I'll check everything over right now to make sure I'm not missing something and try using 10-20 paste instead of velcro saline just to make sure. If it still doesn't work after that, then we'll get it sent to you.
Did you also try the test I suggested, with connecting three electrodes together, Bias, Reference (SRB2), and the channel(s) you are testing. This should result in zero microvolts.
Thanks so much for getting involved, both of you. @wjcroft and @biomurph I really really appreciate it.
So, I did some more tests today before we started the investigation/repair process...and we now have one of those classic situations where once the inspection/tech/repair guy gets involved, the device starts working perfectly.
?!?
I have no idea what I could have done differently from last time, but it appears to be working just fine now. ?????
The only thing I did differently was to try using 10-20 paste, instead of the velcro setup.
I tested Channel 2 first, just using the EKG wrist/arm setup, and it showed a normal rhythm.
Moved the lead wires to Channel 1 and was happily surprised to see Channel 1 repeat the rhythm.
@wjcroft, here I tried the test that you suggested. With just the SRB and Channel 1 touching the microvolts approached zero. Went I touched the BIAS, it went to "Near Railed" and went haywire. I'm assuming that was what was supposed to happen.
I then tried the whole thing over again, this time with the velcro setup.
Worked perfectly, yet again.
?
I have no idea what happened last time. Maybe there was some kind of dust or something on Channel 1's male connector pins? No clue.
But, so long as it doesn't happen again, we're good!
Thanks again for the help.
I'm working through the tutorials on the OpenVIBE website, now. Many new things to learn. Glad we can keep moving forward.
Well, the odd behavior is back, @biomurph and @wjcroft, but I think now that it's most likely me, not the machine part of it is me, and part of it might be the machine.
I'm not sure what I'm doing wrong though. I re-read through the tutorial and have read through everything said here, and can't seem to figure out what's going on.
Here's the deal.
Earlier today I discovered that the EKG signal works on all Channels...but at the time I did this I had no lead wires connected to the SRB 2 and BIAS channels and the daisy module was not connected.
Here's a picture of my setup when the device performs a working EKG.
With this simple setup I received these OpenBCI_GUI results (ignore the hash on the right and left sides of the display box, that was just me moving around to take the screenshot):
Just to make sure, I went through Channels 1-8 and verified that they are all working correctly.
Now, without changing anything else, I'm going to restart the OpenBCI_GUI software and this time I'm going to just change one thing: I'm going to insert the daisy module and start the system up using the full 16 channels.
Here's what the hardware looks like:
With that I started up the GUI, turned off all channels except for Channel 1, excluded Channel 1 from the BIAS and SRB @ channels (as described in the tutorial), and then shifted the PGA Gain down to x8.
Here's the result:
The EKG is still somewhere in there, it would seem, but there's that noise showing up again--the same repetitive noise that I saw earlier.
Nothing has changed other than the addition of the daisy module, the channels of which have all been turned off.
I've seen this kind of noise before and somehow got it to go away, so I'm assuming that there's something we can do here without having to send it in for repairs. Probably something simple.
What are your thoughts?
This is part one of the issue--that is, that the behavior of the OpenBCI board changes with the simple addition of the daisy module.
The second part of the issue is the fact that my board also behave unexpectedly when the SRB2 and BIAS channels are hooked up, even without the daisy module attached.
Take a look at this hardware setup.
I kept the same lead wires connected to my arm and wrist, but removed the daisy module.
Then I added the SRB2 and BIAS channels and hooked 'em up to my ear lobes.
(The reflective properties of light are pretty cool, eh?)
Now, with this changed, I restarted the GUI, turned off all Channels except for Channel 1, scaled down the PGA to x8, removed the SRB2 and BIAS channels from Channel 1, and then started the data stream.
Haywire.
Now, this is where things get REALLY weird...
I made only one change to the hardware setup: I moved the lead wires to Channel 2 and left everything else exactly as you saw it above.
Here's the software setup:
And the result:
?!?
So, Channel 1 behaves properly when SRB2 and BIAS are not hooked up to anything, but misbehaves once they are--even if Channel 1 excludes SRB2 and BIAS from its data stream.
But Channel 2 behaves properly in BOTH circumstances.
I'm starting to think that if there is a hardware issue, it's likely in the chip's connection between the BIAS/SRB2 connections...
Actually just had an idea and confirmed it.
I hooked up the EKG lead wires to Channel 1, left the BIAS wire attached, and then disconnected the SRB2 wire.
Result: Haywire.
Then I reattached SRB2 and disconnected the BIAS channel.
Result: Normal EKG.
It would seem that if there is a problem, it is between the BIAS channel and Channel 1.
@biomurph, when you take a look at this, take a look at that last part of the post:
I hooked up the EKG lead wires to Channel 1, left the BIAS wire attached, and then disconnected the SRB2 wire.
Result: Haywire.
Then I reattached SRB2 and disconnected the BIAS channel.
Result: Normal EKG.
It would seem that if there is a problem, it is between the BIAS channel and Channel 1.
This does not happen when I repeat the same experiment with Channel 2.
I tried it just now.
My eyes are starting to look like this: @_@ :having gone through so many different situations.
I did find that Channel 2 behaved normally both when the BIAS lead wire was connected and disconnected. However, it did behave--in just one instance, not all--with the same odd behavior of Channel 1, but not all. If I connect the BIAS lead wire and then tell the Data Stream to "Don't Include" the BIAS channel then I get haywire. However, if I keep the BIAS lead wire connected and then tell the software to "Include" the BIAS in the data stream, the EKG performs normally.
Channel 1 does not behave normally in either situation. In any circumstance where the BIAS lead wire is connected, Channel 1 goes haywire. In any circumstance where the BIAS lead wire is disconnected, Channel 1 behaves normally.
I can only imagine how confusing it must be to try to make sense of all of that!
Bottom line:
Channels 2-8 all behave in the same manner and can perform normally when the BIAS lead wire is connected. They do misbehave if I have the BIAS lead wire connected and the Data stream set to not include the BIAS, but in that instance I suspect it is due to my own lack of understanding of how the hardware and software work.
Channel 1, however will not behave under any circumstance where the BIAS lead wire is connected. Period. Doesn't matter whether the data stream is set to include or exclude the BIAS channel. But Channel 1 will do a proper EKG signal, so long as the BIAS lead wire is not connected. Period. Again, it doesn't matter whether the data stream is set to include or exclude the BIAS channel.
I do suspect that Channel 1 and the BIAS channel have something odd going on between them.
@bryanbeus, that is some extensive testing that you have performed! Thank you, and let it be a lesson for anyone with any question on any forum anywhere!
It does seem like you have narrowed the mal behavior down to a very specific setup and condition.
The only thing that I can think of trying before you send the unit in for review is to re-flash the firmware to the PIC32 chip.
Bryan, hi. I think you've said previously that you've exchanged some of your electrodes / cables, to eliminate that factor. Did you also confirm your Bias and SRB2 leads? All of the channel electrodes and Bias and reference normally need solid connections to get good data. I realize some of your measurements are using + - differential on a single channel. But check that Bias cable, or any cables that you might suspect by swapping for a known good cable. You can also use an ohmmeter to check electrode cables, as you wiggle around the connection points at either end (cup electrode or header pin connector or touchproof connector).
So cable checks and firmware re-flash, would be next logical step. Also suggest using paste for your tests until the glitch is isolated, just to remove that one extra variable.
The problem is not repeating itself consistently. At the moment the device is working properly. I'm going to move forward with the device if it can continue to be stable. If not, I'm going to mail it in for @biomurph to take a look.
Re-flashing the device does sound like a good idea. I may do that if the problem arises again.
For anyone who was curious about the details concerning Salt Lake Comic Con and a presentation there: I've spoken with the directors and we've agreed to move this back to April. With the technical slow-down here I'm going to need more time.
Bryan, hi. What is your goal at this point? Can you just go ahead with your intention to use OpenViBE and one of Jeremy's tutorials on the speller or motor imagery? You have 8 or 16 channels, so even if one of them might act strangely, you will likely have enough.
The ADS1299 and GUI is potentially daunting as far as the combinations of settings that are possible with Bias, differential signals with P/N both connected, etc. What if you just use the simplest default power up settings and say attach channels 1 and 2 (IN1N and IN2N) to O1 and O2; SRB2 to one ear; Bias to the other. Use paste just to be consistent with what Joel used. With that setup, can you get expected alpha increases on O1 and O2 with eyes closed?
The only time you should need to fiddle with the advanced settings on the Chan Set tab, is with your single differential channel that is measuring ECG. All other settings leave at default.
The Bias lead should always be connected somewhere on the body, normally one of the ears or at a midline site (Electro-Caps use AFz). If Bias is disconnected from the body, then the amplifier has no 'centering' for the differential signals it is measuring; Bias also is used to inject an inverse Common mode noise signal.
@wjcroft Thanks for the support. I really appreciate it.
I've spent a lot of time trying to figure this out, and am starting to think that while the cost barriers have come down, I'm not an advanced enough user to get far in this. Last week I went to my university library (I teach art at BYU) and went reading through as much as I could. Spent three or four hours. The amount of time it will take to actually get to the level I want is starting to seem out of reach, considering this is just a hobby.
At the moment I'm feeling that I should sell what I have and watch for something more user-friendly. Not entire settled on that, but that's the feeling.
Bryan, thanks. Your initial project goal of "moving the cursor with brainwaves" does indeed sound challenging as a first project. But if you scale that back to something along the lines of typical BCI paradigms or neurofeedback, etc. -- then you would have more examples and information on which to base your project.
There are a number of art related projects that have used the Muse to control interactive art works. But you could also use the OpenBCI for this, just as easily. A typical scenario is to setup some kind of classifier / design in a VPL like neuromore, then that outputs OSC Open Sound Controller packets to your driver program which controls the art piece. Packages such as VVVV and PD Pure Data / Max, can accept OSC packets.
A part of me feels like I'm in too deep. I listed my stuff on eBay. However, someone just made an offer to buy them that is basically what I'm asking, but I declined it. Just feeling too attached to the idea of mastering EEG, and using OpenBCI to doing so.
Now I'm coming back here and taking a look through, thinking to try starting again, and feeling once again completely overwhelmed. Been reading through articles throughout the internet for over two hours now, as well as searching around on the OpenBCI forum, and just not finding a place to grab a foothold.
I'm feeling quite frustrated with all of this, but not with a way that I would want to make anyone else feel frustrated. I would like to channel this frustration in a way that benefits the OpenBCI community, and also other people who are interested in the things that interest me. But I'm feeling quite lost, and also worried that people who answer here would feel that I'm too much of a beginner and that I'm wasting their time.
So, I'm going to write this all down anyway. Whatever time anyone can spare to help out will be appreciated, but I know you've got plenty of your own endeavors as well. So I won't be offended if nobody answers at all.
I appreciate those suggestions that you mentioned above, William--those of the artworks. However, the vocabulary you used to describe it was completely foreign to me.
Before going any further, let me make one point that I did not make clear earlier: trying to move the cursor with brainwaves was not meant as an end goal for my EEG endeavors, but as a beginning. It was, as I now see, quite naive. It would require an extensive knowledge of programming and OpenVIBE.
Rather, my actual end goal is to use OpenBCI in meditation. I meditate every day and have not missed in over six years. Establishing some kind of OpenBCI community wherein like-minded EEG meditators can share ideas and learn from each other would be absolutely amazing, and was my original end goal. I thought the cursor-brainwaves was a simpler place to start as it utilized only one area of the brain (that of motion control), and as an artist I had some cool ideas. Whoops.
If this discussion weren't useful for future beginners looking around here on the forum, I wouldn't post it. Would not want to waste your time. But I do think that, with a little help, I might be able to make a pathway through the entry-level forest of OpenBCI in which I currently find myself, and do so in such a way that someone else can benefit from it. I'm willing to give this another go.
The problem is, where in heaven's name should I start? Sorry for feeling frustrated; I'm not frustrated with anyone in the OpenBCI community, just with myself.
I can continue to work through this book, but the problem with it is that it primarily talks about how to use the programs. It is intense and long. It would take months to get through the whole thing, and as a beginner I can't tell if this is the wisest place to put my energy. Just knowing how to analyze the data doesn't help, either. As a meditator, what am I supposed to analyze?
I went through the first few videos. Problem with this one is that you're expected to already know how to code in MATLAB, and they make it clear in the introduction that it's not for beginners.
MATLAB is yet another topic that I keep finding myself getting directed to. As a faculty professor I can purchase it relatively cheap, or even find a place to use it on campus. But I get the impression that it does not work with real-time scenarios, wherein feedback is provided to the user actually during a meditation session. MATLAB instead seems to work only as an analytical tool for processing afterwards.
OpenVIBE seems like an amazing tool for real-time feedback, but after going through all the introductory tutorials I'm not sure what half the stuff is supposed to do, even though I now have an idea how to use it. What are filters for? Which kinds of math do I need to learn so as to put them together properly? And what would be the use?
Just sorting out where to get started seems like a full-time job. Are there other OpenBCI beginners who feel the same as I do? Or do the majority of people who purchase it already know where to start? Or are they at least connected with a program that does?
Here are two things that I would like to learn how to do with the OpenBCI headset. As a beginner, I don't have the experience to know which one is a good place to start, and which one will just keep a beginner spinning his/her wheels.
1 - Have the headset/computer tell me via sounds/chimes levels of brainwave activity during meditation. The sound could be continuous or could occur at milestones of deepening meditation.
2 - Study sleeping patterns as they relate to meditation.
If I might, here is the impetus for why I'm interested in EEG in the first place. Maybe this might be useful in telling me how I can spend my energy in such a way as to find results through OpenBCI.
Briefly: Nine years ago I suffered a severe mental breakdown due to stress working 90-100 hours a week in a mostly isolated environment. The level of severity was extreme. Words can't describe how extreme it was, only to say that it was about as bad as it can get without going permanently insane. The cons of the breakdown are obvious: mental breakdowns stink. And, on top of that, I began having severe lucid vivid nightmares. They are scarier than anything I can possibly describe. But there were good things that came out of it. For the first two years after the breakdown, because my mind had been "opened" so to speak in weird kind of way, I could meditate myself into states that are usually only accessible to highly advanced monks, yogis, etc. It was because of this that I was able to pull my mind back together. As my mind pulled back together, the access to those states gradually, mostly faded away.
There is a lot of negative stigma associated with mental breakdowns. I'm mostly okay talking about it, but I do find that people shy away from me once I tell them I've had a breakdown. This is quite frustrating, and has led to years of being "alone," or feeling like I'm not allowed to let people know who I am. My wonderful wife and the amazing psychiatrist I finally met are probably the only two people who have never run away from me for going through that. Even in meditation communities I sometimes feel uncomfortable, mostly because there people either don't understand what I've been through, and/or they want to add spiritualism to my experiences; while I don't have anything against my spiritualist friends at all, I just don't feel comfortable with spiritualism itself. I want to stay rooted in the practical, more scientific studies of meditation, and keep any spiritual journeys of life more on the personal level.
There were legion psychological problems directly after my breakdown, but the leftovers of the breakdown include two things, and they are neurological, and I hope to use OpenBCI to heal. The conditions that I have are extremely rare; my psychiatrist tells me that most people don't survive what I've been through with their sanity in tact, which means that people who can coherently relate to seem almost non-existent. I read about them in many articles online, but meeting them in-person just never seems to happen; meeting someone who uses EEG to heal from these things at times seems nigh impossible.
The first is a feeling in my chest that feels like a constant sense of painful anxiety, like a sliver between my skin and my heart; also a part of that is a feeling in my head that is as though I am being pulled in many directions at once. After years of enduring these, I have come to the conclusion that they are simply neurological damage left over from the extreme pressure that my body experienced during the breakdown.
Secondly are the lucid vivid nightmares. I take a medication that keeps them away, but if I go off of it, the nightmares come back after about two weeks. I am constantly stressed about the idea that if the world was ever to go through a serious collapse (earthquake, war, etc) and I was unable to get ahold of my medications, that I would soon be once again reduced to the nervous, useless wreck I was years ago.
Meditation is the only solution I have found--and is the reason why I am so motivated to do it. When I can (once frequently and now rarely) meditate myself into a deep state, I can actually meditate myself into a place that I can only describe as "beneath the pain." It's not an out-of-body experience or anything spiritual, but more a state of mind that can only be experienced. It's like being inside your own mind. An awareness within your own awareness. The pain swirls around me, but I am in an inner world of calm emptiness. And in my dreams, back when I achieved lucidity naturally, I could actually sit down and meditate within the nightmare.
However, it is more and more rare these days that I can achieve those states, and I just can't seem to have any luck in finding a community in real life.
My psychiatrist is completely on-board with me using EEG to work through these problems, and wants to get as involved as he has time for. The problem is that he has no experience in EEG, nor does he know anyone who does.
There is a wonderful meditation group in my nearest metropolis and I attend as frequently as I can, but they just do mindfulness things and don't get into the neuroscientific aspects of healing.
I would very, very much like to establish a community in OpenBCI that helps people who have suffered mental breakdowns to use OpenBCI technology to reduce stress and even to repair neurological damage, as I have experienced. I am motivated to get involved because I just can't seem to meet people in-person who understand what I've been through and want to use EEG to better this part of the human experience. OpenBCI seems to have a growing internet community and that's part of what originally attracted me to it--something where I didn't have to fly across the continent to be involved with others, and where we relied on scientific and verifiable methods to create new pathways for healing.
Anyway, just putting all this out there. Many assorted, jumbled thoughts. If anyone can help me direct this energy in such a way as to benefit OpenBCI while achieving some of my goals, I would love to hear it. If nobody responds, that's okay.
Comments
I'll check everything over right now to make sure I'm not missing something and try using 10-20 paste instead of velcro saline just to make sure. If it still doesn't work after that, then we'll get it sent to you.
Mucho appreciated.
So, I did some more tests today before we started the investigation/repair process...and we now have one of those classic situations where once the inspection/tech/repair guy gets involved, the device starts working perfectly.
?!?
I have no idea what I could have done differently from last time, but it appears to be working just fine now. ?????
The only thing I did differently was to try using 10-20 paste, instead of the velcro setup.
I tested Channel 2 first, just using the EKG wrist/arm setup, and it showed a normal rhythm.
Moved the lead wires to Channel 1 and was happily surprised to see Channel 1 repeat the rhythm.
@wjcroft, here I tried the test that you suggested. With just the SRB and Channel 1 touching the microvolts approached zero. Went I touched the BIAS, it went to "Near Railed" and went haywire. I'm assuming that was what was supposed to happen.
I then tried the whole thing over again, this time with the velcro setup.
Worked perfectly, yet again.
?
I have no idea what happened last time. Maybe there was some kind of dust or something on Channel 1's male connector pins? No clue.
But, so long as it doesn't happen again, we're good!
Thanks again for the help.
I'm working through the tutorials on the OpenVIBE website, now. Many new things to learn. Glad we can keep moving forward.
I'm not sure what I'm doing wrong though. I re-read through the tutorial and have read through everything said here, and can't seem to figure out what's going on.
Here's the deal.
Earlier today I discovered that the EKG signal works on all Channels...but at the time I did this I had no lead wires connected to the SRB 2 and BIAS channels and the daisy module was not connected.
[more coming in next post]
With this simple setup I received these OpenBCI_GUI results (ignore the hash on the right and left sides of the display box, that was just me moving around to take the screenshot):
Just to make sure, I went through Channels 1-8 and verified that they are all working correctly.
Now, without changing anything else, I'm going to restart the OpenBCI_GUI software and this time I'm going to just change one thing: I'm going to insert the daisy module and start the system up using the full 16 channels.
Here's what the hardware looks like:
With that I started up the GUI, turned off all channels except for Channel 1, excluded Channel 1 from the BIAS and SRB @ channels (as described in the tutorial), and then shifted the PGA Gain down to x8.
Here's the result:
The EKG is still somewhere in there, it would seem, but there's that noise showing up again--the same repetitive noise that I saw earlier.
Nothing has changed other than the addition of the daisy module, the channels of which have all been turned off.
I've seen this kind of noise before and somehow got it to go away, so I'm assuming that there's something we can do here without having to send it in for repairs. Probably something simple.
What are your thoughts?
This is part one of the issue--that is, that the behavior of the OpenBCI board changes with the simple addition of the daisy module.
The second part of the issue is the fact that my board also behave unexpectedly when the SRB2 and BIAS channels are hooked up, even without the daisy module attached.
Take a look at this hardware setup.
I kept the same lead wires connected to my arm and wrist, but removed the daisy module.
Then I added the SRB2 and BIAS channels and hooked 'em up to my ear lobes.
(The reflective properties of light are pretty cool, eh?)
Now, with this changed, I restarted the GUI, turned off all Channels except for Channel 1, scaled down the PGA to x8, removed the SRB2 and BIAS channels from Channel 1, and then started the data stream.
Haywire.
Now, this is where things get REALLY weird...
I made only one change to the hardware setup: I moved the lead wires to Channel 2 and left everything else exactly as you saw it above.
Here's the software setup:
And the result:
?!?
So, Channel 1 behaves properly when SRB2 and BIAS are not hooked up to anything, but misbehaves once they are--even if Channel 1 excludes SRB2 and BIAS from its data stream.
But Channel 2 behaves properly in BOTH circumstances.
I'm starting to think that if there is a hardware issue, it's likely in the chip's connection between the BIAS/SRB2 connections...
Actually just had an idea and confirmed it.
I hooked up the EKG lead wires to Channel 1, left the BIAS wire attached, and then disconnected the SRB2 wire.
Result: Haywire.
Then I reattached SRB2 and disconnected the BIAS channel.
Result: Normal EKG.
It would seem that if there is a problem, it is between the BIAS channel and Channel 1.
Does all this make sense?
I hooked up the EKG lead wires to Channel 1, left the BIAS wire attached, and then disconnected the SRB2 wire.
Result: Haywire.
Then I reattached SRB2 and disconnected the BIAS channel.
Result: Normal EKG.
It would seem that if there is a problem, it is between the BIAS channel and Channel 1.
This does not happen when I repeat the same experiment with Channel 2.
I tried it just now.
My eyes are starting to look like this: @_@ :having gone through so many different situations.
I did find that Channel 2 behaved normally both when the BIAS lead wire was connected and disconnected. However, it did behave--in just one instance, not all--with the same odd behavior of Channel 1, but not all.
If I connect the BIAS lead wire and then tell the Data Stream to "Don't Include" the BIAS channel then I get haywire. However, if I keep the BIAS lead wire connected and then tell the software to "Include" the BIAS in the data stream, the EKG performs normally.
Channel 1 does not behave normally in either situation. In any circumstance where the BIAS lead wire is connected, Channel 1 goes haywire. In any circumstance where the BIAS lead wire is disconnected, Channel 1 behaves normally.
I can only imagine how confusing it must be to try to make sense of all of that!
Bottom line:
Channels 2-8 all behave in the same manner and can perform normally when the BIAS lead wire is connected. They do misbehave if I have the BIAS lead wire connected and the Data stream set to not include the BIAS, but in that instance I suspect it is due to my own lack of understanding of how the hardware and software work.
Channel 1, however will not behave under any circumstance where the BIAS lead wire is connected. Period. Doesn't matter whether the data stream is set to include or exclude the BIAS channel. But Channel 1 will do a proper EKG signal, so long as the BIAS lead wire is not connected. Period. Again, it doesn't matter whether the data stream is set to include or exclude the BIAS channel.
I do suspect that Channel 1 and the BIAS channel have something odd going on between them.
So cable checks and firmware re-flash, would be next logical step. Also suggest using paste for your tests until the glitch is isolated, just to remove that one extra variable.
Thanks for getting back with me.
@biomurph and I have spoken a little via email.
The problem is not repeating itself consistently. At the moment the device is working properly. I'm going to move forward with the device if it can continue to be stable. If not, I'm going to mail it in for @biomurph to take a look.
Re-flashing the device does sound like a good idea. I may do that if the problem arises again.
For anyone who was curious about the details concerning Salt Lake Comic Con and a presentation there: I've spoken with the directors and we've agreed to move this back to April. With the technical slow-down here I'm going to need more time.
Fun stuff. More coming soon.
The ADS1299 and GUI is potentially daunting as far as the combinations of settings that are possible with Bias, differential signals with P/N both connected, etc. What if you just use the simplest default power up settings and say attach channels 1 and 2 (IN1N and IN2N) to O1 and O2; SRB2 to one ear; Bias to the other. Use paste just to be consistent with what Joel used. With that setup, can you get expected alpha increases on O1 and O2 with eyes closed?
The only time you should need to fiddle with the advanced settings on the Chan Set tab, is with your single differential channel that is measuring ECG. All other settings leave at default.
http://docs.openbci.com/tutorials/01-GettingStarted#getting-started-w-openbci-v-connect-yourself-to-openbci-4-launch-the-gui-and-adjust-your-channel-settings
The Bias lead should always be connected somewhere on the body, normally one of the ears or at a midline site (Electro-Caps use AFz). If Bias is disconnected from the body, then the amplifier has no 'centering' for the differential signals it is measuring; Bias also is used to inject an inverse Common mode noise signal.
There are a number of art related projects that have used the Muse to control interactive art works. But you could also use the OpenBCI for this, just as easily. A typical scenario is to setup some kind of classifier / design in a VPL like neuromore, then that outputs OSC Open Sound Controller packets to your driver program which controls the art piece. Packages such as VVVV and PD Pure Data / Max, can accept OSC packets.
William