Not the same result in .csv obtain with bonsai and with DLC

[tokioshima]

Hello everyone!

I post this message because when I try to run a video to plot with a DLC model in bonsai, I don't have the exact same result when I run the video analysis directly from DLC.
Maybe it's a problem in "MinConfidence" and "ScaleFactor" parameters that I don't well setup but, with the DLC user guide, I found 0.1 and 0.8 values for MinConfidence and ScaleFactor respectively and still don't have the same result for the same video between bonsai and DLC.

Have you some guess to what I can do to improve that?
Thanks in advance!