New DIVA components

[Funkybot's Evil Twin]

Nope, it's a 4-pole design (in hardware). The guy just got a quite messy transfer function when he tried to put it in pole-zero form.

http://www.timstinchcombe.co.uk/index.php?pge=diode

Just so I'm clear: the 303 has a 24db filter?

[EvilDragon]

Yes. Read the article.

[Nokenoku]

Well, we timed the envelopes of our vintage synths and used that as a base. Hence I'm quite unsure what it is everyone is talking about. We replicated them as closely as we possibly could.

I also cannot follow here. It never ever crossed my mind Diva's envelopes could be too slow, and I also never heard anything in that directions from other users.

I'm absolutely sure, they're fast enough. Speed is not what -I- was talking about.

It's probably rather the form of the envelopes. I might be used to ones being a bit more exponential.
Also to me the knob-to-value transition of the decay times feels a bit inconvenient. It's somehow hard to hit the right spot.

That said:
After all I finally got exactly the kind of basses out of DIVA I was searching for.
It's just, that it took a bit long.
Maybe it's just a matter of habituation. But I still think an additional envelope module wouldn't hurt. Best would be some clean digital one, where you can set "bending" the envelope from -100 to +100 (0 being just linear).

That makes me wonder...perhaps DIVA's envelopes aren't slow at all, and maybe people just don't like the scaling on envelope controls. Not enough screen real estate being devoted to the faster side of the envelop could give the envelopes the illusion of being slower. Whereas something like Bassline 101 may have more space devoted to the faster side of things allowing for a wider, fast section, with more fine control (e.g. "faster envelopes").

YES!
That's exactly, what I mean.
The short times don't have enough space in my opinion.

[Sendy]

I can second that the distribution of envelope times is a tad biased towards the longer times. Unless you're making pads, attack and decay times tend to have a sweet spot which is quite small in Diva's widgets. The times are THERE, it just requires more fine adjustment to get at them than on some other synths. Once you get used to that fact, you barely even notice it, though I agree, a fully digital, curve-bendable (from convex to concave) envelope would be a super convenient addition.

[Funkybot's Evil Twin]

Yes. Read the article.

I'm at work right now but I'll check it out later. Truth be told is I opened the link, saw a circuit diagram and got scared.

[fmr]

But you can, of course, allow for multiple selection in the Dual VCO, and that will do the trick

That should be possible, or am I missing something? (Dual VCO, not Dual VCO eco)

Well, I can select more than one wave, but the result sound is not quite as I expect (when I have the saw selected and I select pulse, the sounds changes, but not as strongly as it should, IMO). But maybe it's my ears that are tricking me

OK, I'm at home, and I gave it a test. I selected the Dual VCO, the Feedback, the Uhbie filter and the analogue envelopes. I think this is the closest I get to an Obie in DIVA right now.
I used the OB-Xd as a comparison (it's no secret I am very involved in testing this, as I am an old time fan of Oberheim, although more of the later period - Xpander, Matrix-12 - than the earlier, but nevertheless).
I tried to select both saw and pulse waves. In DIVA, when you have saw selected and select pulse, you have a noticeable difference in sound. BUT, if you now deselect saw, you almost don't notice. It behaves as if the pulse takes over the oscillator. The same happens with the other waves (but then, the Obie only has a sine/triangle when both other waves are deselected).
AFAIK, the Jupiter only allows for one wave at a time, so the behaviour of the Oscs doesn't affect that (actually, it even goes a little further, since there is a colouring in the sound when we select other waves).
But if we want to use those as ans osc section to emulate an Obie sound, it doesn't work the way we (I) expect.
Maybe you could revise the Dual VCO module, with a switch, or something, and re-balancing the different waves in order to have more saw when the pulse is present.
I'm not complaining - you already gave us so much with this synth, I cannot
But I think I had to make my point clearer, for those that are expecting to have Obie sounds now. Anyway, having the filter is already a great thing

[News From The Sky]

The Jupiter 6 allows the user to select multiple waveforms for the oscillators, that's why it's in Diva.

[Sendy]

The problem may be that pulse waves and the square in particular sound a lot louder than other waves, so when you have a square and saw with equal amplitude mixed, the square always dominates. I find this is a "problem" in both the Dual VCO and the DCO modules - being able to adjust the level of the square/pulse in a trimmer somewhere would solve it, volume controls for every waveform would be ideal (but perhaps a bit clunky on the interface).

[pdxindy]

The problem may be that pulse waves and the square in particular sound a lot louder than other waves, so when you have a square and saw with equal amplitude mixed, the square always dominates. I find this is a "problem" in both the Dual VCO and the DCO modules - being able to adjust the level of the square/pulse in a trimmer somewhere would solve it, volume controls for every waveform would be ideal (but perhaps a bit clunky on the interface).

It is the noise level that I most would like to control in the Dual VCO!

[fmr]

Another thing I wonder: If the Uhbie filter had a -24dB switch (as is in the OB-Xa and above), would it sound the same as the Jupiter filter at -24dB?

I could come pretty close to the sounds expected using the Jupiter filter, but I had to ask, as maybe someone has access to one of the real things and can compare. Of course, if the difference is minimum (as I believe it is), we can go over.

(a -24dB switch in the Uhbie wouldn't hurt either)

[Sendy]

The problem may be that pulse waves and the square in particular sound a lot louder than other waves, so when you have a square and saw with equal amplitude mixed, the square always dominates. I find this is a "problem" in both the Dual VCO and the DCO modules - being able to adjust the level of the square/pulse in a trimmer somewhere would solve it, volume controls for every waveform would be ideal (but perhaps a bit clunky on the interface).

It is the noise level that I most would like to control in the Dual VCO!

A little "noise + pulse volume" trimmer that scales the gain of both waves in tandem, tucked away in the trimmers page somewhere would be an elegant solution, at least from the enduser perspective

[fmr]

Another remark, this time regarding the CS-80. This synth is based on a dual architecture (I believe it's like two CS-60 coupled). It has noise source with volume control, high and low pass filters with individual resonance simultaneously (a pair per synth circuit), a sub-oscillator, a very extensive and comprehensive ring-modulator, and an incredible response to touch. Of course, it would be impossible to model all this in DIVA, but I can understand why the CS-80 came immediately after the Oberheim.

The Osc module of the CS-80 would be a great addition, as well as the filter module. The Ring Modulator... I sincerely cannot devise where it would fit (perhaps in one of the sub-panels in the lower part). But this would be a DIVA 2.0, I guess.

CS-80 is just a brass machine? To those heretics take a listen to Beaubourg, by Vangelis.

[Urs]

Another thing I wonder: If the Uhbie filter had a -24dB switch (as is in the OB-Xa and above), would it sound the same as the Jupiter filter at -24dB?

I could come pretty close to the sounds expected using the Jupiter filter, but I had to ask, as maybe someone has access to one of the real things and can compare. Of course, if the difference is minimum (as I believe it is), we can go over.

(a -24dB switch in the Uhbie wouldn't hurt either)

Hehehe, the OB-Xa has two separate filter circuits. One resembles the Uhbie with a SEM-style layout. The other one is an OTA cascade - it'll sound much like the cacsacade filter.

It's not easily possible to make Uhbie 24dB/Oct and preserve the ability to blend between filter modes.

[Urs]

Regarding envelope shapes I'll happily check out our SH-101 - maybe it's "true exponential" without an added bias below zero. I'm also open to something like a 5x faster switch like the MS-20 has.

[Urs]

Regarding the CS-80, first of all it doesn't fit in Diva's voice structure. Like the 800dv it's a layout with two parallel VCO-VCF-VCA stages. We could at most do just one lane in Diva, and that would suck.

Secondly I'm convinced that the CS-80's appeal stems from the hardware (poly-at keyboard, ribbon controller) and from the people who mastered it. If it was for the sound and features of the synthesis engine, we should be having a galore of requests for a CS-30 implementation rather than a CS-80. But I suspect that even if we did a polyphonic CS-30, it would be sniffed at, regardless if it sounds exactly like a CS-80 with a better feature set. I believe that the hype for a CS-80 emulation has more to do with psychology than with the technology itself. I'm sure it'll always disappoint (even if a CS-30 clone would be something worth wile IMHO)

That said, I sincerely hope that some other company comes up with a CS-80 emulation on Diva's level. A companion product to Diva with a dual structure isn't necessarily what I'm into.