"Ghost" edits in note/pitch mode

[ScottParticipant]

Has anyone else encountered “Ghost” edits? A few times so far I’ve had this happen. I’ll be in note/pitch mode, and suddenly a step or group of adjacent steps will start to change by themselves. I can see the LED moving… even though I’m not touching that knob! The first time this happened, it was the last three steps (the three rightmost knobs on the bottom row). When I turned knob #31, the ones to the right and left started to move by themselves, but NOT in the same manner as the knob I was turning (they went their own ways, in different directions and at different speeds). It happened again last night, but this time with the first few steps (knobs 1 through 3).

I thought maybe there was a loop in the MIDI chain, but there was no MIDI in connected (only out A).

So far, this is only something that I encounter while in the note/pitch mode. I made edits in the note length and note velocity modes, but no “Ghosts” there.

I know Halloween is close, but…

Any ideas?

[ChristianKeymaster]

Hi Scott,
thanks for your detailed bug report!

First of all I´m sorry to hear that you´re encountering this problem.
During the two years of development and also the Beta phase we´ve never experienced this behaviour before. The only minor thing that can very seldomly happen is that an encoder is physically right on the border of registering the next turn increment, so it might jump 1 value. We´re already filtering the input to prevent errouneous or stray values, but can only go so far. But that´s it. From your description it sounded that the ‘ghost’ encoders are making bigger turns?

It sounds to me like it might actually be a hardware issue specific to your device, the multiplexer of the controls somehow acting up (somewhere along these lines: http://ricardo-dias.com/2013/07/19/repairing-a-behringer-bcf2000/). Do you have Ghost lights in the LCD display? Segments that slightly light up even though there is nothing displayed in them? This can be an indicator of problems with the multiplexer.

I will keep an eye out for this behaviour though and also am interested to see if more users experience this. Maybe you can also post back here after watching it for some more time?

EDIT: Also, do you use note name or note value display for notes? The multiplexer error is connected to the display as explained in the above link. Depending on what is shown in the display, it might occur or not. That might explain why it´s not in the other modes (velocity etc…)

• This reply was modified 10 years, 2 months ago by  Christian.

[ChaseParticipant]

This has happened to me twice.

Once with the demo, but it kind of went away and then I installed the full version. It was fine when I was using it, then I was showing it to my friend and it began acting up again, then it’s fine. Kind of strange.

For me, it was doing it in maybe not all modes, but at least chord and note delay.

It just kind of starts turning itself slowing increasing in value on some knobs. It seems like it’s different knobs each time. I guess if it does it again I’ll replace the displays.

[ChristianKeymaster]

Hi Christa, thanks for posting here.
Please note that while it may be connected to the LCDs, it´s not proven to be yet.
So, before desoldering your LCDs, you can actually measure the segments with a multimeter like described here: http://ricardo-dias.com/2013/07/19/repairing-a-behringer-bcf2000/

Quote: “You can do that by selecting the continuity check mode in the multimeter and activating a segment at a time… Testing each segment individually revealed a problem in the left display: two segments were dead. As these were connected to the same matrix of the buttons (in fact one of the common pins shared the same transistor of the faulty column), I carefully removed the left display and ran the tests with the top board connected.”

Please report back with your findings!

[PeterParticipant]

Hi, I noticed the same issue like other guys described above. For 2 weeks it was fine when I was using it and then suddenly I observed that on each knob the values are decreasing. It is very frustrating since I cannot select the right value. I detected it when I was trying to select a different value on clock divider knob. Later the strange behavior did occur on other knobs too.

[PeterParticipant]

Hi, I think the unit` strange behavior is more likely that values are decreasing in 1 value, but always end on a minimum value. E.g. i select clock divider value 32 and then it continuously decreased to value 1..

• This reply was modified 10 years, 1 month ago by  Peter.

[ChristianKeymaster]

Hi Peter,
thanks for the feedback and sorry to hear that your BCR is acting up.
We´re about to investigate this behaviour. It is only reported with a few devices so far, so it seems connected to the individual hardware, but maybe we can do something in the software to alleviate it.

In the meantime, have you tried to re-upload the Zaquencer firmware?
It´s a long shot, but at least one user has reported this to help with strange behaviour.

[PeterParticipant]

Yes. I tried it. It worked fine for 5-10 minutes and then the same did occur. I will try to reinitialize to original Behringer firmware to see whether all knobs are working there.

[PeterParticipant]

Weird it works again. Did you try it in more complex midi connection rings? It seems like as a standalone sequencer it works fine at least just now. When ghost edits did occur I had connected it as a 3rd device in my midi setup and it sequenced 2 sound modules.

[ChristianKeymaster]

Thanks for the update!
Yes, we´ve also tried in more complex setups.
We regularly have 4 synths connected to the BCR and extensively test both on internal and external clock with another device or computer giving the clock on the midi in.
We´ve also used it in clock chains of up to 4 devices, each one passing through the clock to the next.

I would like to get to the bottom of this behaviour, please keep me updated here when it happens again. The more detailed you describe the setup/circumstances, the better I can try to reproduce it here.

Thanks!

[ChaseParticipant]

Ok, it did it again. I recorded a 10 second video. Ignore my friend talking in it.

https://drive.google.com/file/d/0B22nHBQ6z84pNUFpQTRINm54cXVvQlRSWVpkeVdUSlZUcDN3/view?usp=sharing

So what you can see is the top 8 knobs with 3 of them ‘turning themselves’ in note mode.

I had just finished showing my friend how to changed all steps at once in a few different modes, then I did it in note mode and when I stopped we were talking and I noticed it turned itself.

[ChristianKeymaster]

Version 1.02 adds a filter for the encoder input to counteract the “ghost encoder” problem.

Wow, a lot of time went into this one.
We´ve found out that it´s a hardware error of a few BCR devices, where there is most likely some sort of crosstalk on the digital encoder lines.

We´ve added a filter to the encoders to remove erroneous input, so that the encoders should not move by themselves anymore.

[TheodorParticipant]

Hello Zaq and congrats for the continuous development of Zaquencer.

Small question based on this issue: provided that one is lucky enough to have a BCR which does not display the ghost encoder problem. I hope the new filtering feature does not give any sort of “overhead” of the processing or anything like that…correct?

(btw I have bumped indeed on that wrongly blinking LED on loading/storing even when it was on a different page)

In simple words, those who do not have the ghost encoder problem are they advised to still update the firmware? I guess so, because of the rest of bug fixes, but just making sure.

Thanks!

[ChristianKeymaster]

Hi Theodor,
I understand your concern.

The processing overhead for the filter should not be noticeable.
There was no indication of a processing speed problem during development
(the original Behringer firmware uses a similar mechanic btw).

Yes it´s advised for everyone to do the update eventually.
All future updates will also be based on this code now.

And lastly, going back and forth between Zaquencer versions is always possible.
If you should ever notice some problem after an update you can simply roll back to a previous version with no hassle at all (as long as the pattern parameters don´t change which they haven´t yet – and if they ever will be changed then only with a big warning being issued along with the update).

• This reply was modified 9 years, 9 months ago by  Christian.

[TheodorParticipant]

Being typically used to firmware updates not being able to roll back, somehow it slipped my mind that you CAN roll back to a previous Zaquencer firmware if something happens to go wrong.

Thanks once more, and a big respect for seeing Zaqaudio refining what was already to begin with, a superb product.

Needless to say Im posting around Zaquencer to anyone I know that might be interested! You deserve it!

[Author]

Posts