HW CV Instrument Unstable Tuning Bug

[astromooses]

I've been trying to work with the HW CV Instrument device in Bitwig but have found it extremely unreliable in it's auto-tuning feature. Essentially I've noticed that every fresh start of Bitwig the device cannot get an accurate curve for the tuning unless I just brute force it for like 30 mins after which the device starts tuning reliably with repeat success. It seems pretty repeatable and documented in Bitwig Support Question pages that this is a bug that's lingered for multiple updates now. Not only that but certain rituals seem to affect the likelihood of tuning which can be anything from, attenuating to absurdly quiet levels, turning off the track's master output (most substantial difference in my experience), cranking oscillators to their maximum octave range, and among other things the occasional swearing at the electronics.

As I said after about 30 mins of brute force tuning, device instances suddenly can tune reliably at least until I close Bitwig and reopen it, after which the whole bug starts again. I've double checked in Ableton's Max for Live CV equivalent which nails the tuning first try every time so I doubt it's my setup at this point. Silent Way does not tune well in Ableton but still performs better than the Bitwig device. In addition Silent Way does not work at all in Bitwig because for whatever reason when I make an audio track to bus the output chains from Silent Way to respective ADAT outs they show signal activity in the selection menu but do not pass anything upon selection.

In terms of erroneous calibrations I've noticed Bitwig seems to struggle most with resolving lower octaves although when I have monitored the input the octaves appear to be striking true during the test.

Relevant BW Question Page: https://answers.bitwig.com/questions/12 ... ot-working

UPDATE: I may have found a sweet spot for accurate and reliable curves by maxing coarse and fine frequency settings, using saws, and leveling between -21db to -24db. This worked on both Intellijel and Make Noise Oscs including after application restart. I have a theory as to why this is but will report back later.

[Taika-Kim]

Hmm interesting, I will have to try that.

[Teezdalien]

Yes, it's not as intuitive as one would hope for it to be...
Personally I've had less trouble with Bitwig's Hardware CV device than using the Silent Way voice controller.

I found these points helpful to achieving quick and reliable callibrations and get roughly 7 octaves with my Serge NTO oscillators and haven't had much trouble once saved as presets. These are mostly similar to your suggestions, but I guess it may also depend on the hardware used to some extent.

- I typically had better results with triangle waveforms than other types.
- You will need to attenuate the audio-return signal level, though not a severely as you've suggested.
- Make sure the audio returning to the CV device in Bitwig IS NOT passed on through to the master output.
- I get best results with the oscillators tuned somewhere in the mid to high frequency range, haven't needed to set them at maximum

However I'm not running via ADAT to an ES-8/9. Instead I use my MOTU interface having DC coupled outs and floating-ring/tip cables to my modular. I'd also love to hear your thoughts and further theories on this procedure.

[Taika-Kim]

I've had huge problems with both FireWire and USB interfaces driving the ES3, with both 12V and 15V oscillators. The strange thing is, that I keep repeating the procedure, about one in eight runs or so produce a straight line, otherwise I might get a few octaves and then a sudden jump to zero. The procedure itself plays clean octaves, it's the BW analysis phase that causes problems somehow.

[pitchdrifter]

Edit: nevermind. I was complaining about something that was caused by the hardware, as it turns out, not Bitwig.
Also, in most cases bypassing Bitwig's automatic tuning function is the way to go.