Emotiv EPOC driver status

Making & changing hardware drivers in Acquisition Server
Nathali
Posts: 20
Joined: Thu Mar 17, 2011 4:47 pm

Re: Emotiv EPOC driver status

Post by Nathali »

continuation of previous post:
Attachments
the results of Emotiv (was started after exiting OpenVibe)
the results of Emotiv (was started after exiting OpenVibe)
Emotiv_res.jpg (48.25 KiB) Viewed 353070 times
scenarios, but as I've mentioned before, even the most simple one (client+display) doesn't return anything, parameters are by default now
scenarios, but as I've mentioned before, even the most simple one (client+display) doesn't return anything, parameters are by default now
scenarios.jpg (65.55 KiB) Viewed 353070 times

ddvlamin
Posts: 160
Joined: Thu Aug 13, 2009 8:39 am
Location: Ghent University
Contact:

Re: Emotiv EPOC driver status

Post by ddvlamin »

Dear Nathali,

I'm not sure what the problem is as I do not have an emotiv, but maybe lbonnet wants to see the log files in your dist\log directory, namely the files openvibe-acquisition-server.log and openvibe-designer.log

There is more information provided in those files (at least in my case). I'm just saying this so he could help you faster.

Did you by any chance change something to the configuration of the acquisition client block? And are the impedances good as I can't see from the above pic? For example if I do not have the ground connected well, all signals are zero too.

Best,
Dieter Devlaminck

Nathali
Posts: 20
Joined: Thu Mar 17, 2011 4:47 pm

Re: Emotiv EPOC driver status

Post by Nathali »

continuation of the previous post: log files included, but there is not a lot of information, anyway that could be useful. Thank you, Dieter.

I've tried to change the parameters of the boxes, of course, for my purposes, but for now they are default. Yesterday I tried both default and my configurations. The result of my configs is the same - 0s everywhere.

About the impedances, they are not always good, I mean green, but at start they are, and try to catch these states.

Thank you for a quick feedback,

Nathali
Attachments
openvibe-designer.log
(16.51 KiB) Downloaded 7989 times
openvibe-acquisition-server.log
(16.73 KiB) Downloaded 7739 times

lbonnet
Site Admin
Posts: 417
Joined: Wed Oct 07, 2009 12:11 pm

Re: Emotiv EPOC driver status

Post by lbonnet »

Hi Nathali,

Thanks for the files (and as Dieter said it's indeed better to have the log files, they carry more information).
However I didn't find any useful message...

I tried to redo the whole setup with my Emotiv, and for some times (the first 2 minutes) I got the same value all the time on every channel. Not 0, but 2.5 if I remember well.
Then I just tried few more times, connecting, disconnecting, etc. And it works nicely since then. I didn't touch anything else than "connect" and "play" buttons.

Maybe the connection with the headset was disrupted somehow... or the battery was too low.
But it definitely works that way, and this was the first and only time I saw this behaviour.
Other users recently reported me that this same setup was also working on their side, as expected.
So, what I have now:

1. Emotiv Developement Kit_v1.0.0.3-PREMIUM
2. src dated as 2011/04/11
3. WinXP sp3 on VM
4. usb dongle
I just don't know how this setup is supposed to work on a virtual machine. The Emotiv SDK is good, the sources too... but is the dongle working through a VM ?
I can test that, but I can't do it right now.

If anyone else manage to get it working in a VM, tell us please :)

anyway, can you try with a "real" windows ?

Now regarding your scenarios, they seems ok to me. I mean they should definitely "work", but you should take a look at the tutorial scenarios (here). For example :
- One acquisition client is enough. In the Designer, you can connect one box output to several box input.
- Spectral analysis should be done on a temporal window, using epoching boxes. look at the tutorial scenario "power-spectrum.xml"

Cheers
Laurent-
Follow us on twitter >> openvibebci

Checkout my (old) blog for some OpenViBE tips & tricks : here !

bxxworld
Posts: 5
Joined: Thu Apr 21, 2011 12:47 am

Re: Emotiv EPOC driver status

Post by bxxworld »

Hi everyone,

I've been trying to build the driver but has been encountering the exact failure ever since.. I'm using Windows 7 + VC++ 2010, and everything before compiling is okay, including installing SDK, run the dependency exe file, but once it comes to compilation, this fatal error would occur, shown as following. I really have no idea where could have possibly been wrong, any insight on this? Thanks in advance!

-bxxworld
Attachments
error.png
error.png (63.06 KiB) Viewed 353021 times

lbonnet
Site Admin
Posts: 417
Joined: Wed Oct 07, 2009 12:11 pm

Re: Emotiv EPOC driver status

Post by lbonnet »

Hi bxxworld and welcome on board !

Yesterday we merged some modifications of the Channel Selector and Reference Channel boxes.
(few words about it here)

The development was made on Linux and it seems like it needed few correction to work on windows 7 + VS 2010.

It should be good now (tested on win7/VS2010 and Ubuntu). Do an SVN update in your openvibe directory (right click, SVN update if you have tortoiseSVN). Then rebuild the software.


PS: I don't have the same warnings as you do about coding format... You have these ones for every file ?

Cheers !

Laurent-
Follow us on twitter >> openvibebci

Checkout my (old) blog for some OpenViBE tips & tricks : here !

Nathali
Posts: 20
Joined: Thu Mar 17, 2011 4:47 pm

Re: Emotiv EPOC driver status

Post by Nathali »

Laurent,

thank you for your advice about the power-spectrum.xml. About the VM - the dongle works with emotiv software without problems. I'll try to find the windows machine to test. And I'll keep trying to get some response from my VM too.

Thank you once again for your time and help.

Regards,
Nathali

bxxworld
Posts: 5
Joined: Thu Apr 21, 2011 12:47 am

Re: Emotiv EPOC driver status

Post by bxxworld »

Hey Laurant,

Thanks a lot for the work, it's finally WORKING!!! I guess the coding format issue arises from the OS, initially I was using a Chinese-based OS, after I switched to an English-based OS, the problem simply solves!

Thanks again!

Best,
bxxworld

toncho11
Posts: 124
Joined: Tue Apr 19, 2011 7:58 pm

Re: Emotiv EPOC driver status

Post by toncho11 »

Hi,

I am trying to display raw EEG signal from my Emotiv Epoch - Research edition.

1. Have compiled a new dist folder + the EPOC driver.
Everything seems normal.

2. I started the acquisition server, selected the diver -> connect -> play

And then I get:
Device drift is too high.
It starts to flash left-right.

3. In the Designer I set: 'acquisition server' connected to a 'basic display'

and in the end I get nothing after clicking play!

No signal is visualized - flat line for every channel!

Scenario is attached. Is this scenario fine indeed???
Attachments
DisplayEEG.xml
My scenario.
(6.35 KiB) Downloaded 7813 times

toncho11
Posts: 124
Joined: Tue Apr 19, 2011 7:58 pm

Re: Emotiv EPOC driver status

Post by toncho11 »

After I restarted the computer - it worked.

Probably when the Emotiv lose signal for a moment, one needs to restart both the acquisition server and the Designer to make it work again.

Cent
Posts: 9
Joined: Mon May 16, 2011 8:12 am

Re: Emotiv EPOC driver status

Post by Cent »

I have the same problem than Nathalie with the acquisition server for the emotiv EPOC.

The drift is too high and in the scenarios, the signal´s level for the sensors are flat.

Can anyone say me why?

lbonnet
Site Admin
Posts: 417
Joined: Wed Oct 07, 2009 12:11 pm

Re: Emotiv EPOC driver status

Post by lbonnet »

Hi Cent,
I have the same problem than Nathalie with the acquisition server for the emotiv EPOC.

The drift is too high and in the scenarios, the signal´s level for the sensors are flat.
The default drift tolerance is set to 2ms.
With a 128Hz sampling frequency, 1 sample corresponds to 1/128 ~ 7.8ms.
The consequence is a correction every time the device drifts with just 1 sample... something that is likely to happen with the EPOC
In the Acquisition server properties, you can adjust the tolerance to 10 ms for example, then drift correction won't occur much. However, be aware that doing so logically lowers the timing precision.
Anyway, this is definitely not the problem. Even with 2ms tolerance and many corrections, you should have some signal...

If you experience the same problem as Nathali, with the same exact setup I'm afraid I don't have much more solution to offer... It is still working on my side and others are reporting so.
The thing is I don't manage to reproduce this behavior you are observing...

Do you acquire the signal through a Virtual Machine too ?


Laurent-
Follow us on twitter >> openvibebci

Checkout my (old) blog for some OpenViBE tips & tricks : here !

Cent
Posts: 9
Joined: Mon May 16, 2011 8:12 am

Re: Emotiv EPOC driver status

Post by Cent »

I am using Windows 7 Professional, no virtual machine.

yrenard
Site Admin
Posts: 645
Joined: Fri Sep 01, 2006 3:39 pm
Contact:

Re: Emotiv EPOC driver status

Post by yrenard »

Dear Cent,

did you solve this problem ?

Yann

Cent
Posts: 9
Joined: Mon May 16, 2011 8:12 am

Re: Emotiv EPOC driver status

Post by Cent »

No, I have the same case always.

At the first time, I have signal, but in seconds the signal becomes high level and never varied.

Post Reply