TK.

Thats a difficult topic, since sometimes (not always) people tend to ask for special support once they made a donation, or they are very unhappy when I don't pay enough attention for their personal wishes although they gave me some money. Therefore I don't promote acitivties into this direction. On the other hand: if somebody wants to make a more or less anonymous donation to my PayPal account, I don't prevent him from doing this - this could lead to a bigger frontpanel in the next design, or to new supported device/module which I normaly wouldn't use by myself, just to have some fun on the driver implemention. (Example: alternative, cheaper DACs for the AOUT module, Motorpots, special graphical LCDs, etc...) Moebius alludes to an important point: I'm very happy with the situation that in the meantime most people get superb help from you guys without any effort from my site. Since some weeks I'm able to totally focus on MIDIbox FM when I find the time without beeing distracted from permanent help requests for which nobody takes care about. Under these conditions you can be sure that I will continue with releasing new projects to the public :) Best Regards, Thorsten.

Hi Marcel, the required software changes are not so easy compared to the small hardware changes you have to make: just cut the tracks to the buttons/encoders which are affected and re-wire the DIN inputs with small cables. This will possibly cost you one hour of work on the ready-made PCB, but this is much less time than writing down an article about the required changes especially for your problem. Best Regards, Thorsten.

Hi Thomas, this is not a bug, but a feature. The SID volume is a 4-bit value. This means that you won't hear any change when selecting a volume between 0 and 7, 8 and 15, etc... the same is true for the other 4-bit parameters like attack/decay/sustain/release... it's a limitation by the SID so, why shouldn't the CS just display the value which is really used? The 4-bit values are patted to allow the proper use with poor MIDI controllers which don't support min/max values (like NOT the MIDIbox controllers ;-)) and with poor sequencers which always display the 7-bit range in the automation The value is not reduced to 4-bit resolution when editing the parameter directly on the control surface. This was to have the same "feeling" when tweaking different values with different resolutions. If you don't like this, then you are allowed to change the source code, but you will propably switch to the original version as soon as you regognize what I mean ;-) Best Regards, Thorsten.

Hi Matteo, this doesn't match with the programming concept, a lot of changes would be required in order to chain the steps without changing the pattern, and some features would not be possible with the current performance anymore (e.g. accurate song position pointer/rewind/forward). I'm also not sure if the song mode would work in this form anymore. Chaining tracks was possible with MIDIbox V1, but I removed this not without special reasons... Best Regards, Thorsten.

Hi Matteo, no, the AOUT implementation is not final, and it still works like described in the changelog - an even more powerfull implementation is planned for next year, but this should not say, that the current version doesn't make a lot of fun :) Best Regards, Thorsten.

Hi Stefan, there is a very simple way to display a message - just use the "print message on LCD" SysEx command of MIOS (http://www.ucapps.de/mios/mios_sysex_implementation.txt) It's an internal MIOS feature which works also in a running application. If you want to prevent that the message vanishes after 2 seconds, you have to modifiy the USER_DISPLAY_Init hook (e.g. the small skeleton: add a "return" below USER_DISPLAY_Init") Example for a message (use MIDI-Ox to send this command) F0 00 00 7E 40 00 08 01 00 00 4d 79 20 66 69 72 73 74 20 4d 65 73 73 61 67 65 21 20 3a 2d 29 f7 Best Regards, Thorsten.

Hallo Stefan, ich vermute, dass es sich hierbei um einen Programmierfehler von meiner Seite handelt -- meine MB16E hat ausser den LED-kraenzen keine zusaetzlichen LEDs, deshalb habe ich dieses Feature nie ausprobiert. Mit der naechsten Version sollte es funktionieren. Gruss, Thorsten.

Hallo Stefan, was hat sich eigentlich mittlerweile beim Merger-Problem ergeben? (http://www.midibox.org/cgi-bin/yabb/YaBB.cgi?board=german;action=display;num=1096807147) Gruss, Thorsten.

ok, so lets assume that it is a stupid error - maybe the reister addresses declared in app_defines.inc are overlapping the C variables? You can check this in the projects.map file, where all addresses are listed. Best Regards, Thorsten.

Hi Nikolai, I remember that the generic remote of Cubase SX is not so well maintained by Steinberg like before. In particular versions one or two functions are not working, in following versions they are fixed but other functions are failing... :-/ Best Regards, Thorsten.

Hi Tim, could you please add this info to the HowToConnectLCD page of the Wiki? Best Regards, Thorsten.

Smash sent me the S&H chips + other stuff some months ago - the package took very long through the customs (ca. 2-3 weeks), but I haven't payed VAT Best Regards, Thorsten.

Copy&Paste from an email to d2k: Currently my MIDIbox FM looks like MBSID Step A, I will probably use a 2x40 LCD in the final version. Due to the flexible concept the display size doesn't matter during the developing phase. A Master/Slave concept like MBSID is not planned and not required, because the application already provides a multitimbral mode with 5 common MIDI channels (Mono/Legato/WT/Poly) + one drum channel The "instruments" can be assigned to 4 audio outputs. 128 Patches are already available (I converted the GM set from the Linux driver) All OPL3 parameters can be modified Development of software based sound engine (Portamento, 2 LFOs + 1 EG + AOUT + 7-track/32-step wavetable sequencer per channel) not started yet - however, the sounds are already impressing :) Plans for complete control surface not before december, I want to create some own sounds before, in order to find out the best usage concept. Best Regards, Thorsten. Addendum: if somebody cannot resist to build more than one MIDIbox FM, then the additional units can receive premade sounds from the master module. It's similar to MIDIbox SID, but w/o realtime control for the slaves

I don't think that another IC will give you better results, because the propability that they are not working is very very low. It's difficult to give you some additional tips so long you don't describe what you've already done. E.g., you are writing that "there doesn't seem to be a short", but you are not writing how you've ensured this. Therefore its unclear if it makes sense to give you some tips for measuring the module, or if the problem could be somewhere else. Best Regards, Thorsten.

Hi, first question: did you also try to control the filter without the joystick? E.g. with the datawheel from the filter menu, or via CC? I'm asking because the AIN driver for MIDIbox SID is totally unsupported and therefore could have bugs Another question: are both SIDs selected on the control surface? For simultanous changes the SID1 and SID2 LED should lit (press both buttons at the same time to select the two SIDs) Best Regards, Thorsten.

Hi, you could have the same problem like Twin-X had with his DIN module, see http://www.midibox.org/cgi-bin/yabb/YaBB.cgi?board=troubleshooting;action=display;num=1096139044;start=28#28 (Reply #28 ) Don't rotate the ICs, this could destroy them! Yes, the DOUT module works also when only one chip is stuffed. Best Regards, Thorsten.

no, I'm not speaking about softcores, they mostly allocate too much macro cells, or the performance is not good enough. Best Regards, Thorsten.

Hi Pilo, well, what should I say to this statement? I don't know your code ;-) no, because matrices are always very hardware dependent and require customizations which are better located in the application, because parametrized code would be too slow. MIOS already provides a hook which allows to prepare the DOUTs before they are updated. However, currently it is not possible to use this hook in C... I know... this will be added to the C wrapper once I find the time. you can use two seperate DIN/DOUT chains, one for the standard MIOS stuff, one for the button matrix stuff. You only need to change the pinning in sm_fast.inc This would be the most performant solution Best Regards, Thorsten.

An extraordinary MIDIbox SID made by Bill More pictures of his DIY stuff can be found under http://www.mylgd.com/public/?rep=.//Photo/Bill/diy

Hi Seppoman, so long the AIN handler is disabled, 128 bytes are free for the application which could be used as alternative RAM storage (0x380-0x3ff). You could store the modified WT data there. The WT sequencer (sid_wt.inc) has to be changed, so that it loads the step information alternatively from RAM. The CS_MENU_MS_Update_WT_Buffer in cs_menu_ms.inc has to be changed, so that it doesn't store the modified data in the buffer, but directly at the appr. RAM address. The CS_MENU_MS_Update_WT_EEPROM function has to be disabled, and CS_MENU_SaveBank in cs_menu.inc has to be changed so that WT data is transfered from RAM to EEPROM if this function is not used to copy data, but to store a new patch Best Regards, Thorsten.

Bingo! Yes, it's the OPL3 :) I've started with a very basic driver which allows to trigger some notes, but - please be patient - a first version won't be released until end of this year. I need this time to find an optimal user interface and an optimal sound architecture (like Duggle mentioned: even more interesting sounds can be expected once the parameters are modulated via LFOs/ENVs/Wavetables). The prototype is not the final one. The selfmade 14.318MHz oscillator will be replaced by an integrated one, the LM314 by a TL074 for high quality sounds, I will add a second YAC512 for two additional audio channels (the YMF262 provides 4 seperate outputs), and finally I want to put this on a PCB. Hopefully the background noise disappears then (I think so...) Here a new example (the last for this month) - drums can be accessed via MIDI now - the bass is a killer! :-) http://www.midibox.org/midibox_fm/mbfm_demo2.mp3 Pilo: like in the first example the delay/chorus/etc effects are made with external VST effects - sorry, no DSP ;-) Andrew: today I received a AtariST from an ebay auction (for 1 EUR) - looking forward for your YM2149F synth! :) Best Regards, Thorsten. P.S.: please don't start to desolder the SMD chips from old soundcards if you never did this before. I will provide a desoldering guide!

w/o words :-) http://www.midibox.org/midibox_fm/mbfm_demo1.mp3 Best Regards, Thorsten.

Der spezielle Drum Algorithmus in der MB64 arbeitet aehnlich wie von Steff beschrieben (er befindet sich in drums.inc): Allerdings betraegt die Sample Rate lediglich 200 uS, da der AIN Handler von MIOS so ausgelegt ist, dass er die restlichen Tasks moeglichst wenig stoert. Mit einer zusaetzlichen Hardware (einem sog. Integrator) laesst sich das jedoch relaxen, jedoch auf Kosten der Latenz (immerhin noch < 3 mS) Hier noch ein weiterer Link zum Thema EDrums: http://edrum.midibox.org Admir hat sich auf das Thema spezialisiert und hat eine hochoptimierte Firmware programmiert. Gruss, Thorsten.

Die Freeware Version vom Keil C Compiler erlaubt lediglich eine binary-Groesse von maximal 4k, und die werden bereits von der MBHP_USB Firmware belegt. Der AN2131 hat insg. nur 8k, viel mehr ist also auch nicht mehr drinn Gruss, Thorsten.

Hi Dan, without a lot of changes in MIOS this isn't possible. A more flexible solution which allows to select either DIN registers or J5 pins would slow down the performance too much at a very sensitive part of the ISR. Best Regards, Thorsten.