Elektron Digitone

[k brown]

Yes, I'm still alive !

Just stumbled on an article about this FM synth and it looks to have a fascinatingly fresh take on 4-OP FM, that combines some simplification with added flexibility (including filters).

Has anyone done or started a project that implements some of it's interesting ideas? If not, I may take a stab at it in my tinker-toy non-coding way. Love the sound of FM, and always keen to look at ways to make it friendlier to pgm. The Digitone appears to make smart and musical choices in it's simplifications - i.e.: OPs B1 and B2 share the same EG, etc.

If I do tear into this, I may come begging fer help with the EGs - they're a bit unusual (though FAR simpler than DX7).

https://www.soundonsound.com/reviews/elektron-digitone
https://www.elektron.se/wp-content/uplo ... 210602.pdf

The basics of it's FM voice:

[Spogg]

Great to see you back in the saddle Kevin!

I would say have a go!

I make use of phase modulation in most of my SIM series synths and their configurations are many. I find that I can add a little “spice” to a sound where appropriate without going over the top. On those synths I generally don’t expose the PM to the user but rather incorporate it into the macro function of one or more knobs.

The principle of PM is simple but implementation is critical in my view. The ideal would be to get access to the full power but having an easy and intuitive way to get there to program it.

I’m already looking forward to seeing and hearing what you come up with and, as always, I will offer my help as and when and if you need it.

[k brown]

Oh, brother - I haven't even started on this thing and I already need help.

When they say that Op C (which is always a Carrier) Ratios are limited to 'integers', I assume that means whole numbers only? (Page 91 of the manual).

I understand what's described about Ops B1 and B2 Ratios control (a great concept), but nothing's coming to mind about how to implement it - anyone have any ideas? See below. (Near top of Page 91).

Also drawing a blank on how a knob might simultaneously control Ops B1 and B2 levels as shown on Page 92. A damned clever concept for FM Ops' level control - a unique triple-action crossfader. One output that goes 0-1 for the first third of it's range, then 1 to a bit above 0 for the second third, then back up to 1 for the last third (!). Second output that's stays at 0 for the first third, then up to 1 for the second third, then stays at 1 for the last third. Sure- no problem! This would probably need Ruby?

The manual's a bit sketchy on the Harmonics control, but I'm sure I can figure out something useful if not exactly the same. Probably using one of Martin's clever oscillators. (Pgs 92 and 93)

The unusual Operator envelopes are described on Pages 49/50 and 91; I may have something in the toolbox that I can adapt, but I'm not terribly confident ; - (


Really some brilliantly fresh thinking behind this FM engine; the non-sine options for the Ops make possible terrific Mix control. For example Algo 7, where one can combine the outputs of two 2-Op stacks, plus the direct outputs of all four operators!


But may have to settle for something similar/simpler, rather than close emulation; or, this may be a road never traveled . . .

[tulamide]

When they say that Op C (which is always a Carrier) Ratios are limited to 'integers', I assume that means whole numbers only? (Page 91 of the manual).

Yes.

I understand what's described about Ops B1 and B2 Ratios control (a great concept), but nothing's coming to mind about how to implement it - anyone have any ideas? See below. (Near top of Page 91).

That clock-like bit? Just make it like a clock. Endless encoder, one 360° turn updates the second value to its next turn, while the encoder value is reset to 0.25.

Also drawing a blank on how a knob might simultaneously control Ops B1 and B2 levels as shown on Page 92. A damned clever concept for FM Ops' level control - a unique triple-action crossfader. One output that goes 0-1 for the first third of it's range, then 1 to a bit above 0 for the second third, then back up to 1 for the last third (!). Second output that's stays at 0 for the first third, then up to 1 for the second third, then stays at 1 for the last third. Sure- no problem! This would probably need Ruby?

Isn't that basically a triangular waveform? That would be y = abs((x % period) - amplitude). This should do it:

Code: Select all

y = abs(( (x - 1) % 2) - 1)

x is the parameter value (I assume its range is 0-127) in the form value / 42, and the result will be in the range 0-1, so you will probably want to multiply it by 127. (x - 1) is phase correction, as the formula would else start at full level. However, this function will always go the full range, even at parameter value 85!

[k brown]

Thanks for looking into this.

I'm too math/coding ignorant to make much use of the info you provided, but if anyone else is dabbling with the some of the Digitone's concepts, it should be of great help to them.

I'll probably have to either scrap this idea or do something way simpler, just using a few of the ideas.

I.E.:

- A four OP with more conventional Ratio selection and output controls, but using the Digitone's Algos and modulate-able Mixing.
- More than SIne available for the OPs.
- 1-pole HP and LP in series with 12/24 LP and 12 HP Multimode filter.
- ADSR Fliter and Amp EGs.
- Simplified versions of the two LFOs.

Would still be a unique and fun FM synth.

[hardwarehost]

As I see it, on the picture is tri. osc multplied by 3 and streamed only in -1 to 1 range, min/max. If sub. or add value 1, it will act as phase shifting. I dindt take in depth look but I was working on something similar so its my association... The MIX controler between this X3 wave and before state should do the dashed lines on the picture.

family jewels : (cant post it all, only parts) x64 alpha

Snap 2021-12-02 at 09.09.37.png (19.16 KiB) Viewed 17274 times

Snap 2021-12-02 at 09.44.26.png (14.21 KiB) Viewed 17274 times

[k brown]

Not sure what you mean - the graphic I posted is not of an oscillator waveform. It shows output levels of two operators at increasing settings of the Level control. I described the resulting output levels in the text.

[hardwarehost]

The level increase from 0 to max is one phase of osc. finished - same as level control . The wave design is here described as output levels of two operators at increasing settings of the Level control

Simply: plug the level into phase and you need the wave - and thats the whole of what i wrote about.
You see on my picture the pretty close match.

The misunderstand was in modulation between the the two in whitch I am interested. If its not there in the real synth, than for emeulation its useless.

[Spogg]

I’ve looked at the relevant technical section of the manual now and none of it looks particularly challenging.

I can adapt something I already have for the cross-fade level control thing and it could be modulated even at audio rates.

I like the HARMonics thing where you morph between additive partial level arrays and I think I can do that too. I would like to hear that actually. It seems like a great alternative to scanning a wavetable. A bit more complex but almost certainly worth the effort.

That weird ratio “watch” thing is interesting too and I never came across that idea before. The review said it sounded notchy, like a zipper, and I can see that.

Regarding the envelopes I would personally choose to use Martin’s ADSR with a delay added (I already have that somewhere).

I’m in the early stages of my next project at the moment but I’m happy to take some time out for some or all of the above. Let me know what you need…

[k brown]

Continuing your well-deserved reputation as a Prince-among-men!

The Harmonics bit I may have some ideas about; we'll see how much trouble I get into there. Going over the manual again, I fear I'll probably need your help with everything else you mentioned

So you think the ADSRs you mentioned can do, or be adapted to do the 'End Level' bit?

By all means finish your current project, I won't have time to do anything on this till probably the end of Jan. - I'm deep into repairing and upgrading a bunch of my recording equipment for an upcoming two-day gig recording piano, tenor and soprano in an acoustically beautiful church. Several previous times I've recorded this soprano, something has gone badly wrong, so I'm really getting my ducks in a row for this. Those mess-ups were live concerts, and were mostly her fault; this is an actual recording session, so I'll have more control over her!! For the live things she would rehearse in one spot on the stage, and I'd set my mics accordingly - then for the concert, she'd stand in a completely different place!

Also, my first time doing a multitrack recording to a computer, rather live mix to 2-track recorder; so I'm sweating that a bit.

I'm almost 'mental' in my single minded focus, so I don't want to take my eye off this ball till it's 'in the can'. For me, Flowstone is not just a 'different hat', but almost a complete brain swap-out!

Thanks again to all who've responded - stay tuned.