Diva cutoff settings

[ho66it]

Urs,
If I am trying to enter settings to emulate a Memorymoog patch from days of yore, and the original setting was 53 on a range of 0-100; would I now enter the original value of 53, or do I need to enter 53% of 120? (Using the Diva ladder filter with cutoff range 0-120)

[Ch00rD]

From what I read earlier I deducted that the full range of the Minimoog filter modulation is 5V = 5 octaves, which would scale to a range of 60 on Diva [EDIT: link added], but I have no idea where on Diva's range the (thus exactly 1/2 smaller) Minimoog would be placed. A value of 64 seems to be MIDI note 60 (middle C), and it scales exactly with one semitone for each interval of 1.00 (so you might do it by ear, temporarily using a high resonance value to home in on the desired filter frequency if needed).

[ho66it]

From what I read earlier I deducted that the full range of the Minimoog filter modulation is 5V = 5 octaves, which would scale to a range of 60 on Diva, but I have no idea where on Diva's range the (thus exactly 1/2 smaller) Minimoog would be placed. A value of 64 seems to be MIDI note 60 (middle C), and it scales exactly with one semitone for each interval of 1.00 (so you might do it by ear, temporarily using a high resonance value to home in on the desired filter frequency if needed).

ch00rD,
Thank you for the advice. I'll give it a try!

[Gonga]

My understanding is that they will not scale precisely, partly because the filter is not a circuit emulation but a simulation which is based on a VCF / VCO combination. For example, Urs cautioned us against expecting to be able to use patch sheets of classic sounds partly for this reason. The best approach with Diva is indeed probably similar to that of using an authentic, unique analog synth - to use your ears and experiment - and this is how those classic patches were created in the first place.

Good luck!

[dmbaer]

I did some calibration measurements and found some interesting information. Key track when at 100 does 100% tracking. In other words, you can only track with the keyboard, not overshoot. The pivot note for the tracking is E below middle C.

I also did measurements on where the cutoff is. A pleasant surprise was that all three LPFs seemed to have the same cutoff for the 30 to 150 setting on the control. That means you should be able to audition different filters quickly without tweaking you settings.

If I have time this weekend, I'll post the cutoff values I recorded. I used the following methodology. I put Voxengo Span after DIVA. For each filter, I set resonance at max and played a cluster of notes in a lower keyboard range. For settings above about 55, the resonance peak was easy to spot. For lower settings, things started to get a bit murky and the measurements became more of a guess.

[Howard]

The best approach with Diva is indeed probably similar to that of using an authentic, unique analog synth - to use your ears and experiment - and this is how those classic patches were created in the first place. Good luck!

[mcnoone]

The best approach with Diva is indeed probably similar to that of using an authentic, unique analog synth - to use your ears and experiment - and this is how those classic patches were created in the first place. Good luck!

+1million.
That's all I ever do.
Specs, and analysis of rates, and technical stuff, gets me bored fast.
I need experiment, by doing things, (even the wrong things) to get anywhere with sounds.
In recreations...should be easier to do by ear.

[Ch00rD]

[...] you can only track with the keyboard, not overshoot.[...]

If you use the multiply or add function, you can get it to overshoot (keyfollow + keyfollow = 2 x keyfollow).

[...]For settings above about 55, the resonance peak was easy to spot. For lower settings, things started to get a bit murky and the measurements became more of a guess.

You could try to improve your measurements a bit by shifting the frequency spectrum upwards: record the notes, play themback at double speed, and divide the values you find by 2 (or even quadruple speed, divide by 4 etc.).

[dmbaer]

If you use the multiply or add function, you can get it to overshoot (keyfollow + keyfollow = 2 x keyfollow).

Yes, well if I knew what the multiply and/or add function was, I might have figured that out.

Much of the DIVA interface is quite intuitive, but some aspects are pretty inscrutible without a manual to learn from.

You could try to improve your measurements a bit by shifting the frequency spectrum upwards: record the notes, play themback at double speed, and divide the values you find by 2 (or even quadruple speed, divide by 4 etc.).

Nice idea!

But really, the lower settings are less important to get accuracy on IMO. I do agree that creating sounds should be done by letting the ear be the final judge. But there are a couple of situations where knowing the calibrations can be a big time-saver. I'm not an ace sound designer by any means. There are some sounds I'd like to port to DIVA from the NI Pro-53, for example. I can do this a whole lot faster if I can dial in some initial settings that are pretty close, so I did calibration measurements on the Pro-53 and then more recently on DIVA. I've also got a great book called "Fred Walsh's Synthesizer Cookbook". To create the sounds in his "recipies", again it really helps to get an initial version up quickly without a lot of experimentation needed.