OpenViBE forum

Hi,
You have the box signal concatenation.
Thibaut

Hi, I can't help you. But for your compute time I have an advise : For me Matlab is not made for online processing. Python is a little better but if you wan't to make big process try to code a Box in C++. It's too hard if you never code in C++. But scripting langage are slower than other in general....

HI,
I can add : If you haven't this option. It's maybe because you have install python with anaconda or other all in one software, you must add to path python manually.
Thibaut

HI,
If box are not white, there is a problem....
Red, is for box doesn't exist. I think you haven't run the good designer, it's in dist/extra not in dist/designer.
Thibaut

Hi,
This bug can appear some times but I don't know why. The panel is only retract on right but when we don't know this it can be disturbing.
Thibaut

THe init --recursive is not enough go in each submodule and get last commit

Hi,
you must update each submodule to latest development commit
Thibaut
PS : Remove your build and dist directory to build from scratch

HI,
I think it depends your algorithm, you can use a band stop filter if your wan't but irt's useless.
Thibaut

HI,
What is your problem ? you haven't the box or the box doesn't work ?
when you launch OpenViBE Designer have you some warning or error message ?
Thibaut

Hi,
If you have a band pass of 0.5-30, you have only frequency between this 2 values so filter of 60Hz is useless.^^
The noise can come from your setup (maybe electromagnetic, move, not enough gel....)
Thibaut

Hi,
If you build from source, with the git source, you can go to the development branch, normally the bug of P33 boxes are fixed. Tell me if not.
Thibaut

HI, Je have made two hello world game on unity one to receive data from openvibe on to send to openvibe, I must make a third to have bidirectionnal transmission but not yet. This hello world use lsl more easy to use than TCP. You can check that at this address : https://gitlab.inria.fr/openvibe/unit...

Thanks Ettore, I think some change have break this box. I create an issue to notify all the new team.
Thibaut

Hi,
It's exactly the same scenario and dataset in input ? If yes, it's certainly a bug in new version. A lot of improvment were done and can break some stuff.
Thibaut

HI,
The input value must be a success rate for each class so a probability not a distance hyperplane
Thibaut