DUNE 2 is out now!!

[Richard_Synapse]

DUNE 2 is realistically the best all around synth ATM (and IMHO of course), but for me to fork out more money for DUNE 3, it would have to resolve some issues and have some features added to it:

-The dials would have to be decimal (an option for keyboard input goes without saying) and they would have to display proper units (i.e. Hz for filters, milliseconds for envelope parameters etc.);

I'm not sure what the point of this would be, to be honest. In my mind this is slower than just dialing in the value you want (very often it will be 0-100 anyway). Besides if we had say decibels this would be quite confusing since it would typically not match the scale you see. Furthermore, if you set up modulations in the mod matrix it would additionally not be identical to what you hear.

-The envelopes themselves would have to be fully customizable, graphical and work like the ones in Serum or Sylenth (those are pretty good envelopes IMHO; the ones in DUNE are not bad, but something just doesn't feel totally right about them);

Which ones do you mean? DUNE 2 has both regular ADSR and graphic MSEGs.

-It would have to become a synth with x (preferably at least 16) voices (where you can detune each voice and do many more things), NOT layers (although having eight layers is definitely not a bad thing);

DUNE 2 does have up to 16 voices of polyphony. This is multiplied by the number of layers in use, as shown in the monitor. So if all eight layers are in use and you hold 16 notes simultaneously, this would equal 128 internal voices.

-CPU increase would have to be no more than 30-40% (although I'm pretty sure Richard can pull off this stuff and keep the CPU increase in under 10-20%).

Good point Actually we work on decreasing CPU consumption a little, in order to make room for extra features. So overall I do not expect to see an increase in performance, it will depend on sound designers in the end though.

The reason I'm not a least bit sentimental about DUNE is that my thread at Synapse's forum regarding those non-integer dials was not answered by anyone from Synapse audio. Yeah, I get it - they don't have to. But, OTOH, a company not answering a legitimate request by a paying customer makes me very (VERY! ) unsentimental about that company.

We do read everything. Sometimes we don't get around to answer though, in particular when it comes to forum posts, sorry.

Note that if you need an immediate answer, the best way is to send a support email. We answer those typically within 24h, often less. We also answer most Facebook messages.

Richard

[Trancit]

nevermind

[Ghost Dog]

I'm not sure what the point of this would be, to be honest. In my mind this is slower than just dialing in the value you want (very often it will be 0-100 anyway). Besides if we had say decibels this would be quite confusing since it would typically not match the scale you see. Furthermore, if you set up modulations in the mod matrix it would additionally not be identical to what you hear ....

The point?

You use the dials to dial in (for example) cutoff frequency and because you're not limited to integer values, there is more resolution there is more resolution. You can dial in 534.74Hz and you can dial in 539.57Hz. If you need to dial in a very specific value (i.e. 560.5HZ), you double click the dial, a simple input textBox appears, you enter the value and press the enter key and you're done.

Being able to dial in 15.6ms attack of 5.65s release time with be way better than dialing 17 or 28. What the hell are those values? I'm guessing 17 is below 50ms (maybe not?!) and 78 is a couple of seconds, but why do I need to guess? Why doesn't it say 15ms or 50ms?!

As for decibels, well ... I personally with be happy with seeing decibels on the master volume fader only.

The mod matrix can remain the same.

.... Which ones do you mean? DUNE 2 has both regular ADSR and graphic MSEGs ....

A fair point. I was refering to ADSR envelopes specifically.

.... DUNE 2 does have up to 16 voices of polyphony. This is multiplied by the number of layers in use, as shown in the monitor. So if all eight layers are in use and you hold 16 notes simultaneously, this would equal 128 internal voices.

I have read the manual, so this explanation wasn't really necessary (although this stuff you can also find in any online review and such; you don't even have to look at the manual).

But what I was refering to is: it would be very nice for DUNE to be similar to Oberheim synths (with the SEM circuits) in terms of being able to detune each voice separately.

DUNE can detune each "voice", but that voice is actually one of the eight layers, not one of the sixteen voices within each of those layers.

[Teksonik]

DUNE can detune each "voice", but that voice is actually one of the eight layers, not one of the sixteen voices within each of those layers.

The 8 Layers each have 2 Oscs and a Sub so that gives you more than the 16 you're asking for and each Osc has a Fine Tune Knob so..........

Not Per Voice Detune but capable of great sounds. I have the feeling that Per Voice Detune (Not per Osc) is not possible in D2 due to the Unison Engine Architecture.
How about something like this ? Workaround to what you're asking but might provide a partial solution until ( if ) D3 gets per Voice Detune.

[fluffy_little_something]

I'm not sure what the point of this would be, to be honest. In my mind this is slower than just dialing in the value you want (very often it will be 0-100 anyway). Besides if we had say decibels this would be quite confusing since it would typically not match the scale you see. Furthermore, if you set up modulations in the mod matrix it would additionally not be identical to what you hear.

Maybe it would make more sense not to display any values at all, just min/max and zero for controls. Since I still show light symptoms of OCD, I am often tempted to set a value of 10 or 15, just in order to avoid values like 11 or 13. Without values one would simply have to trust one's ears and not feel like one has set a weird value

[Richard_Synapse]

How about something like this ?

Yep that's how I would program a patch, if I wanted that typical analog feel of an analog polysynth.

Maybe it would make more sense not to display any values at all, just min/max and zero for controls. Since I still show light symptoms of OCD, I am often tempted to set a value of 10 or 15, just in order to avoid values like 11 or 13. Without values one would simply have to trust one's ears and not feel like one has set a weird value

Makes sense, there is however cases where values are needed. E.g. when it comes to all tempo-sycn'ed controls (like 1/2, 1/4T etc), or if you modulate say the Cutoff parameter set to 10% and want to hit precisely 0% with a LFO or MSEG.

Richard

[Distorted Horizon]

..and want to hit precisely 0% with a LFO or MSEG.

Richard

Visual feedback?

[Caine123]

i love dune 2's sound! sadly i dont get the workflow with it yet. e.g. i muted all oscs and still get a sound somewhere. it is not visible asap, anyone knows a nice tutorial/overview besides the manual?

oh, i fuckn love the MSEG!

also bought recently the cinematic xp, finally a non-dubstep shit, hope to learn with it more with it.

[fluffy_little_something]

DUNE can detune each "voice", but that voice is actually one of the eight layers, not one of the sixteen voices within each of those layers.

The 8 Layers each have 2 Oscs and a Sub so that gives you more than the 16 you're asking for and each Osc has a Fine Tune Knob so..........

Not Per Voice Detune but capable of great sounds. I have the feeling that Per Voice Detune (Not per Osc) is not possible in D2 due to the Unison Engine Architecture.
How about something like this ? Workaround to what you're asking but might provide a partial solution until ( if ) D3 gets per Voice Detune.

Not sure what you mean (I never really understood the Dune architecture), but in Sylenth1 I have set up a slight modulation in my own default init patch: velocity -> (global) pitch. That works well. Without it two notes an octave apart for instance sound ugly when played at the same time. Same goes for overlapping notes with long release.

[Ghost Dog]

DUNE can detune each "voice", but that voice is actually one of the eight layers, not one of the sixteen voices within each of those layers.

The 8 Layers each have 2 Oscs and a Sub so that gives you more than the 16 you're asking for and each Osc has a Fine Tune Knob so..........

Not Per Voice Detune but capable of great sounds. I have the feeling that Per Voice Detune (Not per Osc) is not possible in D2 due to the Unison Engine Architecture.
How about something like this ? Workaround to what you're asking but might provide a partial solution until ( if ) D3 gets per Voice Detune.

Ok, regarding that picture that displays the mod matrix ....

Either you guys are purposely being obtuse (which is rude beyond belief ), or somehow you're failing to understand what I'm getting at (since one of you is a soft synth designer, I do believe you're trying to pull wool over my eyes, but let's remain civil (and of course, if we come to the conclusion that it was me who was wrong, I will be the apologizing, that's not a problem)).

That picture displays random values being assigned to 8 different oscillators on 4 different unison layers (that are called voices here). However, they are all playing SIMULTANEOUSLY and you need to set the unison voices "amount" to (at least) four for that to work. But, again, they are playing simultaneously.

What I was referring to is a (hypothetical) situation like this:

-The "amount" of unison voices is set to one (there is just one layer);
-When you assign random values (within the mod matrix ), it affects a certain voice (as in - voice of polyphony, not layer), but not the others.
-The voice is chosen randomly and, again, only one of the (up to) 16 voices "fires up" (not the others).

On another note, regarding the the displayed values (in ADSR, filters etc), the only synth of the most famous ones (AFAIK), that does this right is Serum. I'm 100% certain this is not difficult to implement, but hey, it isn't implemented (to use a popular phrase) "because .... reasons."

[Arksun]

That picture displays random values being assigned to 8 different oscillators on 4 different unison layers (that are called voices here). However, they are all playing SIMULTANEOUSLY and you need to set the unison voices "amount" to (at least) four for that to work. But, again, they are playing simultaneously.

What I was referring to is a (hypothetical) situation like this:

-The "amount" of unison voices is set to one (there is just one layer);
-When you assign random values (within the mod matrix ), it affects a certain voice (as in - voice of polyphony, not layer), but not the others.
-The voice is chosen randomly and, again, only one of the (up to) 16 voices "fires up" (not the others).

I think you're going to have to give an actual example to show what you mean exactly.

At the moment if you put in the mod matrix a single line of 'Random - Pitch Fine' then every subsequent note you hold down will be a different pitch to the last one even when held at same time, so the synth already does this.

Or did you mean that you wanted all the voices to be the same random amount simultaneously?

Either that or what you're actually saying is, say the synth is set to 8 voices, you want the order of which voice is played next to be random. So instead of say voice activation 1,2,3,4,5 etc, its 2,7,3,2,8,1,1,4,6,2 etc.

If thats the case, I have to ask what would be the point of doing that, is it because you want to assign each voice its own specific frequency point and other different variables rather than a random value?. I think given the fact that Dune is not a simplified fixed number of low voices system, trying to implement such a thing would be extremely difficult and in the end totally impractical.

For achieving specific values for single note in mono mode you can already do this by using the arp or a synced msegm though it would follow the same pattern each time rather than jump to those specific pre-programmed values randomly, though again, given variable voiceage of this synth, very tricky..

[Ghost Dog]

.... I think you're going to have to give an actual example to show what you mean exactly.

Oberheim SEM structure (like I mentioned in one of the previous posts).

Those single voice detunes and pannings are really nice (and not that difficult to implement).

[Richard_Synapse]

.... I think you're going to have to give an actual example to show what you mean exactly.

Oberheim SEM structure (like I mentioned in one of the previous posts).

Those single voice detunes and pannings are really nice (and not that difficult to implement).

Yes, true. In fact we have some unpublished internal version which does this, based on the layers, i.e. we simply cycle through them one by one on each keypress. This does give the desired effect, the problem we face though is that this isn't really compatible with all of D2's features, like the ability to have the arp activated for only some layers not others. In the context of D2, it feels like having two synth concepts in one, in practice it feels somewhat confusing.

Note that for tasks like detune or panning, we will add additional mod matrix tricks to do those easily. Teksonik mentioned the "Random" source in D2 which can be used (in fact is used in quite a few factory patches), sources like "Alternate" and similar would further improve on this (having the voices go left-right-left-right...etc).

Richard

[Cinebient]

Would it even costs a lot more cpu when i just create a patch, then copy several instances of Dune 2 and detune each instance, save it as multi-instrument in my DAW, ready.
This is also the only way to play Dune 2 with a Seaboard if i want to use MPE.
Good that we don't have the limits of hardware from decades.

[Teksonik]

Either you guys are purposely being obtuse (which is rude beyond belief ), or somehow you're failing to understand what I'm getting at

Which part of the following did you not understand ?

Not Per Voice Detune but capable of great sounds. I have the feeling that Per Voice Detune (Not per Osc) is not possible in D2 due to the Unison Engine Architecture.
How about something like this ? Workaround to what you're asking but might provide a partial solution until ( if ) D3 gets per Voice Detune.

I get what you're asking for but like I said it may not be possible due to the Unison Engine Architecture so here's a least a workaround to get some pretty awesome detune capability if you know what you're doing. No, it's not what you're asking for but I was just trying to help your dumb ass. A mistake I shall not make again......