cheater

You need to: - know how to solder - get your pcbs (probably from SmashTV) - get your parts (probably from SmashTV) - get your PIC programmed (probably from SmashTV) - get a case for the Midibox - build it - get a panel for the thing (the shaeffer panel service will make a nice one for you, if you want stuff to look good) don't get too much into the specifics of the panel. you'll probably want to change stuff around once it's ready, so it's best to get a pretty universal configuration. i'm looking into ableton myself. i'd suggest making a panel of vertical stripes. each stripe to have a 10cm fader, and 3-4 knobs, each with a button. you could make them into encoders, actually. try both - make one of the rows out of encoders and not knobs, or something. maybe add a fader bank on the right somewhere so it's neat and looks clean. we all know designing the perfect box is the greatest dream. truth is though that it's not about the box... once you start using it, you'll forget whether the knob is red or black :)

All midi cables at the studio I work in that I haven't made myself have all 5 pins connected! Meaning 5 wires!!

Well... to be honest, I just felt SmashTV's reply was one of those that fuel the fire and let it burn. Oh well, let's leave that. I still love SmashTV :P I never justified that! I just explained it, and made it look less bad ;) Yeah, let's just get back on the topic. So, who's in for deluxe versions of modules? We all know the SID can sound better with a little bit of 12AX7 lovin'... *cough* :P ;) Seriously though: there must be dozens of small mods we all could apply to certain modules, especially since they have been out for quite some time. Perhaps we could have the bypass caps, perhaps we could point out which places would profit from 2% resistors, or audio-grade caps, or a new output stage for the sound chip, be it SID or OPL3 (SmashTV said he's working on one for the SID - Smash: you said you wanted cheap transistors. Perhaps make a costy transistor version too? :) ). Perhaps some better voltage regulators for less noise on the analog lines. Or what ever, really!

SmashTV: man, you're childish. I expected more from you than to toss poop like monkeys do at other monkeys. Seriously - i think Thorsten can handle his problems on his own. He doesn't need support from others, he's not dumb or disabled. Starting flamewars like you have contributed to is the worst bane of all communities. Let me explain KD's tone a bit. Most technical/diy communities work on the basis of knowledge, not community. People come and go, without being remembered, it's their ideas that count in the community. Thus, when someone is pretty knowledgable he's fully allowed to root out from the conversation people who don't really understand the topic. Because, after all, can they add anything to the conversation? No. If you want to take part, go and learn what's needed - and go back. This sounds harsh here, I know, but look at it this way: you're sitting in a university hall, a lecture is going on. You have missed the past 10 lectures and are asking the lecturer over and over what this-and-that means. Would you rather suggest everybody did that (thus making the lecture impossible) or would you rather suggest that people who don't understand the lecture go back to the books, and let the ones that DO understand PROFIT from the lecture? You can't expect people to teach you what you can read in the books. Nobody's going to lose all their time for teaching other people - we need some fruitful conversation here as well to get anywhere at all with the project. Thorsten won't go on with this stuff on his own forever you know. KD: as you have seen, Thorsten, I and some other people have welcomed your contribution to our knowledge base. You must understand, though, that the customs of most DIY communities aren't known or aren't being applied here - so the tone that's completely normal in most mailing lists/news groups/forums that you take part in might seem harsh or offending here. Don't worry, this place is not idiot-ridden: if you ask someone (politely, please :) ) not to take part in a conversation, or if you suggest which topics they could read up on to understand the problem - they WILL do it. If they won't, we'll take care of them, don't you worry ;) But people here are kind enough to understand those things. You don't have to shoot people down at the first hit :) thus, your tone can be a bit more "rounded" and less "edgy" than in most boards - without fear of people bugging you all the time and making the conversations impossible and unfruitable.

Well, I do think KD used some harsh words towards Thorsten here. That's what sniffing rosin fumes does to you after 20 years ;) KD: I'm familiar with synth-diy. This is NOT synth-diy. Behave :) But only a fool looks at the way things are said, and not at the things that are said. I suppose if KD didn't want to help us all he wouldn't have even replied. Well, as I see the situation, it is: 1. KD proposed some changes and TK liked them 2. TK can't apply them because he doesn't own software good enough 3. KD owns software good enough So, KD, if you want all of this done - diy dude! Really: a few boards won't take that much of your time. If you want to help this project, and I guess you're the only one here with professional CAD software, DO send us some PCB designs :) On a side-note: you seem pretty knowledgable with audio stuff. Perhaps take a look at the SID output stage to make it sound more phat? :) SmashTV mentioned something about matching output transistors to the SID... idea: perhaps make two versions of some boards - one that'd be "cost-effective" without all the nice stuff and one that'd be "effective" with matched transistors, HI-FI grade capacitors for audio, extra circuits and whatnot...... that could let people get their hands on a working midibox fast, and it'd move the envelope higher in terms of what's possible - you could build a midibox to get working and uprgade it part by part to get the ULTIMATE sound :) In the end you get a pro-quality synth, but the fun starts much earlier than if you'd have to build it all at once.

my idea: 1. take square piece of styrofoam. make sure to have a second one that's just like that. 2. take chip and put it on its back on the styrofoam. push it in until it's totally in the styrofoam and the legs are level with the styrofoam 3. take the chip out and put a drop of glue where it was. put chip back in and let it stick 4. take electronics pins or whatever and space them out around the chip by pushing them through the styrofoam. glue them in that position. make sure there's at least 2 mm of the pin standing out on the chip side and at least 5 mm on the other side, for soldering. 5. connect the chip to the bolts with nice copper wire (single-strand! much more precise to work with. get a thick one, one strand at least 0.5 mm) 6. make sure all the connections are working great - this is your last chance to fix stuff! 7. cover the whole chip side with glue, and one side of the spare piece of styrofoam. connect and let it dry off i think the best glue for this is the kind that's two-ingredient and turns out rubbery, not hard and resinous. what you get is a nice, indestructible chip that fits on a proto board. pros are that it's pretty stable and that you can't possibly burn it. unless you're less than a newbie ;) cons are that the pinout is kinda reversed (unless you make the pins go through the other piece of styrofoam instead - but that'd make it harder to have the pins stick in a stable way) and that you can't open it anymore. sorry, no pictures!

[glow=limegreen,2,300]They are coming[/glow]

I'm sorry, my reply was a bit harsh indeed. Didn't mean to discourage you at all, I was too trigger happy. I guess SmashTV's harshness wore off on me a bit ;):P 32 oscillators in this chip? Interesting! I wanted to experiment with something like that (load of oscillators) myself - though in analog Make a very, extremely cheap design of an analog oscillator (don't care about the waveforms - cool if you could select at least tri/pw though), that is stable and low-noise. Each oscillator should have separate frequency adjustments. Wonder if it'd be easy (or more importantly: cheap) to make a switch for such an oscillator to turn it into an LFO? Important part: use some 60 of those oscillators :D Make a very crude envelope (maybe just AR), one for each of the oscillators. Throw in 5 cool and complicated envelopes that you could apply to any of the oscillators separately. Perhaps add one analog multimode filter assignable to each oscillator separately (switchable on/bypass for each osc) Combine all that and you got one voice! Cheap VCOs are crucial here. Perhaps this could be realised with this chip? I wonder.... 32 oscs/19 euro = 1.6 euro per oscillator. Not that bad!

Hey tgp, the touch sensitive buttons could be interesting for drum triggers, or poly aftertouch keyboards. Wonder if you could take a normal PC keyboard and fit it with those things - could be interesting as a drum trigger thingie, coupled with the midibox. I know Cubase can scroll the project at different speeds - don't know about Logic. Perhaps it would be possible to use those buttons for this? The harder you press, the faster you wind through the song. Question of my own: does the LC eat up one whole interface? Or just a MIDI channel? Is it really impossible to remap the controller in Protools? (that'd be a downer for me, I was thinking of using it at the studio sometime in the future...)

Hey Ragnar, read the FAQ first: http://www.avishowtech.com/midibox/wiki/index.php/HomePage most notably: http://www.avishowtech.com/midibox/wiki/index.php/Question%20and%20Answers#QA104 (yes, I added it just now. :P) Just wondering - did you find out about the FAQ after you asked your question (like I have)? :)

Hey guys, sorry if you misunderstood what I meant :(. Like Thorsten's, my time is limited as well. So I decided to discuss some ideas before spending 10 years on implementing one and, when it's ready, not being happy with the results. The answer I expected was more in the realms of "yeah, go with that, report back" or "been there, done that, and it sucks". I understand the forum is called "Design Concepts" and not "TK's todo" - and I never said I expected anyone to do it all for me! :o SmashTV: I have built the SID, and it's pretty nice - but it still sounds like many other synths. That's why I'm thinking of ways to spiff it up and make it more "personal" - that's the spirit of the midibox project, isn't it? :) Now I'm getting around the difficult task of learning assembly programming, and drowning in code that's already there :)

OK, another one of my bright ;) ideas I was thinking, if you could route an analog input around the mod matrix, I could just connect some nasty analog LFO and i'd be satisfied with my LFO frequency problems. But then I thought: what about the WTs? So here's the idea: how about hot-patchable digital input, that'd work just like CV input in modular synths... but in digital? The signals run around the MB SID in digital already, so how about just bringing it out? Also, it'd spare you some DACs + quantization errors + hassle for when you got a digital source and want to connect it to a digital thingie like the MB SID. You could make simple external toys, modules, with the PIC and make modular DIGITAL synths. That'd be a first :] Imagine: the MIDIBOX Modular Digital Synthesizer! 8) I can already think of all the craaaaaaaazy digital logic effects you could route the digital CV info through. Also, that's pushing it, but how about making "CV" and "audio" digital stuff the same thing? Both are waves... both are cool stuff. You could take an oscillator and FM another one with it, without any conversion needed. That's just the most basic idea that pops into my mind, out of a gazillion others. I believe some analog modulars had audio signals instead of CV, for simplicity of use. Could S/PDIF be used for this? (just an idea) I'd imagine knowledgable electronics people would think of a much easier thing to use, though :)

http://www.avishowtech.com/midibox/wiki/index.php/MIDIboxSIDFAQ Hope the last "question" isn't too off-topic. I *do* know it's taken damn long. But when your friends build a recording studio and let you in, what else can you do than to get busy? :P TK: I still think LFOs sweepable up to the roof would be great. And longer, sweepable WTs (hell, just listen to the Virus TI on the NAMM demo tape :) ). I know I'm stubborn :p

Hey TK, I'd use the Midibox Wiki... if I knew how to register? o_0

Hey Thorsten, Frustrated? I got over that after my first year of learning mathematics ;) That's exacty what I wanted :) If you say so... :) I still want more knobs :D We'll see how it works out when/if I make an MB FM :P Placeholder/skeleton done (for now) - sorry for the wait, being dragged around by real-lifeTM http://www.salsa.pl/~cheater/MB_FAQ/ Coming up next: I'll distribute this thread around the FAQ. If someone wishes to contribute: use the info at the top of the page :) This FAQ won't exist without other people's contributions. I am solely the maintainer, not one of the big heads ;) Man, it took me like an hour to do that in wordpad on my mom's PC (my own is broken) :) can't wait to get my 'puter working again over the weekend...

Seriously? It'll always be just for fun. Sure, it's cool, but trying to make a stable performance on something like that could drive you insane.

Duh.. what I meant the whole time with "Saw" was "square" :-X I gotta get my synth back together, it's been definitely too long without it! I guess a saw slowly deteriorating into a square wave would still be something to look at, at least for me ;) I'll try to mod the code once I'm ready :) Regarding the SysEx dump: wouldn't making the synth send two dumps, the old-type one and a new-type one (with new extra parameters), solve the compatibility issue? Also, people could choose to just store the "old" stuff in banksticks if they don't need the extra features for a certain sound :) This could e.g. let you store values for external effects modules in the future - right in the memory - very powerful imo. Regarding the FM - since it'll have similar abilities to my synth, I'll tell you what my pains with it are - or rather, what I'd like to be able to change in real time. The most "realtime" way to change parameters in the DX11 is "from the next note", and sometimes I can change some stuff in real time... but it works in odd ways. Is that a hardware limitation, or can you do better with the MB FM? Of course, I hate how the prototype has only 4 knobs! Someone make a hands-on DX synth already ffs ;) My idea would be to have the control surface have a dedicated row of buttons on top (big ones!) for quick algo select, and "hold" (tap on off) to change things and then apply after you tap again, e.g. to make a direct transition to a different wave for an op, without having to sweep when you want a drastic change in a very busy sequence from the MB SID :) under that 3 assignable rows of 5 knobs +1 button each on the left + 1 button on the right in each row. On the top I'd have a list of buttons dedicated for each feature I'll list later on, each with 3 leds under it. Press the button on the right of a row and a button in the list and a led lights up and the synth selects that function for the row. Maybe position the rows and the list selection in such a way that you could press both buttons with the fingers of one hand. Why three rows? That's simple - you could (and you will!) want to change an op's frequency while changing its volume, then quickly change an lfo's speed. Or something else of this kind. No time to surf the menues. Sure, you could do that on the MB 64 - but why have two cumbersome boxes, where one is totally "useless" on your desk and the other is big, cumbersome and not even labeled, bar alone customized for the application? Why have two boxes when you can have one? :) Rant mode off ;) The extra button on the left would do something special to do with the function. The list, for me, would go: --volume - 4 knobs for 4 operators. fifth one for feedback. --adsr - 4 knobs for adsr. fifth knob selects which operator. --waveform - 4 knobs for 4 ops. fifth one for fine or porta, you select. --lfo1 - phase, rate, pitch mod (i guess that's what pitch d. is? d. for "drive" perhaps?), lfo2 mod, waveform, extra button for sync --lfo2 - same --multiplier (is that operator mixing level?) - fifth one does transpose. Transposing the whole voice during a sequence is great! --osc freq - in my DX 11 you can set the freq of an op to a set value or to keytrack. A set freq is useful as modulation, for growling basses (try 4->3->2->1 4,3,2 set very low on sine). --osc freq fine - for that great great unison feel :) --wt - do what ever you want :) --user1 --user2 --user3 The "hold" button could double as a "shift" button - for easy access to menues, or for using the "list" buttons with shift to e.g. quickly turn on unison or the arpeg:) Also, I'd make it readily usable with extra aout modulation things in the patch. For people who'd want e.g. a moog filter to be an integral part of their synth, and want their patches to reflect that. Being able to store info on full modulation of two or three external effects would be great. That's some incompatibility, I know, but to hell with that... :P It'd be much much more betta this way! I'm sure a lot of people would use this feature.

Ja gut.. ist aber immer nur noch ein SID Chip, ob es einen TR909 Interface hat oder nicht :D Der MB SID kann das Alles und noch viel, viel mehr.. und der MB SEQ V2 ist auch viel besser :) Du kriegst 303/909 klangen NICHT aus'm prophet64 :P Jedenfalls nicht solche wie aus dem MB SID :)

Hey Thorsten, thanks for the reply again! Thanks for the code, It's very much appreciated :) I'll experiment with the parallel SIDs, maybe they could sound nicer after all. Good idea with the two types of SIDs. Yeah, I know the works of detuning and slight mis-shaping of things in analog synths :) Lot of experience with VST instruments... Also, post processing is what lots of people forget about while creating sounds. Sometimes the right sound is done more with the effects than the synth itself... BTW: have you ever tried any vintage fuzz for leads? A friend of mine is going to build a Big Muff for my Yamaha DX11, that should ownerize a bit :) Good tip for all synth players: since my DX11 has NO way of interfacing external effects in a real way, I'm gonna use the MB SID with AOUT: I'll make it listen to notes, and then put out the envelope to drive the external fuzz/filter/what ever. SysEx dump: I thought SysEx didn't put a lenght restriction on those things? ??? Oh well - I guess you'd have to make a completely new data structure for what I'm talking about. Too much work I guess :) But the possibility of another 127 knobs to turn sounds tempting... at least to me ;) Hey, What I mean is that the Saw/Pulse waveform is much less detailed, and you could probably make the LFO go to much higher frequencies with it sounding well. Why not make one LFO wave (or mode for all the waves) that'll go up to the max of 1220 Hz, without minding the quality of the wave? Often it's cool to just have that thing to AM the sound into a gritty buzz... yeah, you can tell I listen to goa trance ;) You complain about the oscillators at high frequencies sounding digitally, but I bet many people don't care about that... or even are looking for that! :) FAQ coming up, but it's monday 1 AM so not right now ;D

What's a scribble strip? :)

Hey Thorsten! Thanks for the reply. As always, it's quality reading :) What offense? I'm not the kind of person to take offense when someone answers my question, especially as well as you do it :) Interesting stuff with the AOUT, it sounds like the 8 params won't get old for a very long time :) The 1220Hz wavetable thing sounds just crazy! Indeed, after you explained it, the modulation matrix seems much more complex than I thought at first. No 128 patches yet, but I'll compose this thread into a FAQ and put it on my webpage for everyone. Definitely never enough FAQs. Will keep dumb people like me from posting :P A few small questions, though: Keytracking - what I meant was more about tracking oscillators from midi channel 1 notes, filters from midi channel 2 notes. Is that possible? (as in: will I be able to add it to the code?) You often mentioned the topic of not enough CC parameters. You probably thought of that, but you haven't said so yet ;) - using CCs from two midi channels? The less useful features could be on channel 2, and that could be turned off.. for compatibility or anything. Regarding the multiple SIDs: is a second engine really necessary? What I was asking was if it was simply possible to make two (or more) SIDs listen to the same control data (coming from the PIC) at once - or am I just too newb (yes) with electronics to know that obviously that's more complicated than that and that the PIC needs to react to the SIDs in some sophisticated ways? :) LFOs: wouldn't a saw/PW sound just fine even at low sample frequency? Sweeping resonance: indeed the SID's filter is just 12 dB/oct (and a very... subtly sounding one at times if you ask me).. maybe it would work well with the parallel SID idea I had in mind, where you'd sweep 4 or 8 or what-ever filters at once.. but since you say it won't work, oh well :)

Hey Justin, what I meant was note tracking. Hence the two keyboard idea. The majority of midi controller keyboards can't translate keypresses to another CC. I guess it would go off-tune anyways. moebius: you're right, I'm over-enthusiastic. I tried answering these questions myself, searched the forum etc - I shall try harder and post the answers I find in this topic. Over the weekend.

Hey all, I've gotten around the SID a bit and (as usual of myself) instantly thought of a thousand and one ideas to make it better. The question is how hard these would be to implement, and if that's possible at all :) And of course - what you guys think of these. Okay, here's the meat, beware: 1. controlling the filter separately from the OSCs - or is that possible already? Imagine playing on your keyboard on midi chan1 and having the filter controlled by a stepseq (MB SEQ :) ) or another keyboard... 2. using aout to output LFOs, WTs, ENVs, keytracking info: for controlling external filters (MBSID->moog!) and stuff 3. Thorsten, am I crazy? using multiple SIDs in parallel for one voice (yes, I know you can unison them already, but...) - can it be easy to use one core to control for example 8 SIDs at once, with all of them getting the same modulation info (that's the tradeoff)? Imagine the evil leads this would create. You would have to be able to turn the extra ones off, of course.. Also, you could reuse the filters in a cool way, just imagine: SID1 Â Â SID2 Â Â SID3 Â Â SID4 Â | Â Â Â Â Â | Â Â Â Â | Â Â Â Â | filter1->filter2->filter3->filter4-> out I know I've never seen a setup like this in a synth. Wonder how it'd sound... extremely interesting :) 4. easy one - I know I'm going to do this: take the C64 version C and put a patchbay on the grid in the rear. I mean, isn't that place just MADE for a patchbay? ;D You could e.g. take SID1's out and put it through SID2, or an analog filter built into the patchbay. Could 0wn. 5. more complicated ENVs - look at the Alesis Andromeda A6 for example. Don't tell me the PIC can't do a few attacks or decays more! :) 6. more a wish than plan until I learn how to h4x0r MIOS - more WTs. More powerful, much much longer WTs: we all know that updating a SID every 50th of a millisecond is where it's at. People say that's one of those cool forgotten tricks from the day back, and judging from what FM does (and that's real near that) I do believe them :D Sweepable WTs. For the groove sequencer feel. I do feel the pain of the PICs here - another core would probably be needed for that WT madness, innit? You could have some WTs pre-made for getting some more sophisticated OSC waveforms. That could be real cool. 7. is OSC FM doable with SIDs at all? :-[ 8. LFOs that go only 46 Hz? I need to go up to 500 Hz >: ) Are the LFOs in the SID just software, or do they go analog somewhere along the way? (strong guess: not analog) 9. LFOs: more, more, more waves! :) 10. Mod matrix: as much a wish as a question. Are the modulations just on/off-able or can you set the depth for each crossing? 11. Control surface: morphs. But you'll probably tell me to just use a MB 64 and you'll be right :) 12. Mod matrix: more sources (keytrack, velocity, aftertouch, crap like that) - I guess the diode matrix isn't end-all and some (all?) of those are already in. A cool mod source would be an analog input or three - e.g. for doing FM Â :):):):) or using an external LFO or hooking up to a modular or anything! 13. Mod matrix: more things to modulate. "Filter" probably means cutoff - I'd love resonance in there as well! Perhaps you could put the aout destinations there as well (see point 2) 14. Are more SID modules at once possible? 8 cores + modules? edit: broken ascii ::) moved point 13 to 14, added new point 13.

Read up on slaving cores - like it's done in the MB SID. Oh, and make sure to report back with any good motorfaders you find out about :)

Thanks! One last question: how would you compare that to VFD displays? They seem much brighter, but also somewhat uneasy to the eye. Also, the text seems to be somewhat unsharp, but I can't make sure statements - the only contact with VFDs I've had is a 70s calculator :) The rest is based on photos.. but photos are very misleading! I'm wondering, since there are some VFDs in Germany (I live in Poland) priced about the same. Still, I wouldn't want to buy something that'd prove useless - especially since my money is extremely limited :) Also, another concern is that VFDs are from glass - they might break! :-[