Wed, 05/06/2020 - 23:28
Hello! This month's Astronomy magazine Has Jerry Hubbell discussing high precision photometry (for exoplanets mostly) using defocus and a diffuser instrument called an Enginnered Diffuser, produced by RPC Photonics.
I was wondering if anyone has more information about this?
I am a bit confused. I thought high precision photometry is achieved by having 2 to 3 pixels per FWHM rather than diffusing/defocusing to a larger area, in accordance with the Nyquist theorem.
Thank you for your thoughts. Best regards.
Mike
This issue has been discussed, particularly with reference to 12 bit CMOS astronomical cameras, in the Forum 'CMOS Cameras for Photometry'.
The Nyquist criterion, as far as I understand, states that 'at least' 2 to 3 pixels per FWHM should be achieved. This is to avoid undersampling, which is bad. Oversampling is, in my opinion, not a problem, and under appropriate conditions can give very high precision.
See the followoing posts in 'CMOS Cameras for Photometry' (and probably others as well). I do stress, however, that I am referring to 12 bit cameras. I have no personal experience with using 16 bit CCD cameras.
Page 1, #6, 01/26/2020
Page 1, #29 02/06/2020
Page 1, #43 03/03/2020
Page 2, #54 12/13/2020
Page 2, #64 02/16/2020
Roy
The idea is, at least for bright targets in a high S/N regime, that if you can fill up more pixels with photons, in principle your internal precision will be higher. You might worry about how good the flat-fielding is, as well as other issues when lots of pixels are involved --- so you don't automatically win from this procedure.
\Brian
This technique works best if you can track accurately during the course of the observations. In other words, keep the star in the same pixels all the time. The better the tracking, the smaller any contributions from flatfield errors will be.
Good luck!
If you read my paper disccussing diffuser based photometrry we found that the diffuser mitigates scintillation to a larg degree and also residual errors due to drift.
Jerry - HGRA
You may be interested in the article at https://arxiv.org/abs/1905.02790 (authors include Hubbell and Conti, so you know it's good).
Bob B.
Thanks Mike for starting this thread. I have done a couple of presentations and videos talking about how diffusers work to increase the shot-noise SNR and provide a "seeing-limited" high-precision photometric measurement. The details of our work are detailed in our paper linked by Bob B. above. Please contact me at jhubbell@msroscience.org if you would like to discuss further.
Jerry - HGRA
Thank you!
I enjoyed the paper. The math was a bit in depth for me, but I was surprised by the different types things the could be controlled in order to improve precision. I had not thought about them before.
As I thought about this, it seems to me that defocus can improve precision for photometry, but the downside would be a decrease in limiting magnitude.
For example, if a target is critically sampled at a FWHM over two pixels and then the light is spread out over a larger area, say by a factor of 2.512 for easy comparison, then the limiting magnitude drops by a factor of one from what the limiting magnitude would be for critical sampling over 2 pixels.
Is this accurate?
If so, would a diffuser have the same effect on limiting magniude as defocus? Best regards.
Mike
I have just got around to reading the article. For the diffuser it states that it is fitted into a standard 1.25 inch filter cell. That suggests the photometry would have to be otherwise 'unfiltered' unless there was some way of 'piggy backing' a diffuser and standard photometric filters.
Roy