Hello again,
The dithering process is related to the process of assigning a digital value to the analog incoming data. It is an intentionally applied form of noise, used to randomize quantization error. However, here it apparently creates 'spurious tones' due to 'not-fully-damped resonance in the feedback circuit'. On the Fourier graph (when you zoom in quite a lot at high frequencies) it results in extra sharp 'needles', mainly at multiples of eight, but also (less) at multiples of two. The needles do NOT diminish toward higher frequency (they are 'of nearly infinite width'), and therefore disturb more the higher one's 'treasure data' - which do succumb to the 1/f rule - reside. These technical details are abridged from Biosemi's responses to our questions upon encountering the issue.
It is not possible to filter this contamination out.
In my study the sampling rate was 16384 Hz (Biosemi's Mode 7) with 32+8 electrodes. I was looking (among other things) for brainstem FFR to 530 Hz pure (and stable) tones. Due to presenting the stimulus in alternate polarities (as recommended by the custom practice [but, btw, not by all researchers]), I was looking for the response (in the combined data of the two polarities) at 1060 Hz. This is definitely not a regular region of interest even to FFR researchers, who normally probe speech FFR, at about a quarter of this value; it would disturb them less (but perhaps worth checking anyway). Also, speech FFR would involve gliding frequencies, where you may also still be able to see the beautiful, flexible response. In my case it disturbed more. And, unfortunately, I found it after all the experiments had been conducted…
As I said, Biosemi seem to be on the verge of launching a new product which does not involve the same technology at all. This has been some time since they said that. Hopefully, this will happen soon.
Other than this issue, the data we collected with their equipment is beautiful… :)
Amos
P.S. Alain: Of course we first checked if it is a light bulb or some un-isolated cable etc or a bad electrode. But then Biosemi confirmed the problem resides in the ADC process.
keep in mind that in the Biosemi ActiveTwo system a not-so-widely-known problem exists, of the signal being contaminated by the dithering process in the Sigma-Delta ADC box, at frequency-multiples related to the sampling rate.I did not know about this. How high frequencies are we talking about here? Is there a thorough description of the problem somewhere?
I’m not worried, but it is good to know...
Thanks,Petter
On 18 Aug 2022, at 18:25 , Amos Boasson <agboas@xxxxxxxxx> wrote:
Dear Carlos, and list members:Carlos, keep in mind that in the Biosemi ActiveTwo system a not-so-widely-known problem exists, of the signal being contaminated by the dithering process in the Sigma-Delta ADC box, at frequency-multiples related to the sampling rate.This problem disturbs mainly if you are interested in spectrum analysis at higher frequencies, as for examples in studies of the Frequency Following Response (FFR), where presentation of the stimulus in alternate polarities leads you to look for a response at double the stimulus frequency. Because of the 1/f decrement, mechanical 'side-effects' infiltrating the signal, unremovable, are more and more disturbing. the higher your frequency region of interest.Biosemi intends to implement a whole different technology in their next model (ActiveThree?), which would eliminate this problem, but I have not seen it on their site yet.This problem may not disturb some kinds of analysis, but you may want to consider it.Good luck with your new lab!Sincerely, Amos Boasson
בתאריך יום ד׳, 17 באוג׳ 2022 ב-7:07 מאת Jan Schnupp <000000e042a1ec30-dmarc-request@xxxxxxxxxxxxxxx>:
I cannot recommend strongly enough AGAINST the EGI setup. I had inherited one of their systems. It was terribly buggy. Lost parts of the signal. Lost triggers. My PhD student lost many months trying to debug the device or find work-arounds. Their customer service was unresponsive. It was simply terrible. What rescued my student was that we were able to get an ANT EEGo system from a friend, which perhaps hasn't the greatest S/N but it was quite adequate, and at least it was relatively straight-forward to get to work as advertised and we got some nice studies done with it. ANT now also offer a sponge cap for their EEGo system which can reduce prep time. We ordered one but we haven't received it yet so I can't tell you whether the S/N is adequate.Under absolutely no circumstances would I ever again waste a single dollar or minute of my time on anything EGI.
Best wishes,
Jan
---------------------------------------Prof Jan Schnupp
City University of Hong Kong
Dept. of Neuroscience
On Mon, 15 Aug 2022 at 10:40, Petter Kallioinen <000001c5645d28b7-dmarc-request@xxxxxxxxxxxxxxx> wrote:
Dear Carlos,I have used neuroscan a little and EGI a lot, but I do prefer Biosemi. The system is more simple and modular than EGI. I have done about 250 recordings with 4-6 year olds in pre-schools using Biosemi. While I do think it is possible to have good signal in EGI, it is easier to troubleshoot bad signals in Biosemi, usually a single or a few bad electrodes that can be fixed with more gel. In EGI problems are often more vague and indirect. If you use Biosemi 64 channels with children, time of application can be an issue. Two well trained assistants could do it in 15-20 minutes, but is is still an issue. Depending on your research question 32 channels could be good enough and considerably faster.
Customer support is fast but brief with Biosemi, and a lot of times I have solved things based on forum discussions rather than direct support, I am still more satisfied with support from Biosemi compared to EGI. Biosemi is more to the point and technical, whereas EGI has a slow and somewhat bureaucratic support.
Signal quality should be better with Biosemi active electrodes compared to EGI, but I have actually not tested this thoroughly. When I have looked at amplitudes of MMN responses the effects have been quite similar between EGI and Biosemi. I was surprised by this but it could be a larger difference in paradigms with fewer events.
Best wishes!/Petter Kallioinen, EEG technician at linguistics dept, Stockholm University
On 12 Aug 2022, at 17:20 , Carlos Benitez-Barrera <000001c4037ab8ae-dmarc-request@xxxxxxxxxxxxxxx> wrote:
Dear Auditory List,I am planning on purchasing an EEG system for my new lab at UW-Madison, USA. I have been in contact with several EEG providers, and I’m still undecisive on which system would be best for my lab. I’m familiar with Neuroscan and EGI, but I’m actually leaning towards Biosemi or Brain Products (Brain and Vision in the US). I run simple auditory paradigms with children (ages 3 to 14) and adults. The main characteristics that I’m looking for are (not necessarily in order):
- Ease of use
- Prep cap time (important to minimize with children)
- Customer support
- Signal quality
I’m shooting for a 64-channel system. Also, I’m still hesitating between saline caps or gel caps. I heard that the Biosemi caps despite of being gel based are very fast to get going.Anyway, I’m just looking for some advice from any of you working with these systems. Any experiences or recommendations will help!Thank you in advance for your taking your time to reply!Sincerely,Carlos