dark pixel indices for ST

[ericmorelle]

Hi Andreas,
sorry to bother you again but I managed to find the dark pixel indices with your hints in #192 and wanted to include them.

I also found some codes for the OBP2Protocol on the way that I included for potential future developement ;-)

[ap--]

No need to apologize! Thank you so much for contributing ❤️

2 quick questions before I merge this and prepare a new release:

1. One of the commands you added is GET_OPTICAL_DARK_PIXELS. Did you manage to match the range you set for the dark pixels to data returned from the spectrometer (by reading the packet capture)?

2. (related) are any of the pixels in the range you specified for the dark pixels constant, and don't change their value at all? (If so, those should be excluded)

Basically the readout from dark pixels should be independent of the incoming light but dependent on other noise sources (like temperature, and inherent noise of the electronics). Some pixels might be just pulled to ground and have a constant value, which is why those should be excluded from the range.

It should be easy to test, by just checking over multiple reads if the set of dark pixels you chose has any change in value at all.

Cheers,
Andreas 😄

[ericmorelle]

One of the commands you added is GET_OPTICAL_DARK_PIXELS. Did you manage to match the range you set for the dark pixels to data returned from the spectrometer (by reading the packet capture)?

Not yet. As I said: I'm not really experienced with USB communication. I was able to pinpoint the command and added it because I thought it might be handy in the future. Maybe I can have another look at it in the coming weeks.

(related) are any of the pixels in the range you specified for the dark pixels constant, and don't change their value at all? (If so, those should be excluded)

I checked that: none of the pixels (of the entire spectrum) are constant.

The device also provides a GET_TRANSITION_PIXELS (I hope I remember the name correctly) method that returns the indices of a few pixels between the actual active pixels and the dark pixels. I assume they are to some extend exposed to stray light and therefore not completely 'dark'. As you can imagine, I did not include those into the range for the dark pixels.

[ap--]

Perfect!

Thank you again for contributing!
I'll wait a bit with the new release to see if I can address #193 before