Rigid body error plot

[elska_kaczmarek]

I assume that the error bars on the rigid body error plot are supposed to reflect the same thing as the s.d. written below the graph. But I've been comparing the errors across rigid bodies and noticed that in some the s.d. values are very similar, but the red bars are noticeably wider on one compared to the other (and it's not a result of the axes being different scales). For example, see the rigid body errors for the ceratohyal_left and ceratohyal_right in the attached XMA file. Do you know why this is?

[Benjamin Knorlein]

Hi Elska, The 2 s.d. are actually 2 different ones:

The one drawn is the s.d. per frame. Each frame has an error for each marker contributing to the overall error. The variation/s.d. of these errors per marker are then used to draw the graph. e.g. if you have a high s.d. here it might mean, e.g. that the rigid body does not fit well to the tracked markers and that there is a strong variation of the error across the markers for each (or this) frame.

The s.d. written below is the s.d. of the mean errors of the whole sequence. For this there is no distinction between individual markers of the marker but it only shows the variation of the errors per frame. E.g. if there is a high s.d. here that might mean that some frames are not accurately tracked.

Does that explanation make sense to you?

Best, Ben

[elska_kaczmarek]

Yes, this makes sense! Thanks for explaining.

[Benjamin Knorlein]

OK. I will close the issue then