The long DIVA thread

Official support for: u-he.com
Post Reply New Topic
RELATED
PRODUCTS
Diva$209.00Buy

Post

irrelavent.
Last edited by goldenanalog on Tue Nov 01, 2011 8:51 pm, edited 2 times in total.

Post

:roll:

Post

Where is DIVA fitting into the general U-He development schedule? I'm guessing the next Zebra update is first, but will that mean that DIVA gets all the attention after, or will time be spent on other things? Any chance of DIVA before Christmas?

I haven't been this excited for a synth in a long time.

Post

Forgive me if I am asking an answered question, but 24+ pages is too much for my weary eyes . . .
According to www.u-he.com, "We've been silent for a while, but far from idle! Urs and Clemens are still working feverishly on our NEW SYNTH. If all goes well, there'll be a major announcement soon." Is this synth the one or is it Berlin Modular that is being released soon?

Post

Urs has made some recent posts in the non-sticky'd DIVA teaser thread in this forum about progress on development. Good news sorts of things, worth checking out!

Post

Will DIVA sound much different from Zebra and ACE?

Post

It's gonna sound better and more analog :)

Post

+1 :oops:
circuit modeling and 0-dfb filters are cool

Post

EvilDragon wrote:It's gonna sound better and more analog :)
Sorry, I'm being a bit impatient but in what way? I am not a coder so I haven't a clue but is it a new way of approaching programming like the modelling of individual components?

Post

EvilDragon wrote:It's gonna sound better and more analog :)
I heard it includes ALL of THE MOST successful versions of "this plug-in".

:shock:
I'm not a musician, but I've designed sounds that others use to make music. http://soundcloud.com/obsidiananvil

Post

Basically, DIVA's component modeling is not based on digital filters that employ a 1-sample delay in their feedback loop (Z^-1 block). That delay is responsible for incorrect behavior of filter resonance and filter FM... so you can expect a lot better sound in that regard. Urs and co. are basically using a prediction mechanism instead of a 1-sample delay. This translates to a lot of CPU needed, but also a sound that's a lot more faithful to the venerable analog sound.

At least that's my understanding. Urs, or anybody else, correct me if I'm wrong about something.

Post

EvilDragon wrote:Basically, DIVA's component modeling is not based on digital filters that employ a 1-sample delay in their feedback loop (Z^-1 block). That delay is responsible for incorrect behavior of filter resonance and filter FM... so you can expect a lot better sound in that regard. Urs and co. are basically using a prediction mechanism instead of a 1-sample delay. This translates to a lot of CPU needed, but also a sound that's a lot more faithful to the venerable analog sound.

At least that's my understanding. Urs, or anybody else, correct me if I'm wrong about something.
Do VA hardware synths already employ this type of modelling?

Post

Nope, not a single one from my knowledge. It really needs a lot of CPU, and current DSP CPUs used in VAs (for example Sharc DSP in Virus) aren't fit for the task. Hey, even current i7 CPUs won't give you more than 2-3 notes of polyphony per core on highest quality setting of DIVA!

Post

in short: We hope to release Diva before xmas. We also hope to get Zebra 2.6 into public testing this year. After that we'll push ACE and Berlin Modular forward, with hopefully a bit of other plugs inbetween.

We've etablished a nice workflow for Diva. I hope we can keep that for Berlin Modular next year.

Regarding the technology, Evil Dragon is quite right. We've developed an amazingly fast prediction algorithm that allows us to calculate the output of a filter - with itself as an input variable (chicken and egg problem). In addition we carefully analysed a range of analogue synths and compared their behaviour to papers, emulations, simulations and our algorithms. Latter we tweaked to match as good as it gets.

We're doing Diva first because it's a simple synth and because it might honestly appeal to a wider audience than any of the other stuff in the pipeline. We will take this technology further in the other u-he plugs, which do not necessarily need to sound like this or that but have a certain degree of freedom.

That's it in a nutshell :)

Post

Urs wrote:We also hope to get Zebra 2.6 into public testing this year.
That would be awesome! :)

Cheers
Dennis

Post Reply

Return to “u-he”