It says 8 voice but 4 parts mult timbral.
From looking at demos it seems the osc levels (volume) are tied together for 2 voices? Is this right?
Can you not have 8 independent “osc” volumes/levels?
Not sure exactly how it work, especially with FM.
No one? I love the sound of this synth not cheesy like other FM. Surprisingly really.
Does my question make sense? Is there 4 or 8 “oscillators” that I could mix in separately?
Say it was the Analog Four. It has 2 oscillators per voice. So that gives me 8 oscillators to blend separately. Not sure how FM works vs normal analog wave synths.
Can anyone clarify please?
Give the manual a read through, it explains everything you want to know. It’s available on the support page of the main Elektron website. No real need for anyone to explain the DN architecture when it’s all described there by the great eangman.
1 Like
The meaning of voice is the same as in subtractive synthesis : the complete path from the oscillator(s) to the amp.
You have 4 tracks. Each track can be a different patch. You can also lock different sounds per step in the sequencer.
You can have max 8 voices at a time.
One track can be pushing 8 voices as long as the others aren’t trying.
There is voice stealing and you can limit the voices per track.
1 Like
So how many oscillators are in Digitone?
On Korg OPSIX which is mono timbral there can be up to 6 (all carriers, no modulators)
The Digitone is 4 part multi timbral and 4 OP… so would that mean (up to) 16 oscs? Is this right?
Yes if you are looking at the operators that would be correct. You can choose different algos per track. Think of each track as its own FM machine so to speak.
1 Like
Don’t know where you got the info of 2 voices volume being tied together. That isn’t correct.
Not exactly. Each sound has an overall level, and you can play any number of the 8 voices using a single sound.
The level of each note played can be controlled by note-on velocity, so if you play the Digitone’s sound from an external keyboard then you can control level of each note using velocity.
Each sound has two carrier outputs (X and Y) and you can program the mix between them. This is rather like having different oscillator levels on a subtractive synth voice.
2 Likes
Put aside this oscillators thing for the moment. It doesn’t suit FM much.
You have 4 tracks, so you can play 4 different Sounds at the same time.
8 voices that you have to distribute among the tracks.
For each Sound, you can mix between two outputs on the algo level, so it’s close to A4 2 oscillators architecture you’re describing above.
Now you have to learn a bit about FM synthesis. It’s not hard, you just have to take the time to understand how it works, and take the time to experiment!
https://www.elektronauts.com/search?q=fm%20synthesis%20%20%23elektron-gear:digitone
2 Likes
I just ordered a Digitone the best way to learn is indeed to experiment with it then I will know what is possible and what is not, if it is more limited than I hoped it may well have other features that compensate… and it does sound lovely which is the main thing.
1 Like
The Digitone has:
- Four operators per voice.
- Eight operator algorithms to choose from.
- Eight voices max.
- Four synth tracks
- Four MIDI tracks.
- Multitimbral up to four different sounds at a time (one per synth track).
Operators are sort of like oscillators (4 x 8 = 32 total) but are a little different. In the manual checkout APPENDIX A: THE DIGITONE FM SYNTHESIS. That gives a good overview of how the operators are used to create sound.
3 Likes
Am I wrong that actually the Digitone can play 8 different sounds at the same time? Ok, for the internal sequencer there’s only this trick, you need to use micro-timing trigs, shifting them together as closely as possible, so technically, you can only start 4 different sounds at the exact same time, but practically, there are i.e. many examples where people are doing all their drums (kick, snare, hats, …) on a single track, using sound-locking at the different sequence steps. And this technique only works properly (without cutting voices) because the track would be set to poly. And then there‘s multi-map mode, where you can lay many different sounds from the sound pool over a key range. There’s even an extra midi-channel option dedicated to this. Manual says, this is always available once programmed, and that you can store 8 multi-maps in a project, with up to 128 key ranges/sounds each. I think this would also work with a MIDI loop-back cable. So then you could use the polyphony of the sequencer to trigger more than one sound on a single step, using one or several of the MIDI tracks, for up to 8 different sounds in total. Am I wrong??
there are four “Tracks” per Pattern, i.e., voices . . . BUT the way a Digitone sequences a given pattern you can change voices on each Track (I think) per Step. and the Sound Pool from which these individual voices are drawn is 128 sounds (I think). so in theory, you can have a different sound per Track per Step for a given Pattern if you want to get that granular with the sequencing.
at the end of the day, however, Digitone’s polyphony is still 8 voices across the four Tracks so I think it’s not possible to have more than four actual sounds playing at once. where the additional four “sounds” may factor in is Digitone also has four separate MIDI tracks to sequence external devices. so you can also be triggering four other sounds in addition to sequencing Digitone’s four internal Tracks simultaneously, but those other four “sounds” are not coming from Digitone
You cane fake 8 different sounds at a time indeed.
Make sure notes are different on the same track to avoid voice stealing. And use microtime + sound lock: you should get there.
Actually you can push it further as algos have two different outputs that you mix: with some hard work you could even go to faking up to 16 different sounds!
2 Likes
@chrisroland, I think you are confusing tracks, voices and sounds a little bit. Tracks can indeed be polyphonic (up to 8 voices playing per track, with up to 8 in total over the whole device playing at any single moment), and if you trigger two (or more) different sounds with sound locks in two (or more) consecutive steps on a single track, they will not cut each other off, think of a pad, or a long crash sound, so this way, you can have full drums going on a single track, and with microtiming, you can at least put two sounds very close together, so practically, you can indeed start 8 different sounds (4 tracks x 2) on the “same” step.
Nothing to do with MIDI, so far.
I will get access to a Digitone tomorrow, and I’m very curious if it will indeed work as proper 8+ sound/8 voice drummachine, triggered over the multi-map MIDI channel, so I will try programming, like, 12 drum sounds over a key range (in multi-map mode), and then going to full finger-drumming fun, with Digitone doing the voice allocation, which could also then be sequenced from another device like a Digitakt or, connecting MIDI output to input on the Digitone (loopback trick) and with setting up proper MIDI channels, also recorded in realtime to internal MIDI tracks.
This would not work properly with sustained (holding notes) sounds, because you can’t have different note lengths (or velocities) for the polyphonic notes recorded or sequenced on a single step, but for drums, it should be good.
The OPSix can do this because it allows setting each operator to carrier. So in that case, it gives the impression of 6 ‘voices’, but it’s one ‘voice’ because a voice is the summation of all the oscillators/operators that are triggered with one note.
The Digitone maxes out at 3 waves per voice with one of it’s algorithm setups that features 3 carriers and one modulator.
Other than that, each ‘voice’ can be completely different, with it’s own LFO timing, filter settings, algorithm, etc. And there’s 8 voices. TRACKS limit how one can use the voices, so maybe that’s what you’re asking. 4 tracks can still fire off all 8 voices (playing a chord, or arppagio with ‘tails’ to the voice where they overlap individually, or to slam out 8 different timbres at once, you can use microtiming to trigger up to 3 different timbers at the same time on one track and do this until you run out of available voices.
1 Like
Hi, I know this is a bump, but I was moved to join after I read the post plus it’s a little self serving as I also have a couple Digitone questions.
Firstly, to the OP, what a beautiful and kind thing to do for your friend. I was so touched reading it - it restores my faith in humanity when someone sees a friend in pain and just wants to give them something that can bring them some joy and take their mind off of the painful stuff. Huge kudos.
To my questions if that’s ok.
I actually don’t own any Elektron products, but adore the sound of the Digitone and have been wanting an FM hardware synth for a long time now as I just don’t like the software options thus far. I myself was ill for 5years and had to quit music, and it’s been wonderful getting back into it. I have some great synths that have just been sitting there since 2018 and I sold my iMac Pro and some other stuff but kept most of my synths as I knew that the first day I could, I’d be back into my beloved music. Started in the early 90’s with Cakewalk “dots” on a 386, eventually moved to Logic 3.5 in '97 then to Mac when Apple bought it but I waited till 2008. But I digress…
I have been looking at everything that’s out there, and I just have to say, I don’t like the Opsix at all. All I read is what a monster it is but the raw sounds of it just doesn’t get me, where as I find the sound of the Digitone to be so much more pleasing.
The EssenceFM is out of my price range right now but it has a ton of polyphony and can sound fairly good from what I have heard. But there is something special about the Digitone’s, erm, tone! And it’s all about the tone, isn’t it? 
But I too am concerned about polyphony. I am a pad and chord player, sure I will use it for FM Bass and leads sometimes, but I really enjoy FM Pads, and I am worried that I will have egregious note stealing with the way I play.
Say I play 3 or 4 note long release chords, 8 voices get chewed up fast. But I CAN cope with that, if it’s truly 8 voices, hence my question.
Is it really actually 4 voices, as in 4 voices per 2 operators like the DX7, meaning if I want to use all 4 FM operators it’s only 4 voice total, or is it 8 voices using all 4 FM ops? I think this one might be a dealbreaker for me as I’ll need the 4 ops in many cases to get the huge pad sounds I want.
Anyway, thanks for reading and sorry for the bump, but this topic seemed to be the most relevant one to my question after a Google search.
Thanks very much in advance 
1 Like
i sequence my Digitone externally from Deluge – one midi track per Digitone track/voice. i have yet to run into polyphony cutoff, which i was slightly worried about with 8-voice polyphony. the arpeggiator and chord memory feel like they let me cheat a little more on not hitting that limit, but also i may be more conscious of trying not to have all four tracks sounding at once
And to answer your question: Yes, the Digitone will play eight notes with four operators each.