system error in markers attechment

[sdendorfer (Sebastian Dendorfer)]

@juana_woo

Dear Sebastian,
I've got another question to ask:
as last time I once said as above quoted,
I once changed another smaller c3d file to check the memory problem,
It failed to load,
and reported the error:ERROR(SCR.EXP4) : E:/A..t/A..2/A..2/A..n/E..s/d..o/ModelSetup.any : '/' : Division by zero

This error points to :W = {1/(SampleFreq*0.5)*CutOffFrequency}; in the ModelSetup.any.
Then I set the SampleFreq to 200, it loaded successfully.
So does it mean that the SampleFreq is zero, and causes the ERROR(SCR.EXP4)? And how can I avoid this one since I did't change any parameters of the filter?
Thanks in advance.
Juana

Hi Juana,
it could mean that there is simply some data wrongly stated in your C3d file.
Best regards,
Sebastian

[juana_woo]

Dear Sebastian,
Long time since I discussed on the forum.
I’m still busy with my motion capturing data,
using the demo of ‘GaitLowerExtremity’ .
I’ve attached all markers on the skeleton,
and successfully load the main file.
Now what should I do next?
Should I run ‘RunMotionAndParameterOptimizationSequence’?
I’ve already take this operation, and it reports the error like this:

Position analysis failed : 239 independent constraints and 244 unknowns
Failed to resolve kinematic constraints. Newton relaxation too small.

It then gives a list of ‘Constraint no. * above error tolerance’.
So what should I do to solve this problem?

Thanks a lot!
Best wishes,
Yours sincerely,
Juana

[toerholm (Søren Tørholm)]

Hi Juana,

Yes this is the right operation to run.

If this happens in the very first time step it could mean that initially your model is too far away for the the recorded markers. Please try to manipulate the mannequin.any values to obtain a better starting guess so that each pair of the recorded markers and the markers attached to the bones comes closer…

It could also be that initial marker positions on the segments is not good enough, these can be changed in the ModelSetup.any file.

Best regards
Søren

[juana_woo]

Thank you Søren,
I’m very appreciated that you help me a lot.
Now I’m not very sure the accuracy of the markers that I’ve attached on the mannequin, since I do not know the exact coordinate position of the markers (some of them are bound into rigids) .
So how can I check which marker (or markers) has got the wrong place?

Best wishes,
Juana

[toerholm (Søren Tørholm)]

Hi Juana,

I would do this is in two steps, first change the mannequin joint angle values and pelvis positions so the segments are well aligned with the markers.

Then i would move the markers on the segments so that they visually match up in pairs. Usually it does not need to be very exact it depends on which marker it is and the marker setup in general.

It is important that the orientation of the pelvis is correct though, if the model faces opposite to the markers it may not work.

Best regards
Søren