TK.

I've included some really useful reporting functions into beta5 which will help to find out the root cause. More in the next mail. If you send me the System, Globals, Tracks and TrackInfo report (TrackInfo only for the track which isn't stepped correctly), I probably don't need the files. Thats the intended behaviour. Patterns are not automatically loaded after startup - we always start with the default setup like known from Canon Cameras. ;) The possible USB data rate which can be handled by STM32 is pretty high. It could happen that your host application causes the problem if it receives so many events. E.g., I've a similar issue with MIOS Studio under MacOS when a lot of debugging messages are received. It can take minutes until MIOS Studio displayed everything what MIOS32 has sent within one second. No issue with host applications like Logic Studio... they can handle the data rate without failures. Ok, will be checked, but this can take a bit longer, since I have to open my MBSEQ V3 case to test the BLM with V4 firmware. Maybe a code review helps to find the root cause earlier... Best Regards, Thorsten.

Could you please store the pattern which doesn't play the steps correctly into 1:A1, and thereafter send me following files in a .zip package: MBSEQ_B1.V4, MBSEQ_C.V4 and MBSEQ_G.V4 Before sending, please check if the issue can be reproduced after power-on when pattern 1:A1 is selected again. Best Regards, Thorsten.

I will replace the 128 and 256 steps items by "---" in the length page if not supported by the selected partitioning. (on a graphical screen, we would gray out the items ;)) /Edit: + a warning message will be displayed if 128 or 256 steps selected. No, since the code is implemented in a different way meanwhile. For MBSEQ V3 the available RAM was the limitation, for V4 such limitations are a bit more relaxed, so that such special cases can be covered. It's just only a question of time (and code complexity) to program it properly so that it works under all conditions (e.g. I have to differentiate between "normal" and "drum" tracks...) News: today I implemented a generic list browser - I really like the implementation, it allows to browse through directories (or other lists) very quickly by using the whole screen + GP buttons/encoders and/or Datawheel/SELECT button. Once the SysEx librarian is completely available, I will continue with MIDI file import and export. Best Regards, Thorsten.

Press MENU->EVENT (GP button #2), ensure that a configuration with 64, 128 or 256 steps partitioning is selected (as desired) Press MENU->LENGTH (GP button #6), and either select the length with GP encoder 2/3, or use GP button 14, 15, 16 for quick selection By pressing the "Step View" button you can verify, which step is currently played. The layout of this view depends on the max. number of steps... Thats due to a programming problem (we have to remember which gates were inactive before), but I will try to find a proper solution. I guess that MBSEQ V3 showed the same behaviour? Best Regards, Thorsten.

Yes, it will work with "unsigned char". But: Is only a warning, the project should be built regardless of this message. Are there additional messages? Best Regards, Thorsten.

You can mix them. Whenever an encoder gets a table entry in MIOS_ENC_TABLE, the appr. two pins won't be handled by the DIN driver anymore. But this doesn't exclude, that the remaining 6 pins can still be used for buttons. Best Regards, Thorsten.

With MIOS8 it won't be possible to realize this without doing modifications in MIOS source code: the MIDI parser in MIOS_MPROC_Handler has to be disabled, incoming bytes have to be polled by your application via MIOS_MIDI_RxBufferUsed and MIOS_MIDI_RxBufferGet. Disadvantage: by doing this modification, you won't be able to upload the application via MIDI anymore. Or you would have to change the cables (remove RS232, plug in MIDI cables) and use the first level bootloader after PIC reset to upload code. No fun! The baudrate has to be adapted in MIOS_MIDI_Init, the formulas (and precalculated values for 4800/9600 baud) can be found in the PIC datasheet. MIOS32: easy! Especially since code can still be uploaded via USB MIDI w/o swapping cables :) In your mios32_config.h file, change the UART settings to following values: // Interface assignment: 0 = disabled, 1 = MIDI, 2 = COM #define MIOS32_UART0_ASSIGNMENT 2 #define MIOS32_UART1_ASSIGNMENT 2 // Baudrate of UART first interface #define MIOS32_UART0_BAUDRATE 4800 // Baudrate of UART second interface #define MIOS32_UART1_BAUDRATE 9600 [/code] Thereafter you can use the MIOS32_COM functions to send/receive data: http://www.midibox.org/mios32/manual/group___m_i_o_s32___c_o_m.html thats all! Best Regards, Thorsten.

Are you using the loop points while editing a sequence? In this case, Fx->Loop is the right function. E.g. "Loop in Step View" automatically plays the steps which are displayed on screen. The loop function can be activated under Fx-Loop, or with the shortcut "MENU+Scrub" Or do you want to use different loop zones during a performance? Then just wait for the "Sections" feature, which will work similar to the way how I realized it for MB-808 (see TODO List in the ChangeLog). For A/B/C/D it will be possible to define different step offsets and lengths. And this should be done in the Length page - therefore I'm not happy with your suggestion to put new items into this page (items which have to be removed later)! Best Regards, Thorsten.

Ok, EXIT LED is now active whenever the menu selection page is displayed - this makes more sense anyhow. And for the bling I added a Morph/Mixer/Transpose LED function. Will be available in beta5. Best Regards, Thorsten.

I got the v1.2 PCBs from Ploytec, and prefer to ship these ones in future. The v1.1 PCBs will be held back as spares for the next bulk order (e.g. if more than 100 PCBs are requested) Following orders will be shipped tomorrow: diggi cinhcet mattflo Xem m00dawg afx insectteam Bordes Sébastien horkheimer Best Regards, Thorsten.

Beta4 is available for download. ChangeLog: [tt] o MBSEQ Parameters are controllable via NRPN messages from external MIDI devices now. See also doc/mbseqv4_cc_implementation.txt o these parameters can also be controlled via Internal Loopback (Bus1 Port). o Loopbacked CC Parameters are print in text format when edited o implemented MIDI Remote Keyboard function (note: MBSEQ_HW.V4 has been updated) See also doc/mbseqv4_remote_functions.txt o fixed various minor bugs [/tt] Best Regards, Thorsten.

I never used an EXIT LED by myself... ;) Will be fixed in beta4 (release this evening). Next bug, please. Best Regards, Thorsten.

Press MENU->TRIGGER (GP Button #9) to change the trigger assignments. Note that in distance to MBSEQ V4 we have 8 trigger layers now (A..H) - the 8 available triggers functions are assigned to A..H, accordingly they can all be used in a pattern. Changing the assignments only makes sense if you prefer a different order (e.g. Glide assigned to B instead of D), or if you want to trigger multiple functions with a single trigger (e.g. if Trigger A should activate Glide and Random Velocity) Just change to the Mixer page (MENU->Mixer), and use the datawheel to select the CC page Each page provides 16 CCs, there are 4 pages (accordingly, 64 CCs can be defined) By pressing the SELECT button you can store your mixer configuration in a Mixer Map The required procedure should be described in the user manual of Ableton Live. It works great with Logic :) Best Regards, Thorsten.

Addendum: ca. 15 years ago I experimented with such carbon based pressure sensors: They are quite expensive! Best Regards, Thorsten.

Yes, it's even possible to modulate sequencer parameters via LFO E.g., during this experiment I controlled the amplitude of LFO Track #1 (sine waveform with 8 steps interval) with the LFO of Track #2 (saw waveform with 64 steps interval) Track #1 plays note events + an echo Fx, force to scale has been activated. That's called amplitude modulation :) Best Regards, Thorsten.

The BLM driver is completely implemented, but the inversion mask wasn't changed. I fixed this in beta3 Yes, to mute individual parameter layers. Thats especially useful if a track plays drum instruments - you can mute individual drums. This function is described under User Manual: Menu Pages Haha ;) no, this isn't possible - it would require a *huge* pattern memory to store the parameters for each step. Not only that the appr. SRAM space isn't available... it would also increase the time to change to a new pattern dramatically! However, once the Loopback function does also work for Sequencer Parameters (CCs), you will be able to change up to 16 Fx parameters per track for each individual step! I think that this will really cover your request. Beta3 is available now - from the ChangeLog: [tt] o if global loop mode enabled, a "*LOOPED*" message will flash at the right corner of the EDIT and STEPVIEW screen. o following handling has been added to realize a "smooth startup" for people who are in the progress of building their MBSEQ. The core is already running, they connect a SD Card but don't have LCDs (so that messages cannot be read): - if a SD Card without Banks/Songs/Mixer Maps is connected, the appr. files won't be created automatically anymore. - messages are print on the PATTERN/SONG and MIXER page to inform that files have to be created from the UTILITY->DISK menu - in the DISK->UTILITY menu you will find a special option at the left side which allows to start the formatting process (in fact, the SD Card won't be formatted, but some files will be created) - there is now a nice progress bar which informs about the state so long formatting is in progress o CLEAR button has to be pressed for 2 seconds before action is triggered o fixed BLM inversion mask configuration [/tt] Best Regards, Thorsten.

There is no reason why the length parameter shouldn't work. Could it be that the loop function was activated? (Fx->Loop) There shouldn't be a processing time issue. How can I reproduce this exactly? (describe the configuration steps) Didn't you request the possibility to mute parameter layers? Or who was it? ;) Yes, so long more than 4 parameter layers are configured. I guess you configured it for 8 or 16? If a track only consists of 4 parameter layers, ParLayer C button will toggle between layer C and D Doubleclicks won't really help - I could insert a delay (press button for more than 2 seconds) Best Regards, Thorsten.

Ok, thats a different scenario: the core is able to control all LEDs, correct? In this case the hardware is ok, and it's very likely, that you stored "all-LEDs-on" in the default snapshot. Turn off all LEDs, and press the snapshot button again (for ca. 3 seconds until the store function is triggered) Best Regards, Thorsten.

The MIDIbox Link page describes a concept for linking multiple cores. Sharing the menu interface between multiple cores is described as a possibility, but it hasn't been implemented in any application available at my website. Accordingly this possibility isn't implemented in MIDIbox64, therefore you haven't made anything wrong. Best Regards, Thorsten.

This (extremely simple) application doesn't provide a menu interface. All events are preprogrammed in the C code and can only be changed there. They cannot be changed during runtime. Yes... as mentioned above: no menu interface Where did you found a reference which describes that ain64_din128_dout128 supports a menu? Best Regards, Thorsten.

Protofuse: he uses the MIDIbox64 firmware... No, all LEDs should be off. Are the buttons working? If no: there seems to be a general problem with the SRIO chain, debugging help here: http://www.midibox.org/dokuwiki/doku.php?id=din_module If yes: there seems to be a connection problem between CORE:J8 and DOUTX4:J1 - the srio_interconnection test (as described on the DIN module page) helps to debug this as well. Best Regards, Thorsten.

How does the MIDI Routing in your environment look like? If you route the SUM of all MIDI Inputs to the Sequencer Input (see first snapshot in attachment), the events sent by your MIDI Controller will be sent to the active track as well. This is probably something what you want to prevent. If you separate the MIDI controller by creating a special wire for the appr. input, the MIDI events won't be routed to the MIDI inputs (proposed solution, see second snapshot). Best Regards, Thorsten.

J5 provides Ground, 5V and 8 digital outputs. These digital outputs (0V/5V) are the gates. The clock is available at J6:RC, and the start/stop signal at J6:SC of the core module. No, it's only 0V/5V - you could amplify it to 10V with a transistor circuit. This is also a nice protection against overvoltages - very important for a modular system. Yes, the 24 ppqn MIDI Clock will be output as a 1 mS trigger signal. Best Regards, Thorsten.

MBSID will send the SysEx dumps (and program change events in addition) to slave SIDs whenever a Program Change has been received. If your Alphatrack responses with some unintended Note events this way, a possible solution could be to change the MIDI channel of MBCV, so that Notes on Channel #1 are ignored. The MIDI Channel configuration is part of the patch. It could also make sense to change DEFAULT_PATCH_CHANGE_CHANNEL in the setup_* File of MBCV, so that Program Change events sent from MBSID on Channel #1 are ignored as well. Best Regards, Thorsten.