nebula Posted March 25, 2011 Report Share Posted March 25, 2011 The TR-909 sequencer brought a really cool new feature that was missing in the 808 and 606 before it: dynamics. Aside from the global accent, many of the instruments allowed two distinct dynamic levels on each step. To look at how this worked from a user perspective, first let's look at the TR-808: if you have a blank pattern and press one of the step ("GP") buttons, its corresponding LED lights up and you have activated a drum trigger on that step. Press the button again, the LED turns off, and you have turned off the drum trigger on that step. That's two different states: off and on. On the TR-909, the beginning is similar: the LED for a given step is off when there is no trigger programmed. Press the GP button for a given step, and the LED comes on, but at 1/2 brightness. You have activated the trigger, and the note will play on that step, at 1/2 dynamics. Press the GP button again, the LED comes on at full brightness, and the step will trigger with full dynamics. The step will turn off when you press the button a third time. That's three different states: off, half, and full. There has been some discussion of a hypothetical "MB-9090" in which the MIDIbox 808 software is used to control a 9090 board (from Trevor Page) to make a TR-909 clone. It's a great idea, except that AFAIK TR-909 dynamics can't be realized with the current software. I suspect that modifying the MIDIbox 808 software to allow two different dynamic levels on each step is feasible. But the hardware problem of dimming the LEDs to provide visual feedback of the dynamics per step is a problem - the TR-909 uses PWM to dim the LEDs, and it has been stated elsewhere that 8-bit MIOS can't do PWM dimming effectively due to the required CPU cycles. Perhaps a solution to this would be to use 16 additional DOUT pins, which the builder could use to either light the same LEDs at half brightness (via highter-value resistors) or a second set of LEDs to indicate the lower dynamic level per step. The two-level-dynamic feature would need to be something that can be defined, per instrument, in the setup.asm file. Also, the level would need to affect MIDI velocity, and likely a CV as well. Anyway, this is just a concept. I'm not (yet) a MIOS programmer so I suggest this capability with all modesty, as right now I do not have a clue where one would even begin to make something like this a reality. But at this point I think it's important that the idea gets out there, because it is the only thing standing in the way of making the world's first fully featured TR-909 clone. Quote Link to comment Share on other sites More sharing options...
MPN Posted March 26, 2011 Report Share Posted March 26, 2011 Hi. I just want to point out that the TR9090 responds to midi velocity. Also you can control how long the sampled sounds (hihat, crash and ride cymbal) are sustained by using different note length. That’s why I will use the MBSeq v4 to control my TR9090. Best regards Magnus Quote Link to comment Share on other sites More sharing options...
Altitude Posted March 27, 2011 Report Share Posted March 27, 2011 Not sure haw that would work on the MB808 since the accent is global for all voices. Like MPN stated already, it is already included in the 9090 and you can already do this with the MB808 sequencer is velocity layers for controlling the 9090. It would definitely be cool to do it 909 style though in the MB808 Seq Quote Link to comment Share on other sites More sharing options...
nebula Posted March 27, 2011 Author Report Share Posted March 27, 2011 The TR-909 still has a global accent, which offers further dynamic levels. So as an example, the dynamics output might look would be something like this: soft note velocity = 48 loud note velocity = 96 accented soft note velocity = 72 accented loud note velocity = 128 BTW the best way to permanently connect a 9090 board to a MIDIbox 808 sequencer would be to skip MIDI and use the hardware triggers. Otherwise, you have a bit of latency to deal with. Besides the trigger, each drum voice on the 9090 also has a velocity CV input, so I don't know if we'd need an AOUT or if we could use the multiplexed resistor-matrix DAC on the 9090 board. I plan to order my 9090 board this week. Quote Link to comment Share on other sites More sharing options...
pcbatterij Posted March 27, 2011 Report Share Posted March 27, 2011 Hey all, concerning the leds: It might be an idea to use duo leds. This means 1 led with 3 leads that can have the following: off, orange and red. just an idea. cheers Jef Quote Link to comment Share on other sites More sharing options...
opkod Posted March 28, 2011 Report Share Posted March 28, 2011 The TR-909 still has a global accent, which offers further dynamic levels. So as an example, the dynamics output might look would be something like this: soft note velocity = 48 loud note velocity = 96 accented soft note velocity = 72 accented loud note velocity = 128 BTW the best way to permanently connect a 9090 board to a MIDIbox 808 sequencer would be to skip MIDI and use the hardware triggers. Otherwise, you have a bit of latency to deal with. Besides the trigger, each drum voice on the 9090 also has a velocity CV input, so I don't know if we'd need an AOUT or if we could use the multiplexed resistor-matrix DAC on the 9090 board. I plan to order my 9090 board this week. I have a TR909, and the 2 dynamics levels make the 909 sound. I made some mods on my TR909, but I could make more things with the 9090 pcb... think adding an 808 kick and 606 snare.... IT should be cool to have the same switchs as the TR909. to be continued. Quote Link to comment Share on other sites More sharing options...
nebula Posted March 28, 2011 Author Report Share Posted March 28, 2011 The TR-909 Service Manual refers to the velocity CV as the "accent" for each voice: Accent data on the CPU bus is latched into one of the ACCENT latches (IC2-IC9) selected by Address Decoder (IC612, 613). Latched ACCENT code is converted to analog equivalent at the output of associated resistor array RM0621. The voltage is clamped to the level until it is replaced by the next incoming data. What it doesn't tell us is what the actual data is. Those same resistor arrays also convert MIDI velocities to accent levels, so one could create the same dynamic feel as the TR-909 by identifying the actual accent voltages output from the sequencer, quantified against voltages produced by varied MIDI velocity. I don't know if anybody's ever really tried doing this, besides Roland of course. Maybe Trevor Page? For those who are interested, the TR-909 Service Manual is available from MatrixSynth http://www.matrixsynth.com/blog/media/roland/TR909%20Service%20manual.pdf Quote Link to comment Share on other sites More sharing options...
MPN Posted March 30, 2011 Report Share Posted March 30, 2011 Somehow I missed in my last post that there is independent velocity for each drum sound in the TR9090. From the manual: †The 9090 has a complete MIDI interface. Drum sounds are assigned to individual keys and are velocity sensitive. The interface, located on Board 2, is implemented with a 4MHz PIC 16F84 microcontroller. A number of latches, a discrete 6-bit DAC and demultiplexers provide separate trigger and velocity signals for all drums.†What is more interesting for this sequencer project is this post from Trevor I remember I read in the TR9090 Yahoo group: †The spare connectors adjacent to the PIC were originally intended for a sequencer board to plug into the main boards, but sadly my sequencer board never became reality. However... Those connectors may still be useful to someone, because they provide access to the DAC, multiplexers and trigger latches (full control of the 9090 by means of parallel data). †Best regards Magnus Quote Link to comment Share on other sites More sharing options...
bsilverberg Posted April 1, 2011 Report Share Posted April 1, 2011 You can see one implementation here: http://www.mulletronic.com/diy/mr9090/ (I've built this so I can tell you the circuit is good) Quote Link to comment Share on other sites More sharing options...
pcbatterij Posted April 15, 2011 Report Share Posted April 15, 2011 Hey bsilverberg, the link doesn't seem to work.... cheers Jef Quote Link to comment Share on other sites More sharing options...
pcbatterij Posted April 15, 2011 Report Share Posted April 15, 2011 Hey bsilverberg, I can open the link now cheers Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.