The Official DSI/Sequential Forum

Strange filter 2 behavior

Strange filter 2 behavior
« on: July 11, 2017, 04:15:57 PM »
I just recently acquired a Pro-2 (literally yesterday) and have noticed something odd about the filter 2 cutoff. When using the knob, I can't get the filter to close all of the way. Only from 5 o'clock to about 1 o'clock do I hear any major changes in the sound. From below there, there isn't really any effect. It still only sounds about halfway cutoff. With resonance cranked up, I can actually hear the filter opening back up slightly in that range! I took a really quick video of this behavior here: https://www.instagram.com/p/BWbI7qJHrmd/

I can still close the filter all the way by using the mod matrix or a negative envelope, but it still doesn't scale properly with the cutoff knob if I use that as a workaround. I have tried rebooting and calibrating the filters several times but that does not seem to help. In fact one time the cutoff behavior changed entirely, making it almost bi-polar. In that instance, around noon to 1 o'clock was fully closed, while 7 and 5 o'clock were fully open! That was super strange. Another calibration stopped that, but it's still slightly opening back up when it should be closing all the way.

I'm not sure what else to try at this point. Does not seem this is the way it should be operating. A few guys on the Pro-2 Facebook group said they were also having this issue, but not everyone. Can anyone here replicate the issue?

Edit: I am on 1.3
« Last Edit: July 11, 2017, 04:20:19 PM by particlesintowaves »

Re: Strange filter 2 behavior
« Reply #1 on: July 11, 2017, 04:32:39 PM »
Here's another video, this time showing the weird double polar behavior. Closed in the center, open on either side of the knob.

https://www.instagram.com/p/BWbMCMLHkDD/

Re: Strange filter 2 behavior
« Reply #2 on: July 11, 2017, 04:40:50 PM »
Please contact our dedicated support channel: support@davesmithinstruments.com. We'll get you sorted out.
Dave Smith Instruments

Re: Strange filter 2 behavior
« Reply #3 on: July 11, 2017, 07:32:59 PM »
Please contact our dedicated support channel: support@davesmithinstruments.com. We'll get you sorted out.

Thanks! I will do that

Re: Strange filter 2 behavior
« Reply #4 on: July 12, 2017, 12:18:36 AM »
My replacement board arrives today (assuming the UPS Nazis release it) so should be able to confirm the fix very soon.
Tip for anyone in the UK doing the same - send your board first and that way the return will be assumed 'not new' and therefore will not incur import charges.
Despite much effort, I was unable to convince UPS that this could be a repair situation unless I stuck to an order of events they could understand.

Re: Strange filter 2 behavior
« Reply #5 on: July 12, 2017, 08:31:37 AM »
New board installed, behaviour is exactly the same. Would be grateful if Support watched the video I did again and replicated steps on an actual Pro2. It demonstrates the steps from a Basic Program so should be easy to follow.

https://www.youtube.com/watch?v=2dL7WyzArVE

MDMA

Re: Strange filter 2 behavior
« Reply #6 on: July 12, 2017, 09:20:05 AM »
Great hand signals of despair!
DSI Rev2, Moog sub 37, Roland JDXA, DSI Pro 2, Novation Ultranova, Access Virus TI, Roland Juno 106, Roland Alpha Juno 2, Korg KingKorg

Re: Strange filter 2 behavior
« Reply #7 on: July 12, 2017, 09:30:13 AM »
Confirmed on my end as well. Filter 2 won't close all the way without additional modulation from a slider.

I'm convinced though that the F2 cutoff was working properly, or at least better than it was up until I ran the filter calibration yesterday.

Re: Strange filter 2 behavior
« Reply #8 on: July 12, 2017, 10:33:20 AM »
By any chance, is this a problem only when the filter Key Amount is being used?

Re: Strange filter 2 behavior
« Reply #9 on: July 12, 2017, 10:38:04 AM »
By any chance, is this a problem only when the filter Key Amount is being used?
This happens on an Init patch with no key tracking engaged

Re: Strange filter 2 behavior
« Reply #10 on: July 12, 2017, 10:50:44 AM »
That's wrong.

With keyboard tracking you sometimes have the use a DC offset to lower it but if it's doing it without and keyboard tracking it is ill.

Re: Strange filter 2 behavior
« Reply #11 on: July 12, 2017, 11:10:39 AM »
If you can't figure out the cause, contact DSI Support:

Phone: (415) 830-6393 x1
Fax (707) 286-5501
email: support@davesmithinstruments.com


Re: Strange filter 2 behavior
« Reply #12 on: July 12, 2017, 12:20:01 PM »
If you can't figure out the cause, contact DSI Support:

Phone: (415) 830-6393 x1
Fax (707) 286-5501
email: support@davesmithinstruments.com

Done that, will let you know how it goes. Someone just calibrated filters with cutoff at max on both and is now happy, didn't fix mine though.

Re: Strange filter 2 behavior
« Reply #13 on: July 12, 2017, 12:51:51 PM »
My filter 2 cutoff problem (mimicking BongMystic's) started after I recalibrated my filters to see if I could recreate what Particlesintowaves' filter calibration was doing. Lol.

This may help some people though.
I recalibrated my filters a second time, but with both the cutoff knobs cranked all the way open to 164. Now my Filter 2 cutoff is behaving the way I'm used to; there's an extremely slight bleed that can only be heard in headphones but on the PA, it's closed.

Re: Strange filter 2 behavior
« Reply #14 on: July 12, 2017, 02:47:57 PM »
I also just fixed my issue with a recalibration - except my F2 knob was at 0, not 164. Maybe try that too?

Re: Strange filter 2 behavior
« Reply #15 on: July 12, 2017, 06:19:44 PM »
Want to add that before I did this last calibration with the knob turned to zero, I ran the calibration clear function in the debug menu

Re: Strange filter 2 behavior
« Reply #16 on: July 14, 2017, 02:28:36 AM »
I too found that calibrating with the filter knobs at zero gave the best results. I also tried clearing the calibration and even initializing it, no messages on screen to say anything happened when taking the option though.
The filter isn't quite closed but I believe it's as good as it is gonna get, can live with it. No need to return main boards for this one folks (returning the one I got today).

Some speculation (for what it's worth):
Maybe the calibration routine affects only the pot range when it compensates for component differences in the filter circuitry. So rather than compensating for all modulation sources it may just add an offset to the cut-off pot's range. That offset then buggers up the overall range available to control from that pot. Since the filter's control CV is generated entirely by software this suggests a problem in the calibration coding. This may explain why, despite an apparent range of greater than 7-bit (0-127), the actual filter movement (for both VCFs) is not spread over the knob's full range of travel. Anyway, hey ho, back to making music. ;)

Re: Strange filter 2 behavior
« Reply #17 on: August 03, 2017, 02:53:23 AM »
Im so glad someone started this thread. Mine is doing exactly the same thing. Not the double open but not being closed at 0. Basically until i turn the F2 cutoff to about 90 nothing changes in sound!
Hardware or software issue?

Re: Strange filter 2 behavior
« Reply #18 on: August 03, 2017, 03:05:12 AM »
Want to add that before I did this last calibration with the knob turned to zero, I ran the calibration clear function in the debug menu

How do you access the debug menu?

Re: Strange filter 2 behavior
« Reply #19 on: August 03, 2017, 06:57:31 AM »
Im so glad someone started this thread. Mine is doing exactly the same thing. Not the double open but not being closed at 0. Basically until i turn the F2 cutoff to about 90 nothing changes in sound!
Hardware or software issue?

Bizarrely, DSI say this particular filter design they implemented will never close fully. The 'explanation' is that it's due to the architecture of the filter's electronics. A cynical person (not I) might say they spent too long getting their filters direct from Curtis chips... fortunately I like the synth enough to tolerate this unfortunate kink.