Issue with fourier.calibrate

[Yyuuuun]

I encounter an issue while using fourier.calibrate. Here is the code I run:

fs.fourier_calibrate(array_shape=20,array_pitch=20,plot=1)

The final output (the image) has a very unfitting format regarding the size of the camera's captor (1024,1024). Also, the zoom is very strange. See the image I provided. This image is fully black (I'm not sure why) but I had the same issue with an illuminated one.

[ichristen]

Hi @Yyuuuun ,

I think that your camera is underexposed and the calibration is trying to calibrate on noise. A few things to try:

• cam.get_image() after calling the calibration will capture the pattern that was used for calibration. cam.plot() will plot this and cam.save() will save it to an .h5 file which you can share.
• If you see an array of spots that you think should be able to be used for calibration, then you can remove the faulty calibration (del fs.calibrations["fourier"]) and recalibrate with the plot=2 debug keyword. This will give some extra plots of the FFT-based pitch calibration which are useful to debug.
• If you don't see an array of spots or your camera is underexposed, then you will need to debug your hardware before calibration. Make sure that you're able to see the 0th order of the SLM. You can use cam.live() for a live viewer inside jupyter.

Hope this helps!

Best,
Ian

[ichristen]

(But regardless, the fitting was obviously silly as it was so far off of the camera's field of view. We should add some checks for this.)