[fabian]

Use the new rMEA package (available on CRAN) for easy checking of rawdata.

Cheers, Fabian

[fabian]

A beta-version of MEA 4 is now available.
The process of moving MEA to the APP-Store is highly complicated, and will not be completed in the near future. Simply use the download-form to get access to MEA 4.0b.

Thanks for your patience!

Feedback is always welcomed and appreciated.

Best, Fabian

[fabian]

Hi Briana,

flickering is a potentially difficult thing to deal with.
I assume that you do not see the flickering outside the program, because it is generated by the codec used for the video. This is usually not visible to the naked eye, but when loaded into MEA, the frame-differencing algorithm picks up these slight changes.

If this is a strong noise, it will definitely have an effect on the validity of your data.

I suggest the following:
Try adjusting the threshold in MEA, and see how this has an effect on the frame-differenced video (the black-and-white screen).

When you do not see the flickering anymore, maybe you can still use your videos.

Just make sure to check the raw data of MEA for signs of anomalies.

Best, Fabian

[fabian]

I am working on a new version of MEA that will be available on Apple’s APP-Store.
It will remain free, because I believe that MEA should be accessible to as many researchers as possible.

For suggestions/ideas, please feel free to comment here or to contact me directly via email.

Thanks and best wishes, Fabian

[fabian]

Hi Ryan,

please change your security settings in your system preferences. I have added a screen-shot for this setting to the top of this discussion/faq – page.

[fabian]

Hi Christoph,

unfortunately, I do not have the resources to implement this . I agree that such an option to individually edit ROIs would be very important.
One simple solution to the problem you describe is shown in the manual:
You can use one of the ROIs that are not needed for your analysis to cover/edit the relevant ROI (see pictures in the manual). However, this only works if you need less than 8 ROIs.

Best, Fabian

[fabian]

The hardware requirements are quite low:
For an efficient analysis, you just have to make sure that you are running MEA in 64-bit mode. (See the manual on how to adjust this).
Otherwise, since MEA 3.0, the program runs at a speed that allows MEA to capture each frame in sequence. Thus, on a slow computer, MEA will simply run slower. The speed of the analysis also essentially depends on the resolution of your source’s video:
High-resolution videos (19020 X 1080; HD) run much slower than e.g. files with a resolution of 640 X 480. The same is true for the frame-rate: A video-file with a high frame-rate (e.g. 50 fps) takes much longer than one with a lower rate.

If you have a lot of videos to analyze, you might want to consider recoding your files to a lower resolution with lower frame-rate. (e.g. to 640 X 480 at 10 fps).

Computers with SSD-drives generally run much faster.

[fabian]

Hi Paulo,
Yes, this is possible.
You could set up at least 4 ROIs (one for each person). Then you would analyze the synchrony between each possible pair or you could calculate a multiple corraltion coefficient for triads / quartett.

You should bear one thing in mind: The individual subjects cannot cross/invade each other’s ROIs. (i.e. they cannot move in front of each other). MEA cannot distinguish between subjects moving from one ROI to another. Subjects have to “remain” within their designated ROI.

[fabian]

Dear Asef,
The most convenient way is to use R.
Cross-correlation analysis may be done with e.g. the “ccf” function.
A scripted version for the synchrony-analysis in R will be published in the near future.
Best, Fabian

[fabian]

Hi Konrad,
thanks for this suggestion!

I hope that users will add their papers soon.
This could be made into a list somewhere easily visible on the page.

Best, Fabian

[Author]

Posts