3D-calibration fitting function

[zitmen]

What steps will reproduce the problem?
1.Do a 3D bead calibration using thunderstorm method
2.
3.

What is the expected output? What do you see instead?
The defocus fitting is not very good on one of the two sigma(z).

What version of the product are you using? On what operating system?
1.3(2014-11-18)

Please provide any additional information below.

Is this common because of imperfect system or in fitting?

Original issue reported on code.google.com by novayang on 2015-03-03 15:55:26

---

• Attachment: Sigma-ftting.jpg

[zitmen]

The blue one should be better than this. It is not obvious to me why this happenned.
There are certain restrictions on parameters of the calibration curves to avoid overfitting.
It might cause this, I am not sure. What calibration method did you use? There are
two to choose from. If you can attach the raw data, I can try to replicate your result
and figure out where the problem is.

Original issue reported on code.google.com by zitmen on 2015-03-04 15:56:20

[zitmen]

I would try again with a new sample. To me this looks like a problem with the data,
especially since the red curve looks ok. hard to say if its the sample or the microscope
setup but it is easy to get bad data in these experiments.
-Guy

Original issue reported on code.google.com by guyhagen1 on 2015-03-19 22:11:06