TK.

It sounds like a power problem with your Hub, are you using an external PSU for it? If not it's time to try this now, because if the hub is only supplied by your PC the total current drain shouldn't exceed 500 mA With an external supply it mainly depends on the specs of your PSU I've connected three GM5 interfaces and two MIOS32 cores (one is a complete MBSEQ V4) to a single hub, the total consumption is more than 1 ampere. If I disconnect the PSU the MBSEQ LCD backlights are dimmed and the devices won't work reliable anymore, regardless if the Hub is connected to a Macbook Pro or to a Dell PC Therefore my assumption that you noticed the same effect. Best Regards, Thorsten.

P.S.: note that I updated the binary with improved pinning for your special case - I hope that you like it, because I won't create another configuration

Hi, I created a special .asm file for your configuration here: http://svnmios.midibox.org/filedetails.php?repname=svn.mios&path=%2Ftrunk%2Fapps%2Fcontrollers%2Fmidibox64e%2Fsetup_midibox16e_without_ledrings.asm It will be automatically part of the next release package. Attached you will find the prebuilt .hex file of the current release. Best Regards, Thorsten. setup_midibox16e_without_ledrings.hex

There are no conflicts at the hardware side, and not that many adaptions have to be done at the software side. The code will look better after the migration, e.g. instead of // send mapped midi-note with 127 or 0 velocity MIOS_MIDI_BeginStream(); // midilink encapsulation header MIOS_MIDI_TxBufferPut((unsigned char) button_event_map[pin][0]); // first byte from table = CHANNEL MIOS_MIDI_TxBufferPut((unsigned char) button_event_map[pin][1]); // second byte from table = NOTE MIOS_MIDI_TxBufferPut(pin_value ? 0x00 : 0x7f); MIOS_MIDI_EndStream(); // midilink encapsulation tail [/code] you can now write: [code] MIOS32_MIDI_SendEvent(DEFAULT, button_event_map[pin][0], button_event_map[pin][1], pin_value ? 0x00 : 0x7f); where DEFAULT is a MIDI port (usually assigned to USB0) of course... Some tutorials can be found here: http://www.ucapps.de/mios32_c.html Recommented reads: #001: Forwarding MIDI Events #002: Parsing MIDI Events #003: Debug Message #004: Sending MIDI #009: Controlling up to 128 LEDs with DOUTX4 Modules #010: Scanning up to 128 buttons connected to DINX4 Modules #012: Scanning up to 64 analog pots (or even more) Best Regards, Thorsten.

This effect is caused by the infamous ADSR bug of the SID. More informations and workaround: go to my website, enter "adsr bug" into the "search ucapps" form Best Regards, Thorsten.

Happy Birthday - have a great day! Best Regards, Thorsten.

MBSEQ V4 can control 16 drum instruments per track - accordingly you will get up to 256 "drum tracks" :) Best Regards, Thorsten.

You only need to change the base address in iic.asm header. With IIC_DEVICE_ADDRESS EQU 0x18 the new device addresses will be 0x18, 0x1a, 0x1c, 0x1e No, it isn't possible to update the firmware via MIDI or IIC Best Regards, Thorsten.

Great! :) Best Regards, Thorsten.

Hallo, ich habe Dir eine PM mit einem Link zur aktuellen MIOS Studio Version geschickt. Wenn das LCD nicht initialisiert wird, ueberpruefe mal, ob Du auch die richtige MIOS Release erwischt hast (sie befindet sich unter pic18f4685/midi/mios_v1_9g_pic18f4685.hex) Die hex2syx.pl Fehlermeldung ("ERROR keine .ini Datei") kommt mir uebrigens etwas seltsam vor - das Script selbst gibt keine deutsche Fehlermeldung aus, und benoetigt auch keine .ini Datei. Wie dem auch sei, mit der aktuellen MIOS Studio Version ist es obsolet Gruss, Thorsten.

But it worked before, correct? I guess that you changed something on the MIOS Studio configuration which causes this effect, somehow it looks like a feedback loop, since messages are received multiple times. It could also be, that a second application is running in background which forwards MIDI events (you are working under MacOS, right?) -> check the MIDI routings. Best Regards, Thorsten.

One possible reason (beside of bad solderings or wrong connections between MBHP_CORE_STM32 and MBHP_DOUTX4 module) is that the 220 Ohm Pull-up resistor array R31 is not mounted. You can test this by measuring the voltage between ground and pin J8:SO, J8:SC, J8:RC when the DOUT module is *not* connected. Note that due to the AC waveforms generated by the SPI transfers you won't measure exact DC values with a common multimeter, but they could give you some hints. E.g., J8:SO should be 0V when your application sends 0 to all DOUT pins, and it should be ca. 5V when you application sends all-1 J8:SC should be between 4V..5V (each mS this pin sends a burst of 1 MHz for ca. 300 uS) J8:RC should be ca. 5V (only short pulses sent each mS) If you always read 0V, the pins are probably floating, accordingly the pull-ups are not connected. Btw: the usage of MIOS32_MIDI_SendDebugMessage() is really recommented for debugging, it's more useful than a LCD. See also http://svnmios.midibox.org/listing.php?repname=svn.mios32&path=%2Ftrunk%2Fapps%2Ftutorials%2F003_debug_messages%2F Best Regards, Thorsten.

In the hope that these pictures are saying more than 100 words ;) Best Regards, Thorsten.

for which MIDIbox projects? post the links please. Best Regards, Thorsten.

Shipped today: Hias lumstar Gerbil phunk Following guys will get their GM5 chips via Nils together with the GM5x5x5 PCBs: frailn vcfool kokiPsiho madox cavey magneto Best Regards, Thorsten.

On a Mac it was really simple, just click on the example Xcode project, edit it, build it - done. On a PC Phil noticed a lot of compatibility issues between gcc and MSVC while trying to compile my code, since I'm sometimes using code constructs which are not supported by the Microsoft compiler, such as variable declarations inside the function body, or casting one union to another in a single line. And there are problems with the latest MinGW based gcc version as well (I will give you access to the programmers lounge where you can read more about the details). However, for somebody like me Juce is still very attractive - without reading any documentation I was able to adapt the AU/VST wrapper example code for my needs within a couple of hours. :) No, it's because the latest Xcode version which was shipped with Snow Leopard uses gcc 4.2, but the MacOS 10.4 SDK isn't compatible to this version. I guess that the Plugin could be compiled successfully with an older version - sooner or later I will "sacrifice" an external HD for a Tiger installation. Best Regards, Thorsten.

Selecting sections with a keyboard is officially supported with the latest version, see the ChangeLog: http://www.ucapps.de/midibox_seq_changelog.html There is even a video which demonstrates the usage. The documentation hasn't been updated yet, therefore you probably missed this. I'm planning to write a tutorial about the usage, because a) this powerful feature can be quickly overlooked, and b) the perfect configuration in conjunction with your gear requires some learning. Thats how phrase mode is working, it can also set mutes, send mixer maps, change tempo, etc... It really makes sense to work with MBSEQ V4 first before proposing new features, which probably already exist or have been implemented on a different way. Best Regards, Thorsten.

In the last weeks I re-implemented the MIDIbox SID sound engines in C, so that it is possible to run MBSID on a STM32 core, but also to emulate it accurately on a Mac or PC. The AU/VST gives you the possibility to try some MBSID patches before building the hardware. The sound isn't so fat like on the original, the filter sounds weak and there are a lot of aliasing effects when higher frequencies are played... but probably you will like it anyhow. The AU/VST is based on Juce, Dag Lem's reSID is used for SID emulation, and the part of my MBSID code is to control the SIDs with the possibilities described in the MIDIbox SID Manual. Download (current version 0.2 from 2014-09-28) AU (for MacOS): http://www.ucapps.de/midibox_sid/virtual/MIDIboxSID.component_v0_2.zip (64bit AU which works with Logic Studio 10) VST (for Windows and Linux): http://www.ucapps.de/midibox_sid/virtual/MIDIboxSID_v0_2.dll.zip Known issues: - only the first preset sound bank with 128 sounds is available. No possibility to access the sound parameters yet. - no access to CCs, NRPNs, SysEx yet This will be added very soon, so that it is possible to automate parameter changes, and to edit complete patches with Rutger's editor (Java), or Nils' editor under Windows - the VST doesn't work with Logic 5.5 under Windows - the AU requires MacOS 10.6 or higher - only 64bit supported anymore - only a single instance can be opened! To support multiple instances it will be required to implement the MBSID code in C++, but this will also increase the memory consumption which is bad for the STM32 build. I will evaluate the separation in classes later. Best Regards, Thorsten. P.S.: big thanks to Philetaylor for building the VST under Windows - it wasn't straightforward!

Again: Sold Out! Let's try a 5th bulk order Best Regards, Thorsten.

Finally... :)

This has to be done by somebody with Windows skills. The instructions to compile the VST are given in the "extras/audio\ plugins/How\ to\ use\ this\ framework.txt" document of the juce library. Best Regards, Thorsten.

Too bad :-( However, for those who are interested, here the AU Plugin which allows you to emulate a MBSID w/o hardware: /edit: deleted link - see separate posting here: It has to be copied into your Library/Audio/Plug-Ins/Components directory (if it doesn't already exist, create it!) Thereafter (re)start your DAW - done :) It's built for MacOS 10.6 (Snow Leopard) - a build for older OS versions is available on request. Note that the high CPU consumption, the weak filter sounds and the aliasing effects are caused by the reSID emulation. Only the real hardware will produce the sounds we really love! :) Best Regards, Thorsten.

Are you interested to implement this in Juce instead of Synthmaker? You could take the plugins of the ctrlr project as inspiration. It would have many advantages, e.g. we could not only compile a VST for Windows, but also for Linux. And we could compile an AU component for MacOS. Background: last weekend I took the code of MBSID V3, combined it with the reSID emulator, and put it into an AU plugin: This plugin allows to emulate a MBSID on a Mac accurately (a VST for Windows could probably be compiled as well). All "Vintage Preset" patches can already be played, even Multi/Bassline/Drum patches! Support for sending SysEx messages to a real MBSID could be easily added, since the appr. informations are already available in the MBSID V3 code. See also the source code as a reference. (the simple Juce editor component is located under juce/src/EditorComponent.cpp) As you can see in the snapshot above, the GUI is currently very minimalist. You could help me in creating the GUI while I could continue with the low-level stuff. :) Best Regards, Thorsten.

Basically it would look like this example: http://www.ucapps.de/mios_c_cc_to_nrpn.html But instead of void MyFun_Send_MC303_NRPN() you would write something like: void MyFun_Send_OldSynth_SysEx( unsigned char evnt0, unsigned char sysex_parameter, unsigned char value) { MIOS_MIDI_TxBufferPut(0xf0); // SysEx Header MIOS_MIDI_TxBufferPut(0x11); // to your MIOS_MIDI_TxBufferPut(0x22); // old Synth MIOS_MIDI_TxBufferPut(0x33); // however it looks like MIOS_MIDI_TxBufferPut(0x44); // too bad that you haven't posted a spec MIOS_MIDI_TxBufferPut(sysex_parameter); // parameter number (specified when calling this function) MIOS_MIDI_TxBufferPut(value); // the CC value MIOS_MIDI_TxBufferPut(0xf7); // end of SysEx (or do you need a checksum before?) } [/code] Best Regards, Thorsten.

Hi, this CC parameter is only sent over eight MIDI channels (B0 5B .. B7 5B) If you are using the MIDIbox64 application with default configuration, they would be sent by the first eight pots multiplexed by IC1 of the first MBHP_AINX4 module. Possible errors: - AIN module not properly powered (check Ground/+5V connections to the 4051) - pots not properly powered (check Ground/+5V connections at outer terminals) - the three selection pins of MBHP_AINX4:J6 are not connected to MBHP_CORE:J6 Best Regards, Thorsten.