The CPU is a Coldfire MCF5441 with up to 385 dhrystone 2.1 MIPS @ 250 MHz … Same one as in our other current lineup of products sans the Octatrack. (But before anyone asks - no that doesn’t mean we can run anything on everything, it’s a general purpose CPU and it does very different things across all machines.)
All DSP is done in assembler. Our process for this project was that first I made prototypes in Max which I made Max4live Devices out of and sent out to all my colleagues. We then collectively tested them and improved the patches based on the feedback received. After we were happy with how they sounded we started translating the patches to run on hardware, and this was quite a long process where I worked closely with our amazing DSP engineer Oscar A who dealt with my stupidity for a few months until we had all of the machines running on hardware. Thanks Oscar!
4 Operators are a very good amount I think, it’s a perfect balance between easy to handle and versatility. Add different waveforms to the mix and some other tricks, not really necessary to go above unless you want to do some very specific things.
The FM is not FM at all; it’s phase modulation just like the Yamaha stuff etc. PM has many advantages, but the biggest one is the possibility of doing feedback without detuning the sound.
Not really, I was just messing around with some ideas. At the time I was obsessed with trying to make all parameters feel ‘smooth’, so I came up with the oscillator balancing thing. At first it was sort of a mix between additive synthesis and chord intervals, but we later as a team made it much more musical by introducing the idea of using chord inversions for the interval mixing.
The voice itself is simple - it’s just four oscillators that each play the same wavetable. I made the waveforms with a max patch which later turned into Hard/Softcore. All of the tables in the Cycles are made strictly with additive synthesis. Sorting them in the particular order took a long time, it was really hard to find an order that felt logical and nice to sweep through, hehe.
I would like to add graphs of all the waveforms in the manual, but I haven’t had the time to plot them out yet.
The Kick drum is inspired by classic drum machines but with some FM sprinkles. It started off a bit wilder, with a ratio control and stuff, but Tone is so much better for that and I wanted the Kick drum to be fool-proof and always sound fat and punchy. Plus, having the setup it currently has is super specific and allows for some very cool ticks. I think there is so much more to FM than what you would expect, so having the triangular waveform shaping that goes to feedback is kind of unique and something I think sounds really good. Can’t do that on the Digitone!
I hope so, I think it’s very necessary for live performance.
From what I understand it was to keep the form factor slimmer and the handle was an interesting design decision. The cable is quite stiff, so it doesn’t really pose any issues.
Theoretically yes, but in practice the Rytm does a lot of other stuff so it won’t have enough power to run it. So short answer: no. Long answer: nooope.
Yes. There were one patch that didn’t make the cut, but it was mostly due to time constraints and its structure being radically different to all the other patches. The system in the Cycles that Oscar (DSP engineer) put in place is quite ingenious. It’s a huge structure that can dynamically change, this is what allows us to swap machine per step. It’s wild!
No, it was all made in Max and we used Ableton Live to try the different machines out as Max4Live patches. (No, I can’t share the patches and they also sound worse than the final product, much clickier and less refined overall. But the sound is pretty much the same)
Yeah we should try to improve upon that, I also think it’s too hard to distinguish.
The Reverb does have a Tone parameter (Hold FUNC and turn Reverb Size), but the Delay has a fixed HP/LP setting that rolls of nicely.
Yes, it’s a soft Thru, but it’s done right so… Shouldn’t pose any issues. 
Yes, it’s the same. I think I’ve commented on this before, but it’s simply the material being put into the reverb that is different, and therefore the tone of the effect itself is perceived as different too.
Depends on how you want to use it. You can see it as a save point you can go back to in case you made a change to the pattern you disliked. Or, you can use it as a performance control, save current pattern status, mess with it and then reload to the point before.
Apparently yes, I had no idea. Really cool trick! Hehe.
No, the mapping and scaling is our secret sauce. The algorithms, maybe.