Diva 1.1 - Grand Performance Boost

up to 40% less CPU

Good news! Here's the Diva update you have been waiting for. We locked ourselves away in our lab and squeezed as many CPU cycles out of Diva as humanly possible, all for the sake of one thing: Performance!

Diva 1.1 is now ready for download on her brand new product page. As always with u-he products, the update is free. Simply download the appropriate installer, follow the instructions and you're good to go.

Note: We have also made subtle changes to the basic sound quality, improving the transient behaviour and adding some "sheen". Although these changes will certainly be recognized as improvements, we recommend that you render your existing projects before upgrading.

The most important changes are...

1. Multicore Support

Finally play as many voices as you like on selected systems. Requires multicore system with on-CPU memory interface e.g. i5, i7, Nehalem.

2. up to 40% less CPU usage on expensive presets

We have improved the feedback prediction algorithim in the filters so dramatically that highly resonant patches will now use up to 40% less CPU than older versions. Can't get that sound any faster now!

3. Improved sound quality

Diva 1.1 sounds like a freshly refurbished piece of analogue hardware. We've polished the attack transients and added a bit more sheen to the VCO models.

Actually, what happened is this: To improve on the crossmodulation behaviour of early betas, we abandoned the widely used miniBLEPs in favour of a version with a more accurate phase response. A few people noticed that Diva's release version (1.0) sounded a tad duller than the beta, so we have fixed that. Now she has all the glamour of those early betas, plus all the goodness that we added in the meantime.

4. Revised User Guide

More info, more detail, more tips / tutorials, more fun to read! (included in downloads)

Version history 1.0.1 to 1.1


- Multi-threaded operation allows use of multi-core capabilities
- In many cases, Diva's filters are now a lot faster


- White-noise DC-offset removed
- Panning inverted to make it correct
- Triggered notes sound more coherent
- Sonic improvements (more harmonics, less noise in VCF and synced oscillators)
- New feature "Transient Mode" can reset DC-blockers and filter states on attack


- Completely new MIDI-learn mode
- Presets can be revealed in Finder (OSX) or opened in Explorer (Windows)
- drag & drop presets in Cocoa view (Mac)
- Better file error messages for windows
- log written if a file named Diva.log exists on the desktop


- deadlocks that some experienced under Windows are gone
- no more unwanted legato mode after preset selection
- no more crashes after invalid file access
- All Notes Off releases current notes instead of killing them
- MIDI preset change no longer forgets modulator section
- Init preset is correct in the presence of MIDI Banks
- no more crashes after long silence

© 2011-2018 by Urs Heckmann | Imprint / Impressum | Press Resources | Privacy Statement / Datenschutzerklärung | Jobs
Highlight hotspots
Close X