.bvh transfer error

Hello AnyBody,

we are doing a research project with an interface between AnyBody and the human model from the software ema Work Designer where a .bvh file from ema is supposed to be transfered into AnyBody. It used to work perfectly in 2019 when it was created with the initial research project but now it does not work anymore.

In a .bat file where the initial transfer from the .bvh (exported from ema) into AnyBody is executed we get this error message:

"TrialSpecificData.any(22) : 'PelvisPos.Data' : The matrix must contain the same number of columns as T has elements, i.e., the number of data and parameter values must match"

Here is a screenshot from the TrialSpecificData.any:

And from the .bvh file:
image
...
image ...
...

We have mapped the body parts of both models to each other in the MarkerProtocol.any, so that should be correct. I am not sure what other specification I have to do in PelvisPos.Data, it used to work exactly like that.

I am using AnyBody V7.4.28845.

Thanks in advance!

Best regards
Juliette

Hi Juliette,

The error describes a mismatch between the expected and input data format. The Data entry in the AnyFunInterpol must be transposed. Please try this:

Data = {Motion'[0],Motion'[1],Motion'[2]};

I hope this will fix it.

Best regards,
Dave

Hi Dave,

that solved it, thank you very much!

Best regards,
Juliette

Hello,

I have encountered another problem:


The model is far from the bvh data at the knees and head/upper back.
Also the left knee can't hold itself as straight as the right one during the movement:
image

I think there is something wrong with the transfer of the bvh data to the human model, should I change something in the MarkerProtocol_BVH.any file (if yes, sRelOpt or sRelOnBVH?) Or at some other point?

Thanks in advance!

Best regards,
Juliette

Hi Juliette,

This indeed looks very odd. I would definitely recommend you double check the marker protocol file:
the sRel values (both sRelOpt and sRelOnBVH) and also the marker placement on the human model and the bvh model. Are they on the corresponding segments? Maybe the left knee marker is being driven to the right leg on the bvh model?

From the first image, it looks like the marker just below the chin is somehow coinciding with a marker attached to the bvh arm model, but it could just be my false interpretation from the view.

Another aspect that you may want to look into is if there are some extra drivers in the model that are preventing a good solution.

One last suggestion could be to check if improving the load time position of the manikin helps the solver. Maybe the load-time position was a bit off and the solver ended up with a strange solution. If you can try to change the load-time position or change the starting frame of the analysis, it may solve this problem. Instead, if the model is consistently resulting in this strange solution, then there are some other issues that I mentioned before.

Another tip for you could be to visualize the bvh stick figure model. You can find the BVH_StickFigure study in the Studies folder. Please run this study to be sure that the raw data makes sense.

I hope this helps

Best regards,
Dave

This topic was automatically closed 125 days after the last reply. New replies are no longer allowed.