TK.

Confirmed that this is an imperfection which will be (hopefully) fixed sooner or later. Best Regards, Thorsten.

They won't be deleted, which means that they could still be taken over if enough tracks are selected during a PASTE operation. Best Regards, Thorsten.

You are right, I haven't considered this (relatively new) feature which was actually implemented for another purpose. It should work! :) Best Regards, Thorsten.

It's still on the wishlist, but not implemented. See also: (you are the only guy who requested this yet) Could you please check if this also happens if your Akai device is clocked from a PC via USB MIDI? Best Regards, Thorsten.

yes Best Regards, Thorsten.

Yes, 16x16 matrices can be handled in C w/o problems, especially because the actual DOUT transfers are handled by MIOS8 in background with an optimized assembly based routine. The mechanism is always the same: SR_Service_Prepare() will be called before the SRIO scan (each mS) In this function, increment a counter variable on each cycle, reset to 0 once it reaches 16 Transfer the new pattern (which corresponds to the counter and is ideally stored in an array) via MIOS_DOUT_SRSet Please note: a 16x16 LED matrix means that the 16 LED rows will only be visible for 6.25% of the time - the brightness won't be optimal! It's better to implement two 8x16 matrices instead (counter from 0..7) Best Regards, Thorsten.

I agree with the ALT button behaviour change, although it will lead to the minor issue that BLM Emulation users who use a mouse to activate the buttons won't be able to access ALT+<function-key> anymore, since a mouse can only select a single button. However, sooner or later I will add a special key for this (like I did for the SHIFT button, which can be activated with the space key) Drum instrument selection is now supported in Grid mode. http://www.ucapps.de/mios32/midibox_seq_v4_087_pre3.zip Best Regards, Thorsten.

Here a new beta version with enhanced BLM support as discussed above + the ALL button fix: http://www.ucapps.de/mios32/midibox_seq_v4_087_pre2.zip I also created a new version of the BLM emulation which shows the new layout (with ALT and UpperTrk button): http://www.ucapps.de/midibox_seq_manual_blm.html Best Regards, Thorsten.

Of course, it's possible to create a LED matrix with a PIC18F in C. Unfortunately there is no simple example available, only complex examples like this one: http://svnmios.midibox.org/listing.php?repname=svn.mios&path=%2Ftrunk%2Fapps%2Fcontrollers%2Fmidibox_mm%2F&# (see SR_Service_Prepare() function in src/main.c and the appr. code in src/mm_vpot.c) Best Regards, Thorsten.

Well, it seems that the SysEx communication is working. It's normal that no dumps are received for a bank (because no BankStick is connected) or patches (because your MIDIbox isn't configured for patch mode). If the MIDIbox editor doesn't work at your side, unfortunately I can't change this, because this is a quite old software which I haven't implemented by myself. There is an alternative possibility: you could generate a .syx file with the mk_syx.pl script which is part of the installation package (-> see mk_syx/README.txt) It requires a perl installation on your PC, available here: http://aspn.activestate.com/ASPN/Downloads/ActivePerl/index/ The .syx file can then be uploaded with MIOS Studio, you've already proven that this communication is working, so that I'm sure that this will lead to a successful re-configuration which is just a little bit less comfortable than the GUI :) Best Regards, Thorsten.

This makes sense, I added the request to the wish list. Best Regards, Thorsten.

Good ideas, added to wishlist, I prefer option 1 Best Regards, Thorsten.

Request 1 and 2 can be easily implemented, for Request 3 I've doubts that people will really spend the time to organize the SYSEX directory this way - I will wait for more people who like this approach (or who have alternative ideas). Best Regards, Thorsten.

Yes, this was due a conflict with the step recording feature. On the other hand it's probably a good idea to allow selection mask changes as long as the ALL button is pressed, I will consider this in the next release. Yes and no - so: Yo ;-) The ramp is a special implementation that I did for the EDIT page. I would have to write dedicated code for other pages as well to make this happen, on the other hand I wouldn't do this without knowing an attractive use case. Best Regards, Thorsten.

Also doch der LM317... ich dachte eigentlich, der waere "unkaputtbar" Das ist so gedacht. Wenn Du nur einen einzigen Fader bewegen moechtest, koenntest Du dies auch mit dem "Direct Move" Wert veranlassen. Gruss, Thorsten.

That's strange indeed! Did you build a PIC16F or PIC18F based MIDIbox64? And does the Query button in MIOS Studio detect your MIDIbox? Best Regards, Thorsten.

No, there is no MIDI Learn function for SysEx, and I'm not planning to add this. Best Regards, Thorsten.

Hi Kike, are you using your MIDIbox together with a LCD and the 4 menu navigation buttons? Then you could just assign different MIDI events, e.g. via MIDI learn. Or are you working under Windows? Then you could configure the MIDIbox remotely with the MIDIbox64 Editor: http://www.ucapps.de/midibox64.html Best Regards, Thorsten.

Thank's a lot for starting this initiative! Your articles are very nice to read and contain many details which might not be obvious for many people who haven't followed the development progress in the last years. I hope that other users will contribute as well :) Best Regards, Thorsten.

I agree that these functions will be useful; let's collect more ideas before I implement the changes. To summarize from my point of view: - if 16x8 layout is selected, Button 3 should toggle between track 1-8/9-16 and indicate the current selected track range with the LED colour. - a button to select alternative functions will be necessary. I will call it ALT (instead of SHIFT, because SHIFT is used for other purposes). - ALT available at the lowest button (16x8 layout: Button 8, 16x16 layout: Button 16) - Users of the original BLM16x16+X hardware have to press SHIFT and ALT at once to get access over the functions. Users of the emulation have direct access to the ALT button. In GRID and KEYBOARD mode the ALT button will allow to select the octave with button 1-7. Octave from C-0, C-1, C-2, C-3, C-4, C-5, C-6 (C-3 by default) 16x16 layout: all octaves can be selected, not only 7 In TRACKS and PATTERNS mode the ALT functions are not allocated yet. Feel free to request features for the "free slots". Best Regards, Thorsten.

Great! :) ok, I added this info to the webpage: apt-get install libfreetype6-dev:i386 libasound2-dev:i386 libasound2-plugins:i386 libgl1-mesa-glx:i386 libgl1-mesa-dev:i386 Best Regards, Thorsten.

This is strange, something is different in your toolchain setup - I never had the situation that close() isn't available under Linux, maybe google will help you? I'm using "gcc (Ubuntu/Linaro 4.6.3-1ubuntu5) 4.6.3" However, I think that the change was successful: now the access to midiInQueue is protected with a lock. Without the lock I can reproduce the error, with the lock messages won't get lost. Here the prebuilt binary: http://www.ucapps.de/midibox_blm/MIDIbox_BLM_1_2.tar.gz Best Regards, Thorsten.

Update: I remember that I had a similar issue in MIOS Studio and solved this with a ScopeLock - I will keep you updated if it helps. Best Regards, Thorsten.

Ok, these are no good news. :-( As you can see in the diff, for Linux I'm calling the MIDI parser (which is doing some GUI updates) directly whenever new MIDI messages are received instead of putting the messages into a queue for later processing from a timer callback: http://svnmios.midibox.org/comp.php?repname=svn.mios32&compare[]=%2Ftrunk%2Ftools%2Fblm_scalar_emulation%2Fjuce%2FSource@2056&compare[]=%2Ftrunk%2Ftools%2Fblm_scalar_emulation%2Fjuce%2FSource@2058 This measure was necessary for MacOS, and it seems that it's also required for Linux (although the app didn't crash at my side... maybe because the timing conditions on my slow laptop are different). On the other hand it seems that the queue doesn't work reliable under Linux - maybe because it isn't thread-safe. I need some time to understand the backgrounds before I can provide a proper solution for Linux. If anybody wants to try it out by himself: the sources are in the repository under tools/blm_scalar_emulation/juce In order to build the application, just change to the Builds/Linux directory and enter "make" Best Regards, Thorsten.

Thanks for testing! :smile: I was able to reproduce the problem, it only happened under Linux and was related to a special method that I added for MacOS. It should be fixed now. The official binaries (MIDIbox_BLM_1_1) are now available under: http://www.ucapps.de/midibox_seq_manual_blm.html The 32bit build is intended, because my good old Linux laptop which I'm using for testing can't run in 64bit mode, and I guess that some other users could have the same problem. A 64bit build won't give us any advantages anyhow. Best Regards, Thorsten.