Affiliation
None
Sat, 11/13/2021 - 00:22
Can a CMOS camera be used to calculate star limb darkening and other measurements used in exoplanet research? I use this site to calculate limb darkening:
EXOFAST - Quadratic Limb Darkening (osu.edu)
If you are trying to calculate limb darkening coefficients for exoplanet modeling, there is no need for you to do that yourself. Rather, use the EXOFAST website you mention.
Thanks Dennis. I was under the impression that exoplanet work required us to calculate the star limb-darkening ourselves. A CCD camera vendor has been pushing me into accepting a CMOS camera instead of a CCD. I do not like that idea at all because a CCD can use the standard astronomical filters to do much more research in various fields. A CMOS is limited in that way. I would like my camera to be as versatile as possible so that my research will not be limited in any way.
I have been talking to the vendor about a CCD camera that I believe would be best for exoplanetary work, but the sensor may or may not be available right away. There could be a 6 to 9 month wait. They put out a request for availability of the e2v CCD42-40-1-368 (Midband) sensor to Teledyne/E2V yesterday (Friday). I should hear back on Monday or Tuesday.
If the wait is unacceptable I will probably accept the E2V CCD23042 (Midband) sensor which has larger pixels (15 microns rather than 13.5 microns). Both cameras have a 96% QE at about 650 nm and they are both sensitive well into the infrared. I like the smaller pixels however because they would spread the FWHM over a larger number of pixels in order to fit better within the 3 to 5 pixel optimal range. The drawback (and there always is one, right?) is that the camera with the smaller pixels has a significantly smaller Full-Well capacity of 100,000 electrons, vs. 150,000 electrons, making an exoplanet observation more difficult to obtain maximum SNR without taking the chance of saturation.
I hope to get this mess over with soon. Dealing with this vendor has been very stressful.
Ed:
Do you know that there are now many CMOS monochrome cameras available with filter wheels that will accept standard photometric filters?
Have you done the calculations to match camera pixel size with your scope and seeing (FWHM=2-3 pixels) at your site?
Ken
I pretty much know…
Ken,
I pretty much know nothing about CMOS cameras or how they work (as opposed to CCD cameras). I was not aware that there are monochrome CMOS cameras are now available. Or that they can be used with standard Johnson-Cousins or SDSS filters.
My previous CCD camera had 15 micron pixels which (assuming a FWHM of 3 arcseconds) produced a plate scale is 0.9 arcseconds/pixel or 3.3 pixels across the 3 arcsecond FWHM. That seemed a little low for me. What size pixel would you recommend as ideal for such a setup?
Would you recommend a CMOS or a CCD for exoplanet photometry? Could either be used with AstroImageJ?
I am in the market right now for a new camera to replace my old Apogee CCD camera, so any advice you can provide would be extremely welcome. The vendor is currently trying to have me consider an FLI KL400 with 11 micron pixels rather than a PL230-42 CCD or a ML4240 CCD. I do not know which one would be best for exoplanet work using AstroImageJ.
Ed:
1. Do a search on 'ccd vs cmos' on the aavso web page. Read the forum threads. You will pick up guidance on many of your questions.
2. Read the CCD Photometry Guide to get the best answer about how many pixels per FWHM (i.e., 2-3 pix). What is the typical seeing range at your site (3" +-)? Your current camera matches your seeing reasonably well right now but it could be slightly better. (Not so much to worry about it.) The rule of thumb about matching FWHM and pixel size/image scale are true for both CCD and CMOS cameras.
3. There is nothing inherently different with ccd vs cmos cameras that makes the latter unsuitable for photometry. New CMOS cameras often have less read noise and thermal noise than CCDs. That is a better thing for precise photometry like exoplanet imaging.
4. AIJ reads fits images. It does not know whether it came from a ccd or cmos camera, and doesn't care.
5. You must have discovered that CCDs are harder to purchase now. CMOS cameras are taking over the imaging market. If money is not an object, CCDs will continue to be available at the high end ($$$). More CMOS cameras are being produced for the photometry market. 16 bit cameras are replacing the 12-14 bit cameras. Remember that you can bin smaller pixels to match your seeing. However, keep in mind that smaller pixels do have more limited well depth and are often noisier. Not as critical now since most are adequate? Also, look for a new camera that has no/very little amplifier glow. This makes a difference in how you select exposures and whether you scale your darks or not. You can modify your procedures, so that either work. IMHO, a field of view of 30 arcmin is a sweet spot for photometry (different for astrophotography?). You can buy a very large chip if you really want to, but they may not be worth the extra cost.
6. Do some more reading and assess the comments that I hope others will provide. Do not trust only one comment including mine. CMOS imaging hardware is developing and improving rapidly. Alternatively, don't obsess about it forever, many CMOS cameras ($ to $$$) are suitable for variable star and exoplanet photometry.
Ken
Hi Ken
I thought I had done my research and spoken with several AAVSO folks and settled on a ASI294 MM Pro for variables and possibly Exoplanet transit timing. The 294 provides a FOV of 1.37x 0.94 ( in my 800 mm F4 newt), is 16 bit, has px scale at 1.19” and has a FW of 66k.
I just read your comment on FOV at 0.3 being a sweet spot for photometry. I realize that, unless the perfect camera is out there for my scope, there’s a balance I need to find among the various camera selection criteria. Any comment on how you’d rank the criteria : FOV, FW, bits, px scale, etc in order of importance? Many thanks for any input you can provide.
Gary
Gary:
I hesitate to give a hard and fast ranking. Lots of compromises can be made.
However, I would always try to match the image scale with your seeing range first. That said, under-sampling (FWHM<2 pixels) is worse than oversampling (FWHM>3). Try to be reasonably close to FWHM=2-3 pix. BTW, binning of small pixels can get you in the desired image scale range, and it will shrink the size of the image file/ download!.
At f/4, you will normally get a relatively larger FOV than with f/8 or f/10 focal ratio. My 'opinion' about a FOV of 30 arcmin is that if you have a FOV<15 arcmin, you may have difficulty finding good comps, and a really large FOV may not provide any more comps near the target than needed. My key point is that you don't need to buy the newest, biggest, most expensive camera to do effective photometry. As noted above, a larger FOV is better than a very small FOV, but it doesn't have to be humongous! That might just be wasted space and $$. Of course, if you want to study all the ZTF targets we see on many images now, bigger FOV will show more.
I prefer 16 bit cameras since they provide more dynamic range but 12-14 bit CMOS cameras do work and binning (through software, not actually on the chip) of CMOS pixels actually increases the effective bit size. BTW, the ASI294 camera is actually native 14 bit not 16 bit.
I prefer larger well depths but again it does not have to be 100,000 e-.
I'd say your camera is a good fit to your scope and seeing.
HTH, Ken
Hi Ken and thank you for your time and help with camera selection. I’ll dive into this a bit more now with your input.
In case you hadn’t delved into this camera, it has a 2 mode feature that seems appealing, at least on the surface - it’s binning is on-chip-switchable with the results summarized in the text below that I lifted from the ASI website for this camera:
Unlocked Bin1: 12bit ADC, 2.3um pixel size, 47 megapixels, 8288*5644 resolution, 14k full well capacity.
Bin2: 14bit ADC, 4.6um pixel size, 11.7 megapixels, 4144*2822 resolution, 66k full well capacity.
Bin 2 is the normal mode but can be switched to Bin 1 in Maximdl, NINA, Sharpcap, etc. There are some obvious trade-offs with the Bin 1 mode but it may provide some useful flexibility with resolution and ‘seeing’ in some situations. Forgive me if you’re already familiar with this device. In case you are not, there’s a link below to the ZWO website with all the detail.
thank you again for your help!
Gary
https://astronomy-imaging-camera.com/product/asi294mm-pro/comment-page-2?add-to-cart=21752
Hi Dennis
I’ve worked through the ‘Guide’s’ tutorial example using your sample data from a transit of WASP-12b to get experience with the steps and rhythm of the AIJ process. I had recently captured a large portion of a transit of WASP-11b so I thought it would be good to try to run my data through the whole AIJ process for practice. In filling out the Observation Worksheet for the LDC’s, I went to the EXOFAST link in the spreadsheet just above line 27 and input WASP-12b ( from the tutorial) to understand where the sample Worksheet got the quadratic LD u1 and u2 values. The EXOFAST query (WASP-12b and passband u) returned 2 numbers, similar in length, but quite different from the ones used in the Tutorial. So, now I’m confused and would appreciate guidance on how exactly, do I find the LDC coefficients for my observation data using the link to EXOFAST? I used a V filter. Thank you!
Gary
Hi Gary,
Not sure why you put a passband of u in for the band, since the example in the tutorial uses V, as specified in line 26 of the worksheet. In fact, when you put in WASP-12b for the planet, and V for the band, you do get exactly what is in the worksheet.
Hope this helps,
Dennis
Hi Dennis
I entered ‘u’ only because I noted that the output values on the two spreadsheet lines were labeled u1 and u2 - I had no other criteria for what to enter and I didn’t recall at the time reading that a V filter had been used in the sample data. Also, when ExoFast returns the two calculated numbers, there is no label saying they are u1 and u2… - now I know.
When you’re new to all this, every little uncertainty can become a snag in the learning process. For example, I easily calibrated my data yesterday ( really like AIJ approach a lot ) but the reduced images all lost the auto stretch AIJ initially applied, so in the dark images I could only see 5-6 of the brightest stars. Now I need to figure out how to get all the images stretched again so I can see all the prior stars, apply the apertures and process the images. I guess that’s my snag for this weekend.
Thank you for your note - enjoy a relaxing weekend…
Gary
Gary,
To get the auto-contrast working correctly, on the Image Display under the Contrast tab, make sure you have only the following selected: Auto brightness and contrast. And then click on Reset auto brightness and contrast to defaults.
Dennis
Perfect Dennis. Thank you…Gary