MacOS 9.x or earlier and a serial MIDI interface otherwise MacOSX 10.4.11 and Classic and a USB MIDI interface.

http://groups.yahoo.com/neo/groups/opcodeusers/info
#159
20th October 2013
Old 20th October 2013
  #159
Gear interested
 
FreshTrooperXBL's Avatar
 
Joined: Apr 2013
Posts: 13

I'm following this page as I'm interested in both a VFD replacement (if available) and an eprom upgrade. I just got my DSS-1 upgraded so the JX-10 is next on the list. I'm taking my first computer programming class and have worked on electronics for a few years now so I have a lot of respect for Fred. Cheers!
#160
20th October 2013
Old 20th October 2013
  #160
Gear addict
 
Joined: May 2010
Posts: 359

Quote:
Originally Posted by coolveco View Post
Some update.

The JX10 code is now almost complete, and includes support for the keyboard (duh!), the pedals and the controllers. As a bonus, I already added support to tune tone parameters with C1 or C2 sliders for those without a PG800.

I also improved the midi implementation on the JX10. The parameters for aftertouch, bender, etc.. which were ON/OFF are now OFF/RECV/SEND/ON, so you can decide what gets received and what gets sent (or both). Per tone is still supported...

There are still a few minor issues that I need to fix, but the code is now really usable on the JX10. The MKS70 is of course still supported, and is sharing the same code base.

I must say a big thanks to Dave, who has been beta-testing with me. Without him, the code wouldn't be where it is now.

We are now about to start adding new features. I have yet to decide if more RAM is required on the sound board. Roland's implementation (which is surprisingly good compared to the assigner) uses almost all RAM, and the 80C320 only provides 128 bytes extra. So will be just the 80C320 or maybe more RAM and also a new crystal to get even more cpu cycles...

Finally, to answer a previous question : yes, you can update the sound boards to the 80C320 and still use the old Roland code on the assigner. You will need the new code (same as Roland for now, but with adapted timing to support the faster CPU).

-Fred
i am in
#161
21st October 2013
Old 21st October 2013
  #161
Gear interested
 
Joined: May 2013
Location: Oxford, England
Posts: 21
My Studio

Quote:
Originally Posted by wavesequence View Post
Subscribe the Opcode Users Group on Yahoo, you'll find it under Files directory.
Remember that you can run Galaxy plus if you have a Mac running MacOS 9.x or earlier and a serial MIDI interface otherwise MacOSX 10.4.11 and Classic and a USB MIDI interface.

Yahoo! Groups
Thanks for that, was accepted swiftly and found the files. Interestingly I alighted upon this site. I would be surprised if it's still a stock item.

Opcode Galaxy Plus Synthesizer Editor & Librarian - MES Musikelektronik
#162
21st October 2013
Old 21st October 2013
  #162
Gear interested
 
Joined: Sep 2013
Posts: 25

Looking for a single page from the JX-10 service manual. The page showing the actual voice parts that are repeated 6 times on each voice board.

Service manual most commonly available online are missing this page.
#163
21st October 2013
Old 21st October 2013
  #163
Gear nut
 
wavesequence's Avatar
 
Joined: Mar 2013
Location: Mestre - Venezia, ITALY
Posts: 81

Quote:
Originally Posted by Louis1960 View Post
Thanks for that, was accepted swiftly and found the files. Interestingly I alighted upon this site. I would be surprised if it's still a stock item.

Opcode Galaxy Plus Synthesizer Editor & Librarian - MES Musikelektronik
The Opcode Galaxy Plus you're pointing out not only requires MacOS 7.x - 9.2.x and a floppy disk driver but it's also copy protected with an ancient and surpassed method. If the diskette fails you're out of luck.
You'd rather download the Opcode Galaxy Plus pace-free from Opcode users yahoo group.
#164
21st October 2013
Old 21st October 2013
  #164
Gear Head
 
Joined: Sep 2013
Posts: 35

Quote:
Originally Posted by brukernavn32 View Post
Looking for a single page from the JX-10 service manual. The page showing the actual voice parts that are repeated 6 times on each voice board.

Service manual most commonly available online are missing this page.
Grab the MKS70 manual from synfo, this has the info you are missing
#165
21st October 2013
Old 21st October 2013
  #165
Gear interested
 
Joined: Sep 2013
Posts: 25

Thanks! Good thinking.

Finally got to see the difference between JX8 and Super JX's VCF. JX8 filter will not self oscillate. Not so sure if this is only hw related thought?

#166
22nd October 2013
Old 22nd October 2013
  #166
Gear nut
 
wavesequence's Avatar
 
Joined: Mar 2013
Location: Mestre - Venezia, ITALY
Posts: 81

I'm so envious of your knowledge
#167
22nd October 2013
Old 22nd October 2013
  #167
Gear interested
 
Joined: Oct 2013
Posts: 25

Hey folks, I am new in these parts...Long time JX10 user...since the early ninties. I found this thread and can't stop my excitment.

How can we make contributions to keep this project going? I can think of nothing better...

By the way, I am also a JP6 owner/user and love both of these boards. I don't share the opinion that the attack is too slow....The JX10 rocks!

Look forward to watching things here and I am ready to be a beta tester at anytime.

Stephen
#168
22nd October 2013
Old 22nd October 2013
  #168
Lives for gear
 
grumphh's Avatar
 
Joined: Dec 2009
Location: Denmark
Posts: 4,061

Quote:
Originally Posted by Stephencl View Post
I don't share the opinion that the attack is too slow....
May i draw your attention to this clip which shows the difference between the fastest JP-4 attack time and the fastest JX-10? *

Now, i am not saying that it should be attempted to get the JX as fast as the JP, but certainly, if some of the mush in the JX attack phase could be cleared it would be a far more interesting synth.

Try and look at the clip in a waveform editor - the JX has a weird rise time of 5 - 10 ms.

...i have no idea whether this is due to the programming or the electronics - but one thing is certain: The JX does have a very slow attack.



* Just one oscillator, attack at zero and trying to match the rest of the envelope between the synths as well.
Basically, all as equal as those two synths could be programmed.
#169
23rd October 2013
Old 23rd October 2013
  #169
Gear interested
 
Joined: May 2013
Location: Oxford, England
Posts: 21
My Studio

Quote:
Originally Posted by grumphh View Post
May i draw your attention to this clip which shows the difference between the fastest JP-4 attack time and the fastest JX-10? *

Now, i am not saying that it should be attempted to get the JX as fast as the JP, but certainly, if some of the mush in the JX attack phase could be cleared it would be a far more interesting synth.

...i have no idea whether this is due to the programming or the electronics - but one thing is certain: The JX does have a very slow attack.
That's an interesting and quite revealing clip. Granted, it would open up the JX-10 to more percussive/sharper attack sounds. Saying that, with it's bi-timbral, velocity and aftertouch abilities, it's a highly expressive-sounding instrument. It really does cast a shadow over my Synthex somewhat and at approximately 10% of the latter's value. However, I'm all ears as to what improvements may come for the JX.
#170
23rd October 2013
Old 23rd October 2013
  #170
Lives for gear
 
grumphh's Avatar
 
Joined: Dec 2009
Location: Denmark
Posts: 4,061

Quote:
Originally Posted by Louis1960 View Post
That's an interesting and quite revealing clip. Granted, it would open up the JX-10 to more percussive/sharper attack sounds. Saying that, with it's bi-timbral, velocity and aftertouch abilities, it's a highly expressive-sounding instrument. It really does cast a shadow over my Synthex somewhat and at approximately 10% of the latter's value. However, I'm all ears as to what improvements may come for the JX.
Oh yes, it is a very good sounding synth - with a few drawbacks, such as the attack speed, and if that can be fixed in software and is not a hardware limitation the JX-10 would have a small renaissance, i think...
#171
23rd October 2013
Old 23rd October 2013
  #171
Gear interested
 
Joined: Sep 2013
Posts: 25

Quote:
Originally Posted by grumphh View Post
May i draw your attention to this clip which shows the difference between the fastest JP-4 attack time and the fastest JX-10? *

Now, i am not saying that it should be attempted to get the JX as fast as the JP, but certainly, if some of the mush in the JX attack phase could be cleared it would be a far more interesting synth.

Try and look at the clip in a waveform editor - the JX has a weird rise time of 5 - 10 ms.

...i have no idea whether this is due to the programming or the electronics - but one thing is certain: The JX does have a very slow attack.



* Just one oscillator, attack at zero and trying to match the rest of the envelope between the synths as well.
Basically, all as equal as those two synths could be programmed.
Thanks! Not having access to PG800 I admittedly don't know much about sound programming JX-10. I'm wondering how the VCA setting was set in this comparison?

Is it possible to set VCA to Gate or Trigger on JX-10? And will this alter the performance? I'm asking as I've not seen anyone explain the envelopes chart on analog.no showing 1ms for JX-10.
#172
23rd October 2013
Old 23rd October 2013
  #172
Lives for gear
 
grumphh's Avatar
 
Joined: Dec 2009
Location: Denmark
Posts: 4,061

Quote:
Originally Posted by brukernavn32 View Post
Thanks! Not having access to PG800 I admittedly don't know much about sound programming JX-10. I'm wondering how the VCA setting was set in this comparison?

Is it possible to set VCA to Gate or Trigger on JX-10? And will this alter the performance? I'm asking as I've not seen anyone explain the envelopes chart on analog.no showing 1ms for JX-10.
As for my clip, i am pretty sure that i had both filter and vca set to the same eg on the JX. In any case attack would have been set to zero on both eg's - i have no interest in creating skewed comparison clips.
This is the shortest attack i could program for a bass note on those two synths.

Honestly, i have no idea how the 1 ms figure on analog.no is justified.

There is a possibility that the high notes could get there if you have eg keytrack on (haven't measured it), but as it is obvious from this clip, in the low range the shortest actual attack time is much higher than 1 ms.

Also, the gate setting on the VCA does not help in shortening the attack time audibly.
#173
23rd October 2013
Old 23rd October 2013
  #173
Gear interested
 
Joined: Sep 2013
Posts: 25

Quote:
Originally Posted by grumphh View Post
There is a possibility that the high notes could get there if you have eg keytrack on (haven't measured it), but as it is obvious from this clip, in the low range the shortest actual attack time is much higher than 1 ms.
Thanks!

From analog.no "Measurements have been done using the VCA envelope at minimum attack, full sustain and filter wide open.
Sampled at 48kHz using a high pitched tone."

"Filter full open and a high tone" might imply we should test if problem can be related to the key tracker circuit?
#174
23rd October 2013
Old 23rd October 2013
  #174
Lives for gear
 
grumphh's Avatar
 
Joined: Dec 2009
Location: Denmark
Posts: 4,061

Quote:
Originally Posted by brukernavn32 View Post
Thanks!

From analog.no "Measurements have been done using the VCA envelope at minimum attack, full sustain and filter wide open.
Sampled at 48kHz using a high pitched tone."

"Filter full open and a high tone" might imply we should test if problem can be related to the key tracker circuit?
I am not even going to venture guesses where the problem might be as i have no technical knowledge whatsoever.

It is just that on the net i have always read that the software envelopes/processor speed could be the issue.
#175
23rd October 2013
Old 23rd October 2013
  #175
Gear interested
 
Joined: Sep 2013
Posts: 25

Yeah I know that is the common understanding. But now that we actually have a programmer on our "pay-list" we better point him to the correct location before he loses interest

I don't know where the fault is. But the gay over at analog.no has been around long enough to take interest in the subject and document his findings. In my mind we should trust this for now and see if we can tweak tones until we locate other suspicious parameters.

Someone have posted an idea that slow VCA are the problem. I've not seen details about this theory, but he might not meant the actual VCA function in the Super JX. But rather VCA chips used all over the sound boards.

Anyone know other Roland synths using the infamous IR3R05 that are known to have "slow envelopes"? Or devices that use the same chip but are known to be fast?

Edit: I need to rephrase that sentence implying the IR3R05 are slow. That is not what I'm trying to say. What I'm trying to figure out is if the actual filter design around this chip differs from other Roland synths.
#176
23rd October 2013
Old 23rd October 2013
  #176
Gear maniac
 
AstroZon's Avatar
 
Joined: Dec 2011
Location: Titan
Posts: 226

I need the VFD replaced on my MKS-70 as well. Count me in.
#177
23rd October 2013
Old 23rd October 2013
  #177
Gear maniac
 
AstroZon's Avatar
 
Joined: Dec 2011
Location: Titan
Posts: 226

Quote:
Originally Posted by brukernavn32 View Post
Anyone know other Roland synths using the infamous IR3R05 that are known to have "slow envelopes"? Or devices that use the same chip but are known to be fast?

Edit: I need to rephrase that sentence implying the IR3R05 are slow. That is not what I'm trying to say. What I'm trying to figure out is if the actual filter design around this chip differs from other Roland synths.
The MKS-50 and Alpha Juno 1 & 2 all use the IR3R05 Filter/Amp IC - one per voice.

However it's not the filter that controls the attack but the envelope which is in software. Roland had two different design shops back then in their synth division. One did the analog JX series synths and the other did the Junos and Jupiters. Apparently, they wrote different envelope software.

Here's a quick MKS-70 demo I made when my VFD still worked. Everything is the MKS-70 except drums.

MKS-70 Demo

The stabs after the mid break show that it can make some fast attacks. No PG-800 either, just me, the Alpha Dial, and sub menus. One think I discovered early on was to use just one half of the unit - effectively making it a rack mount JX8P. It just seem to produce more useful sounds. I think that those stabs used both boards though.
#178
23rd October 2013
Old 23rd October 2013
  #178
Gear interested
 
Joined: Sep 2013
Posts: 25

Yes I understand the attack are software controlled. But if someone can measure 1ms attack time with high notes I suggest slower attach is eater purposely coded with lower notes (key tracking).

Or hw gates or sw is waiting for sync signal that impact the attach with lower notes.
A note at 220Hz could in worst case then give a whooping 4,5ms delay (1/Hz*1000). Not that I've seen report that the attach is actually deviating. One reason I can think of for an implementation like this would be to eliminate pops when the attack opens.

Just thinking out load here.
#179
23rd October 2013
Old 23rd October 2013
  #179
Gear maniac
 
AstroZon's Avatar
 
Joined: Dec 2011
Location: Titan
Posts: 226

Quote:
Originally Posted by brukernavn32 View Post
Yes I understand the attack are software controlled. But if someone can measure 1ms attack time with high notes I suggest slower attach is eater purposely coded with lower notes (key tracking).

Or hw gates or sw is waiting for sync signal that impact the attach with lower notes.
A note at 220Hz could in worst case then give a whooping 4,5ms delay (1/Hz*1000). Not that I've seen report that the attach is actually deviating. One reason I can think of for an implementation like this would be to eliminate pops when the attack opens.

Just thinking out load here.
Oh, I see where you're going. You could be right on the programming of the lower note's attack times, and it may have been done out of compromise. You suspect that Roland divided the attack along with the note?
#180
23rd October 2013
Old 23rd October 2013
  #180
Gear interested
 
Joined: Sep 2013
Posts: 25

Yes, to better simulate some acoustic instruments. If so it should be implemented as an option with the key tracker. Maybe they didn't have enough memory/pins to implement it as an optional parameter? Or it could be that there is logic there waiting for for sync signal to open attack when the "wave" are close to its null point to avoid pops.
Topic:
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Similar Threads
Thread
Thread Starter / Forum
Replies
Meriphew / Electronic Music Instruments & Electronic Music Production
13
FlamFive5 / Music Computers
1
jdjustice / Music Computers
2
sbmayer99 / So much gear, so little time!
2
racemize / Music Computers
9

Forum Jump
 
Register FAQ Search Today's Posts Mark Forums Read

SEO by vBSEO ©2011, Crawlability, Inc.