jsynthlib-devel — Mailing list for developers and users interested in improving the JSynthLib.

[Jsynthlib-devel] Complex sysex parameters

[federico <xa...@in...>]

hi all,

today i decided to develop my MKS7-Editor on JSynthLib.
it was everithing ok with envelopes and linear, one-value, parameters.

now i'm getting in trouble with complex parameters.
here is an example:
param16:

 bit 6: unused
 bit 5: chorus on/off
 bit 4: saw-wave on/off
 bit 3: pulse-wave on/off
 bit 2,1,0: octave range (001=16', 010=8', 100=4')
 
how can i make 4 different widgets that manipulate the same param_number?
please consider that if go to change octave range, i should mask other 
bits, like this:
 new_val = (old_val & 0x28) | (oct_range & 0x03)
to avoid re-setting to zero saw/pulse/chorus

thanks

-- 
Federico

Re: [Jsynthlib-devel] JSynthLib-0.20.0 is released

[Thorsten K. <Tho...@gm...>]

Hi,

I noticed that the small images for waveforms and algorithms
are not displayed in the MIDIbox FM editor when JSynthLib
is started from the .jar file (java -jar JSynthLib-0.20.0.jar)

They are displayed when I unpack the .jar and start JSynthLib
with "java JSynthLib"

Example for start from .jar:
http://www.jsynthlib.org/screen8.jpg

Example for direct start:
http://www.ucapps.de/midibox_fm/mbfm_patch_panel_pre1.gif

Do I have to change anything in my code to get this working?

Best Regards, Thorsten.

deprecated class/member (Re: [Jsynthlib-devel] NameValue.java)

[Hiroo H. <hir...@co...>]

Rib> NameValue doesn't seem to be used anywhere.  Should we remove it?

Yes. I've removed the deprecated members and NameValue class.

And I made the layout of tab windows in the preference window consistent. 
-- 
Hiroo Hayashi

[Jsynthlib-devel] NameValue.java

[Rib R. <ri...@gm...>]

NameValue doesn't seem to be used anywhere.  Should we remove it?

Re: [Jsynthlib-devel] BankDriver requests

[Bill Z. <wrz...@po...>]

Steven Schmidt wrote:

> From the dialog I select Bank E, patch 9.  Patch 10 from Bank G gets 
> stored at Bank E, patch 9.

Perhaps you've stumbled across the 'copy patch' command instead of the 
'upload patch' command?  Or maybe the synth mangled the sysex it sent 
you so that when you send it back, it doesn't work.  Only suggestion I 
can think of is make sure you have the right command, and make sure the 
bytes are correct.

-Bill

Re: [Jsynthlib-devel] BankDriver requests

[Steven S. <ste...@co...>]

Hiroo Hayashi wrote:

>Well I might misunderstand you problem.
>
>Steven> Whoops - spoke to soon.  I can request any patch from the synth and save 
>Steven> it in a patchlib, but when I try to store a patch from a patchlib to the 
>Steven> synth, it grabs the GM patch (bank G) at patchNum + 1 and saves it to 
>Steven> the bank I specified.  If I then try to play the patchlib patch, it 
>Steven> plays the GM patch.  What's up with that?  I'm hoping this is familiar 
>Steven> to someone with some experience at writing synth support who can show me 
>Steven> the error of my ways. 
>
>The only problem here is "when I try to store a patch from a patchlib to the 
>synth, it grabs the GM patch (bank G) at patchNum + 1 and saves it to 
>the bank I specified".  In other words, store command sends a patch
>which is next to the patch specified.  Am I correct?
>  
>
Correct-sort of.  I am expecting it to load the patch that I have 
highlighted in the patchlib, but it is getting a GM patch from the 
synthesizer and storing it instead. Here is a more detailed example:  
 From JSL, I select new library.  From 'unsaved library', I select 
Patch->get.  From that dialog, I choose Bank A, patch 27.  Patch A-27 is 
retrieved.  I highlight the patch and select 'store'.  From the dialog I 
select Bank E, patch 9.  Patch 10 from Bank G gets stored at Bank E, 
patch 9.  Obviously, its only doing that because I must be telling it to 
do that, I just can't see where I'm telling it to do that.  Any help 
would be appreciated. 

Do you think a different list for synth-driver development might be 
helpful?  Issues like the very hard to decipher Midi Implementation 
documents could be dealt with there along with how to use them in driver 
deveopment.  Just a thought, in case the core developers prefer not to 
deal with these issues.

Re: [Jsynthlib-devel] BankDriver requests

[Hiroo H. <hir...@co...>]

Well I might misunderstand you problem.

Steven> Whoops - spoke to soon.  I can request any patch from the synth and save 
Steven> it in a patchlib, but when I try to store a patch from a patchlib to the 
Steven> synth, it grabs the GM patch (bank G) at patchNum + 1 and saves it to 
Steven> the bank I specified.  If I then try to play the patchlib patch, it 
Steven> plays the GM patch.  What's up with that?  I'm hoping this is familiar 
Steven> to someone with some experience at writing synth support who can show me 
Steven> the error of my ways. 

The only problem here is "when I try to store a patch from a patchlib to the 
synth, it grabs the GM patch (bank G) at patchNum + 1 and saves it to 
the bank I specified".  In other words, store command sends a patch
which is next to the patch specified.  Am I correct?
-- 
Hiroo Hayashi

Re: [Jsynthlib-devel] BankDriver requests

[Steven S. <ste...@co...>]

Steven Schmidt wrote:

> Hiroo Hayashi wrote:
>
>> For your single editor problem, I recommend you to use Windows->MIDI
>> Monitor.  Using debugger may be a good way.
>>
> Would Windows-MIDI Monitor be a windows program?  I'm working in 
> Linux, not Windows.

Ok - I see what you mean - yes I've been using that.

Re: [Jsynthlib-devel] Change in DeviceAddDialog

[Joe E. <jo...@em...>]

Hiroo Hayashi wrote:

>Someday when we have a hundreds of synths
>support for a manufacture we might want to have GROUP_MANUF_TYPE (it
>might be only if the manufacture names product names inconsistently).
>  
>
Right. Types, at this point, aren't really useful because there aren't 
that many devices supported. However, as the list grows, I see this 
becoming useful someday... which is why I wrote support for it now. The 
reason I committed the source with "GROUP_TYPE_MANUF" is just so that 
everybody could see (and test) the full working of the tree.

Also, one use for GROUP_TYPE_MANUF (or, at least, GROUP_TYPE) that I can 
think of is this. Suppose I'm about to go out and buy a new synth, drum 
machine, controller board, or whatever.... and that I want to make sure 
that the item I buy is supported by JSL. In that case, I'd want to 
browse the items by their type.

>And as you wrote grouping synths into types is not easy.
>  
>
Well, as I see it, it kinda breaks down into:
- Controller *and* synth in one -
   Keyboard Synth (like a Korg M1)
   Guitar Synth (like a Roland GR-33)
   Drum Module
   Other Synth (Can't think of any at the moment)
- Controller *only* -
   Keyboard Controller (like a Yamaha AX-7)
   Guitar Controller (like a Yamaha G-50)
   Other Controllers (like wind controllers, like Yamaha's midi saxophone)
- Synth *only* -
   Rackmount/Standalone Synth (like a Kurzweil K2600R)
   Drum Machines (like a Boss DR-505)

Of course, I'm probably giving guitars too much prestige by setting them 
apart because that's what I deal with a lot.

>And if synths are grouped by manufacture names, it is superfluous to have
>a manufacture name on each node.
>  
>
I'm not sure what you mean here. Do you mean that the manufacturer 
folders shouldn't have the manufacturer *names* on them, or that we 
don't need manufacturer folders at all if the list is merely sorted?

>And minor think; use ErrorMsg.reportStatus() instead of
>System.out.println().
>  
>
Argh! I thought I took all of those out. I use println's when I need a 
quick debugging aid. Where did I leave one in?

- Joe

Re: [Jsynthlib-devel] BankDriver requests

[Steven S. <ste...@co...>]

Hiroo Hayashi wrote:

>For your single editor problem, I recommend you to use Windows->MIDI
>Monitor.  Using debugger may be a good way.
>
Would Windows-MIDI Monitor be a windows program?  I'm working in Linux, 
not Windows.

My question about the bank driver though, is it intended to process an 
entire bank?

Re: [Jsynthlib-devel] CCSender Class for Alesis DM5

[Jeff W. <jww...@ya...>]

--- Hiroo Hayashi <hir...@co...> wrote:
> Jeff> deleted. Can one of you who have admin
> authority take
> Jeff> care of this?
> 
> I believe you can do that by yourself.

I tried removing it with my CVS client. It updated the
version on my local copy to -1.1 and then it told me I
had to do a commit. When I did the commit I got some
sort of error back from the CVS server.

I'm not a CVS expert so I would feel more comfortable
if someone who knew what they were doing could take
care of it.

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

Re: JSLFrame related fixes (Re: SDI mode (Re: [Jsynthlib-devel] change proposal))

[Hiroo H. <hir...@co...>]

Hi,

Rib> You're probably right about the version string.  I also agree that
Rib> checking for the ScreenMenuBar probably shouldn't be in isMac().  The
Rib> problem is that without it, JSynthLib starts up with an empty menu,
Rib> and a single frame (with no title bar) containing a menu.  Someone who
Rib> see's that is going to think that JSynthLib is broken.  By making
Rib> isMac return false in this case, the behavior is acceptable.  Another
Rib> solution, if we had a library open automatically, would be to hide the
Rib> "invisible" frame and the toolbar on a Mac without the screen menu
Rib> bar, but still make JSynthLib quit when you close the last frame.  But
Rib> this seems too complicated.

Why don't we enable 'invisible window mode' only when MacOS and
ScreenMenuBar mode?

    if (isMac && System.getProperty("apple.laf.useScreenMenuBar").equals("true"))
        ((JSLFakeDesktop) proxy).createInvisibleWindow(mb);
    else
        ((JSLFakeDesktop) proxy).createToolBarWindow(title + " Tool Bar", mb, tb);

I've checked in this fix.

Rib> > And if we define default file, we can use "Patch Library Path" which
Rib> > user specifies.  But adding a user preference "default library" may be
Rib> > better idea.
Rib> 
Rib> I didn't know you could do that.  However I assume that most Mac users

I added the feature Oct 31, 2004.

Rib> will use the Application Bundle, which doesn't have a way to specify
Rib> command line arguments.  So I think it would still be a good idea to
Rib> have an option to open a library on startup.
Rib> 
Rib> This would mean another place in AppConfig where we test for
Rib> MacUtils.isMac().  Maybe we need to come up with a better way have
Rib> default options be OS dependent.

I agree with you it's good idea to add preference option to specify the
default library file for any OSs.  I can do that easily.

Thanks.
-- 
Hiroo Hayashi

Re: [Jsynthlib-devel] BankDriver requests

[Hiroo H. <hir...@co...>]

Hi,

Steven> lets me find the answer myself.  Not so with this one.  I would think 
Steven> that a bank driver requests a dump of the entire bank to store as a 
Steven> patchlib, but the code looks like it looks for a single patch from the 
Steven> synth, not from the big dump.

If you request a big dump to your synth, the synth will return big dump.

Steven> Then again, when requesting a patch from 
Steven> the bank driver, I can't select a specific patch.  What exactly is the 

I may be misunderstanding your question, but let me try.  The methods BankDriver
(IBankPatch interface) do the job.  You may have to override some of
them.

Steven> Also, my (Korg Triton) device and single driver are working.  Should I 
Steven> check them in or wait till the editor is done?
I can check-in your working driver.  Editor is optional.

For your single editor problem, I recommend you to use Windows->MIDI
Monitor.  Using debugger may be a good way.
-- 
Hiroo Hayashi

Re: [Jsynthlib-devel] CCSender Class for Alesis DM5

[Hiroo H. <hir...@co...>]

Jeff> deleted. Can one of you who have admin authority take
Jeff> care of this?

I believe you can do that by yourself.
-- 
Hiroo Hayashi

revision numbering (Re: [Jsynthlib-devel] Drivers/Editor for Behringer V-Amp 2)

[Hiroo H. <hir...@co...>]

Bill> > If there are any showstopper bugs we can fix them without having to go 
Bill> > to 0.21.
Bill> >
Bill> You should go to 0.21 anyway, if any problems appear.  Every public 
Bill> release should get a new number and date stamp.  Otherwise everyone has 
Bill> to keep track of *which* 0.20 they have.

Don't worry. I tagged and named the release 0.20.0.  Brian just calls it
0.20-beta only on the JSynthLib home page.  If we don't see any
showstopper, he will just change the home page.  If we fix a bug, we
will release 0.20.1 with the bug fix without adding features added into
0.21 (main branch).

For 0.19, after releasing 0.19-alpha, 0.19-pre1, and 0.19-pre2, nobody
release 0.19.  Nobody (including me) did not want to spend his time to
release a same thing.  I decided to release 0.20.0, 0.20.1, and so on. 
It was my bad not to write this issue on the ML. 
-- 
Hiroo Hayashi

Re: [Jsynthlib-devel] Eclipse instructions

[Hiroo H. <hir...@co...>]

Rib> I wrote some more detailed instructions for setting up Eclipse for
Rib> JSL, which are now available at http://ribrdb.com/jsl/eclipse.html
Rib> Should we maybe put this or a link to it in programming.html?

This is what I wanted.  I've merged it into programming.html.

Thanks.
-- 
Hiroo Hayashi

Re: [Jsynthlib-devel] Change in DeviceAddDialog

[Hiroo H. <hir...@co...>]

Hi Joe,

Joe> ---------------------------------
Joe> Explaination of the layout
Joe> ---------------------------------
Joe> The tree supports grouping by manufacturer, type (ie, synth, controller, 
Joe> drum machine, etc.), manufacturer-then-type, type-then-manufacturer, or 
Joe> no grouping at all (in which case it looks just like the old JList, 
Joe> pretty much).
Joe> 
Joe> It is *currently* configured to group by Type, then by Manufacturer. The 
Joe> problem is, devices don't have types yet. In the meantime, so that we 
Joe> have something to test, the type is the first letter of the manufacturer 
Joe> name. So, a "Kawai K4" will have type "K". This is why the tree 
Joe> initially starts out with just a bunch of folders with single-letter 
Joe> names. So, the Kawai K4 is in "/K/Kawai/Kawai K4".
Joe> 
Joe> If you want to experiment with different groupings, look in 
Joe> DeviceSelectionTree and change the groupStyle field to GROUP_NONE, 
Joe> GROUP_MANUF, GROUP_TYPE, GROUP_MANUF_TYPE, or GROUP_TYPE_MANUF (which is 
Joe> what it is now). Should we just hard-code this a certain way, or let the 
Joe> user change it with a setting?

It think only having GROUP_MANUF is enough.  This dialog is used for a
user to select one of synths he owns. He knows the manufacture of his
synths.  I don't think it is worth to add complexity of having a setting
combobox (or whatever). Someday when we have a hundreds of synths
support for a manufacture we might want to have GROUP_MANUF_TYPE (it
might be only if the manufacture names product names inconsistently).
And as you wrote grouping synths into types is not easy.  There are lots
of kinds of MIDI devices.  Grouping policy may be different between
users.

And if synths are grouped by manufacture names, it is superfluous to have
a manufacture name on each node.

And minor think; use ErrorMsg.reportStatus() instead of
System.out.println().


Anyway great job!  Thank you.
-- 
Hiroo Hayashi

Re: [Jsynthlib-devel] How to write requestPatchDump?

[Jeff W. <jww...@ya...>]

Bill>Meanwhile, I discovered that core.Driver,
AlesisDM5SgSetEditor, Bill>and 
XMLDriverImplementation all send NOTE_ON twice when
told Bill>to play a patch.  Shouldn't one of those be
an OFF?

In the case of the AlesisDM5SgSeteditor, the second
note on has a velocity of zero, which makes it the
same thing as a note off.
Sending a note on with velocity of zero rather than a
note on is sort of a MIDI hack that came about to make
use of running status. Here's a link to the MIDI spec
for further reading:
http://www.borg.com/~jglatt/tech/midispec.htm


	
		
__________________________________ 
Celebrate Yahoo!'s 10th Birthday! 
Yahoo! Netrospective: 100 Moments of the Web 
http://birthday.yahoo.com/netrospective/

Re: [Jsynthlib-devel] No-frills sysex device/driver/editor?

[Jeff W. <jww...@ya...>]

Joe>Well, I don't really care about *editing* as much
as just *seeing*. 
Joe>Something that just broke the data into your
standard hex-editor 
Joe>*style* 

There's a static method called hexDump in Core.Utility
that you might want to check out. You give it a
bytearray and it returns a string type containing a
formatted hex dump of whatever you need. You can just
put it in a System.out.println call. Nothing fancy,
but it's quick and easy and it does the job.


	
		
__________________________________ 
Celebrate Yahoo!'s 10th Birthday! 
Yahoo! Netrospective: 100 Moments of the Web 
http://birthday.yahoo.com/netrospective/

[Jsynthlib-devel] generic editor / library

[federico <xa...@in...>]

have you ever considered to build a generic driver for JSynthLib?

i found the driver creation process too much complicated (i am not a 
master in java)..
generally, the sysex parameters could be described with bitfields.

i.e.: F0 31 42 0[4bits:channel]  [8bits:param] [8bits:value] F7

make a synth driver by creating a text file that describes it, would be 
much simpler.
isn't it?

Re: [Jsynthlib-devel] How to write requestPatchDump?

[Joe E. <jo...@em...>]

Bill Zwicky wrote:

> Single driver is now fully functional for Casio CZ-101 and 1000.  So 
> .. what do I do with the code?  I have a SourceForge account, but do I 
> need to beg permission for write access?

Developer access isn't that difficult to get (after all, *I* got it). 
Any of the admins (Brian, Hiroo, Torsten, RdbRob, and.... Garret, I 
think) should be able to add you. I guess they just need to know your 
sourceforge ID.

And... don't forget... do an *update* right before you commit.

>
> Meanwhile, I discovered that core.Driver, AlesisDM5SgSetEditor, and 
> XMLDriverImplementation all send NOTE_ON twice when told to play a 
> patch.  Shouldn't one of those be an OFF?

I guess so. However, I think that the MIDI spec actually says that a 
NOTE_ON with velocity of 0 is supposed to be considered a NOTE_OFF.

Re: [Jsynthlib-devel] How to write requestPatchDump?

[Bill Z. <wrz...@po...>]

All fixed!  Thanks for your help!  ;)

Single driver is now fully functional for Casio CZ-101 and 1000.  So .. 
what do I do with the code?  I have a SourceForge account, but do I need 
to beg permission for write access?

Meanwhile, I discovered that core.Driver, AlesisDM5SgSetEditor, and 
XMLDriverImplementation all send NOTE_ON twice when told to play a 
patch.  Shouldn't one of those be an OFF?

-Bill

Re: [Jsynthlib-devel] No-frills sysex device/driver/editor?

[Steven S. <ste...@co...>]

Joe Emenaker wrote:

> Brian wrote:
>
>> Joe Emenaker wrote:
>>
>>> Is there a device/driver/editor for JSL that is just bare-bones?
>>
>>
>> I've recently looked into seeing if there are any GPL hex-editor 
>> controls for Java. Though eclipse has a hex editor plugin, the code 
>> is quite large and most likely is pretty integrated into eclipse 
>> utilities. 
>
I just installed EHEP for eclipse.  It looks like it does what you're 
asking for.

Re: [Jsynthlib-devel] No-frills sysex device/driver/editor?

[Joe E. <jo...@em...>]

Brian wrote:

> Joe Emenaker wrote:
>
>> Is there a device/driver/editor for JSL that is just bare-bones?
>
> I've recently looked into seeing if there are any GPL hex-editor 
> controls for Java. Though eclipse has a hex editor plugin, the code is 
> quite large and most likely is pretty integrated into eclipse utilities.

Well, I don't really care about *editing* as much as just *seeing*. 
Something that just broke the data into your standard hex-editor *style* 
(like 16 hex bytes, separated by spaces, and then the ascii equivalents 
on each line). That way, there's a generic item that people can use to, 
at least, make sure that their code to fetch patches is even working 
correctly.

The one time I every tried making a Device/Driver for JSL, it didn't 
work at all... and I had no clue at what stage the failure was. Before I 
try making another, it would be nice if I knew that I could proceed with 
"baby steps" so that I could make sure that each piece was working 
before moving on.

- Joe

Re: [Jsynthlib-devel] BankDriver requests

[Steven S. <ste...@co...>]

Steven Schmidt wrote:

> Also, my (Korg Triton) device and single driver are working.  Should I 
> check them in or wait till the editor is done?

Whoops - spoke to soon.  I can request any patch from the synth and save 
it in a patchlib, but when I try to store a patch from a patchlib to the 
synth, it grabs the GM patch (bank G) at patchNum + 1 and saves it to 
the bank I specified.  If I then try to play the patchlib patch, it 
plays the GM patch.  What's up with that?  I'm hoping this is familiar 
to someone with some experience at writing synth support who can show me 
the error of my ways.