u-He Zebra II Vs Arturia Complete (Analog)

[PAK]

DIVA beta is up in the u-he forum

I just checked it and did a first 1:1 comparison with my Moog Slim Phatty.
My first impression: You could forget all other Moog emulations in comparison with Diva...

Yes indeed. It's still way better than any other native offering even in Economy mode - if anyone's worried about CPU. The fact that the Moog section is only part of it, and the other parts on offer are also excellent, make Diva a pretty special product. The comparison to Arturia would really be with their Origin hardware rather than their software. I've never laid hands on that hardware, but I do think the sound quality in Diva dates Arturia's native stuff quite badly and, when mentioning the two side by side, the more apt comparison might now be talk of "next generation" Vs "previous generation". I'll leave peoples ears to decide which one belongs where

Re Diva specifically as a Moog emulation. Some hardware zealots will likely still "nitpick" aspects, but I think they're going to have to cherry pick their Moog A/B examples very carefully to attempt to prove something is "superior". I'd only make two observations..

First, to my ears, Diva's filter has a (slightly) different "driven" character to it than some Mini's I've heard. Turning up the mixer Feedback knob does help with a perception of a more rounded tone, but also introduces a "fuzzy" element to the character along with that I will admit that I wouldn't mind a switch on Diva which gives a nod to those "other Mini's" and attempts to make reproducing a slightly more "rounded/smoother" tone easier. I suppose I should give a "real world" example, so I'll pick a sound many should know. Keep in mind it's youtube, and a higher bitrate source would be required to make judgements about relatively small tonal differences. Anyway, the synthbass is exposed at 1:47.

I chose that because there's a rounded element to the tone, but yet the resonance remains rather "bitey". These two things conflict, which is why such a simple sound is actually rather tricky to copy exactly. Of course its a full production where other elements may influence the sound, and a really subtle chorus (can't be done in the current Diva beta - no chorus "mix" knob!) would maybe help Diva out. But I still think it's a little tricky to obtain a "roundness" to the sound with Diva - it leans towards being more "buzzy". Here's the Diva preset I made for anyone to try - maybe you can do better than me?
http://dl.dropbox.com/u/37542806/PAK%20 ... 20Diva.h2p

Be sure to mess with the feedback knob to hear some of what I was attempting to reference above. I mapped resonance to velocity in order to give some control of keeping the bite whilst closing the filter enough.

Secondly, if pushed, I get the perception that some hardware Mini's may have "more umphh", specifically related to certain envelope settings and how Diva reacts to them sometimes. I'd need a Mini beside me to make a proper comment about this, because I'm not totally sure what I'm picking up on. Beyond that it'll be interesting to see what people come up with if they want to criticize Diva (at least the Moog portion), because there really isn't room as my ears hear it - Urs and co have done a very good job.

Indeed it's so good that it's likely to cause some worried reactions amongst a few hardware vendors, much less software vendors. The fact that it's offered at a price much lower than many lesser emulations, and a fraction of the cost of hardware, is just icing on the already tastey cake. I hope everyone inclined to support these efforts does so. At the very least you need to help fund Urs' ability to obtain a Rev 2 Prophet 5 to emulate I think a really good SSM emulation would be next on many peoples list..

[Gonga]

PAK, I often find your posts very interesting and once again you've said it perfectly! I'll surely be getting Diva before Dec 31 but don't have the cash on hand atm...

Urs has said that the attack architecture is different than Zebra's (which I love), and yet I've been getting wonderful attacks from Diva as well. I feel bad that I am busy with other projects and can't get to an analog project atm, but I cannot wait to sink my teeth into Diva, Zebra and ACE this winter!

Diva is synthesizer history in the making. I have wondered though why others haven't tried these new approaches that Urs has taken, and how long before they do?

[hakey]

I have wondered though why others haven't tried these new approaches that Urs has taken, and how long before they do?

Not strictly speaking a synth, but afaik Cytomic's The Drop uses a similar approach.

[Lotuzia]

Diva is synthesizer history in the making. I have wondered though why others haven't tried these new approaches that Urs has taken, and how long before they do?

Maybe a lot of not, or poorly, informed people are mistaking apples and oranges.

SMM are integrated circuits, and not a suite of separated components like transistors, capacitors, diodes etc.

So they cant be deciphered by techniques like Spyke, often mentioned in the commercial part of promo, where a thorough study of the components and circuitry will add something, but will be hard to model completely due to the overwhelming number of equations you have to translate into the digital domain. ( And incredibly costly CPU wise ).

For INTEGRATED circuits you have much much less information. (But you can have some. Few actually )

These informations about integrated circuits, as well as deciphering of the circuitry of analog monsters are one of the basis of -I believe- many companies works since a lot of years. ( I know at least three companies using these techniques ), but I'd say there might as well be some more. Its not like if something new was invented here. What I can reveal without breaking any rules or NDA is that The Xils-Lab PolyKB and Synthix for example are also emulating integrated circuits, as well as analog circuitry.

Then obviously after decrypting all these components, choices have to be made. And additional things can be found.

BTW : I'm not speaking here of the intrinseq quality, or less quality, of Diva.

Just about some of the different techniques that can be used to build analog models.

So to answer your question, nothing new under the sun. ( At least for me :hrug: ) its just the popping out into the global community of some very technical points developpers usually did not care to discuss about or share with users, maybe because theu thought it would be very annoying for them, but maybe the audience is ready, or even willingfull, to discuss such things now, I dont know.

Anyway, this will add some info about what are really integrated circuits, opposed to a suite of components, etc, etc. You really have to distinguish antique units and what components they used to understand the difference. Hope all the above is not too technical btw.

LtZ

[hakey]

nothing new under the sun

Zero-delay-feedback filters?

[Lotuzia]

nothing new under the sun

Zero-delay-feedback filters?

If you read carefully the two precedent posts <EDIT Before the one you quoted> you"ll see they were refering to SMM emulation, and some of the analog circuitry emulation methods, wich was what I was discussing.

These methods might have been seen as "total innovation" by non informed users, as these are quite technical points, so it was a pleasure to discuss them a bit here.

So I can see that you have nothing to say or discuss about the actual content of my post, and therefore might totally agree with it. I'm deeply honored

[goldenanalog]

It's very interesting that the Minimax is *STILL* considered one of the very best Minimoog emulations to date; and that it's a 12-note poly to boot. Wow...

Any opinions on the Creamware Prodyssey, especially by direct comparison to an actual ARP Odyssey?

[PAK]

SMM are integrated circuits, and not a suite of separated components like transistors, capacitors, diodes etc.

You mean SSM? I wouldn't have mentioned it but you misquoted the name a few times. Btw, even taking that into account, there is more than "just the chip" when it comes to copying a particular synths usage of it, since different synths can use the same chip and still sound different. There's also discrete copies of the SSM 2040 out there too, and I'm not aware of a software emulation which sounds as good as those (modular based) copies, so an exact copy of them would still be more than I think we have now

developpers usually did not care to discuss about or share with users, maybe because theu thought it would be very annoying for them, but maybe the audience is ready, or even willingfull, to discuss such things now, I dont know

I do find it interesting for developers to explain things, like overall techniques related to how they attempt to improve things, or their ways and methods of measuring things. It serves to give end users a hint about someones dedication and determination to achieve a certain level of detail. Ultimately the user judges the end products sound for themselves though, so I don't think it's that important for overall marketing purposes

In U-He's case, whilst there's an element of marketing to those sorts of things, ultimately the sound must speak for itself. It's rather bold and confident when someone more-or-less states "we attempted to blatantly copy this in excruciating detail, forget all that 'inspired by' stuff". Such invitations to comparisons usually invite a rain of arrows on the coders head. I bet Urs is surprised by how little of that he's had with Diva (so far!) especially when it includes an attempt to model arguably the best known synth ever made (well, outside of the DX7 and M1 maybe ) I think Urs deserves some kudos for that. Whilst many developers might perform similar techniques I think considerably fewer are able to pull it off with the style that Diva has. Just my opinion

[Gonga]

Thanks for the explanations. I was actually referring to the zero-delay approach, which is what I thought makes Diva particularly special. My understanding is not long ago Urs was saying he wasn't sure even high-powered computers would be able to run more than a single voice, so I assumed it was a ground-breaking approach. The uniquely beautiful sound of Diva seems to underscore the notion that something new is under the hood. Not only the oscillators and filters, but the attack also impresses me very much as being superlative in sound character.

[pdxindy]

nothing new under the sun

Zero-delay-feedback filters?

So I can see that you have nothing to say or discuss about the actual content of my post, and therefore might totally agree with it. I'm deeply honored

He commented on the actual content of your post. You claimed there was nothing new and he responded with Zero-delay-feedback filters which is something new and therefore related to your post. A perfectly reasonable reply on his part.

[Ingonator]

You could maybe check a comparison with the Arturia Minimoog demo version which i have created yesterday:

Hi,
i just started recreating some Minimoog sounds based on patchbooks i found in the web. For better comparison and for adjusting the parameters i got the 15 days evaluation version of Arturia Minimoog V.

Here is a comparison (wav file) of the same preset, first in Diva and then in Arturia Minimoog V (Arturia starts around 1:38 ):

https://rapidshare.com/files/2628398261 ... son_01.zip

This is a single instance of each synth and no internal or external FXs were added.

The Diva preset is available here:
https://rapidshare.com/files/1658076369 ... ad__IW.zip

I think this shows which problems the Arturia synths got in comparison to Diva or the real thing...

More Minimoog and other presets will follow soon.


Ingo

I am working on recreating more Minimoog patches with Diva now.

my opinion is that the Xils synths (especially Synthix) are some of the few which seem to be in the same league as Diva in terms of analog sound.
Synthmaster 2.5 got some great filters too. I'll try to replicate some of my Diva presets with it soon.

You mean SSM? I wouldn't have mentioned it but you misquoted the name a few times. Btw, even taking that into account, there is more than "just the chip" when it comes to copying a particular synths usage of it, since different synths can use the same chip and still sound different. There's also discrete copies of the SSM 2040 out there too, and I'm not aware of a software emulation which sounds as good as those (modular based) copies, so an exact copy of them would still be more than I think we have now

About the SSM filter: While i made a few Diva presets inspired by the Polysix (based on the fact the Diva Ensemble is more or less modeled on it) Urs mentioned he tries to get a Polysix (with a SSM2044) or an older Prophet 5 version (Rev2 with SSM2040 i guess) for the replication of the SSM filter. His Siel Opera got it too but sounds totally different.
BTW Waldorf PPG Wave 3.V models a SSM 2044 too and it sounds fantastic IMO.


Ingo

[hakey]

nothing new under the sun

Zero-delay-feedback filters?

If you read carefully the two precedent posts <EDIT Before the one you quoted> you"ll see they were refering to SMM emulation, and some of the analog circuitry emulation methods, wich was what I was discussing.

Hmm, yeah, but you started your post with the following quote from Gonga:

Diva is synthesizer history in the making. I have wondered though why others haven't tried these , and how long before they do?

And, after a load of semi-coherent waffle, you went on to say:

So to answer your question, nothing new under the sun.

As Gonga's question referred to the 'new approaches that Urs has taken', and as the zero-delay-feedback filter approach is, afaiaa, a new approach, it seemed that your answer to Gonga was misleading; as such, I corrected it.

And, sorry, but there was no indication that that particular claim was intended to address any other post than Gonga's (but every indication that you were engaging in a not-so-subtle attempt at talking down the work of a competitor - real classy ).

-

[hakey]

A perfectly reasonable reply on his part.

Thanks.

[Fritze]

Here is a comparison (wav file) of the same preset, first in Diva and then in Arturia Minimoog V (Arturia starts around 1:38 ):

https://rapidshare.com/files/2628398261 ... son_01.zip

Ohhhh, even after some loudness correction the differences are stunning. The Arturia sounds like an organ in comparisson. It's that static. Are you sure you have the detuning set correctly and everything? And -can- you switch the legato off in the Arturia? I never played with it as I once owned the Moog Modular by Arturia and didn't expect anything above that from the Mini.
Diva ownes them, hands down.
Gotta try the Xils.

[Urs]

First, to my ears, Diva's filter has a (slightly) different "driven" character to it than some Mini's I've heard. Turning up the mixer Feedback knob does help with a perception of a more rounded tone, but also introduces a "fuzzy" element to the character along with that I will admit that I wouldn't mind a switch on Diva which gives a nod to those "other Mini's" and attempts to make reproducing a slightly more "rounded/smoother" tone easier. I suppose I should give a "real world" example, so I'll pick a sound many should know. Keep in mind it's youtube, and a higher bitrate source would be required to make judgements about relatively small tonal differences. Anyway, the synthbass is exposed at 1:47.

I chose that because there's a rounded element to the tone, but yet the resonance remains rather "bitey". These two things conflict, which is why such a simple sound is actually rather tricky to copy exactly.

--- snip ---

Secondly, if pushed, I get the perception that some hardware Mini's may have "more umphh", specifically related to certain envelope settings and how Diva reacts to them sometimes. I'd need a Mini beside me to make a proper comment about this, because I'm not totally sure what I'm picking up on. Beyond that it'll be interesting to see what people come up with if they want to criticize Diva (at least the Moog portion), because there really isn't room as my ears hear it - Urs and co have done a very good job.

I do think these are related maybe. Most Minimoog emulations and replica suggest that the filter operates at a much lower cutoff range than ours, maybe in the vicinity of an octave lower. Our Mini will not produce a resonance spike much below 70 Hz, but it will do so far beyond 20kHz. So I think the trimpots and parts are at exceptional levels, and we trust Mr. Zimmer's observation that this is his best sounding Mini, maybe because of these unusual settings.

Should we get hold of other Minis in future then we might compare them and add flavours

Also... it is very important to note that we emulated VCA distortion and how we implemented it. The VCA distortion on a Mini is related *only* to the volume envelope, which can be observed by the harmonics added with higher Sustain levels (but the output knob is perfectly linear). Diva however has a dynamic VCA with a volume knob and a choice of envelope. You can influence the VCA distortion by changing the VCA volume in addition to envelope settings. On top, Diva has a master output parameter that's perfectly linear.

Thus, whenever Diva lacks drive or adds too much, adjusting the VCA volume and compensating by master output might do the trick, specifically in the Transistor ladder (our VCA model is built into and swapped with the filters, to allow for authentic zero delay feedback modeling... not just resonance...)