Hello All,
I'm a new member here. I am looking to contribute to data collection. I also have an interest in spectroscopy. This isn't the same old CMOS v. CCD question; but more of a mentoring and workflow question.
I have the identical equipment to a mentor that I would be seeking out; a C8 (actually same vintage) and the ZWO ASI533MC (OSC). On the other hand, I have the opportunity to get a used CCD - the Starlight Xpress 694M with the ICX695AL progressive scan chip - and a set of UVBRI filters (I'd probably go with the Baader set). I am working my way through the AAVSO manuals - which are excellent.
I am coming over from astrophotography and I'm an avid observer (and budding "math guy"). I just want to get a workflow going and stick with it. I am driven to work hard, fail, succeed, and eventually make a contribution to this community.
Question: Go with the CCD and filters or stick with what my mentor has and go from there? I just seems that there is a push
Thanks much for any insights.
Tony
Hi Tony,
I think the bigger question with your two choices is: one-shot-color or monochrome with filters. For photometry, the latter is best: monochrome with filters. You have much more flexibility (you can still do astro-imaging with an LRGB set, or even photometric BVR filters), will end up with more standard photometry, have better spatial resolution and will in general have more throughput by going monochrome.
The secondary decision is CMOS vs. CCD. There, the choice is more difficult, as both work well. CMOS is the current leader because it is cheaper for the same sensor size and QE, but you can get CCD cameras second-hand that work nearly as well, usually have larger pixels, and always have 16-bit digitization.
The tertiary decision is getting an identical system as your mentor has. I'll bet your mentor can help you with either system.
Arne
You must…
Thanks, Arne.
You must have read my mind. I am considering the ZWO ASI533MM (CMOS) - monochrome - with the filters versus the CCD. As I am plodding through the AAVSO manuals, I am also keeping an eye on the boards and comments and it seems that contemporary monochrome CMOS can deliver as well as a CCD. (But, what do I know? Hence the question.) There just seems to be this undercurrent drift toward CMOS.
I have the ZWO ASI533MC (CMOS - One Shot Color). I like the square format and worked out fields of view with my optics (61mm refractor, 203mm SCT and a 235 SCT, all with excellent reducers). So, complimenting my gear with the same camera in monochrome would solve the "learning a new workflow" problem, not to mention being able to gather data with the ZWO ASI AIR. The ASI AIR does a remarkably good job at plate solving. While I can star-hop pretty well and have a fairly good knowledge of the sky, this really helps. As well, this particular CMOS chip is back illuminated and is free of amp glow. Notwithstanding, I always shoot calibration frames.
On the other hand, if CCD is the way to go due to pre-existing calibration data with AAVSO, and the purity of the data, I'll go with the CCD. I am told by the person offering to sell the CCD that AAVSO would more readily accept data from the CCD as this particular chip has been in use for some time. (That's what I'm told . . . . Somehow, I think I see some eyerolling out there.)
For what it's worth, I had dropped out of astronomy for a while because - just like photography clubs - it became "Equipment Wars" and was more about that than the science. I have owned many, many, many lenses, filters, scopes, gadgets and I am hoping to resolve on something that won't impede my growth and offer a reasonably good result until I know what I'm doing.
Thanks a bunch for your feedback!
Tony
Hi, Tony:
Welcome to the world of variable star observing!
The choice of monochrome vs color is *much* more important than the choice between CCD and CMOS. Earlier this year I was analyzing lightcurve scatter in the AAVSO database, and it's clear to me that it is possible to do excellent photometry with either a CCD or a CMOS camera. Experience, skill, discipline, following instructions, and meticulous calibration images make all the difference in the world, whether you're using CCD or CMOS.
CCD cameras do not have a data "purity" advantage over CMOS, and the only pre-existing calibration data that matters is a mix of the calibration that you perform specific to your camera serial number and the calibration (color transformation parameters) that you measure for your particular photometry filters.
Unless you're under pristine, really dark skies, your photometry is likely limited by skyglow (light pollution) noise rather than camera noise. So don't spend a lot of time agonizing over one camera vs another camera. Instead, invest your brain cells in getting more experience and learning the camera. Be careful about focal reducers (requires extra care during calibration) and choose your photometry filters carefully.
And ask lots of questions here in the forums.
- Mark (MMU)
Your…
Thanks, Mark.
Your statement, "Instead, invest your brain cells in getting more experience and learning the camera[,]" is precisely the statement that I was looking for. Like I mentioned, I want to settle on a workflow and master it. Frankly, I really want to work with the CMOS monochrome.
I appreciate you inviting me to ask questions. I will have many over time.
Thanks very much!
Tony
Dear All,
I currently use an ST-10 with Astrodon B and V filters and am quite happy working with these. However, the camera is old and so, one day, will die. I have spoken with several observers who say that monochrome CMOS is as good as CCD and more convenient. But, despite repeated requests to places like Diffraction Limited, I have yet to see a plot of the linearity of a CMOS camera. It should be easy to create.
So, if anyone on this group could supply such a plot, I would be most grateful.
Sorry to miss everyone at this year's meeting but I will not be able to make it.
Thanks,
Mark Spearman
(SIQ)
Mark,
Check out Brian Skiff's post several months ago about a paper with detailed analysis of several popular CMOS cameras. Arne Henden also has a thread with a number of test results for CMOS cameras.
Clearest skies,
Walt Cooney
Mark:
I can't make any general statement about the linearity of CMOS compared to the linearity of CCD cameras (I just don't have the data). However, I can compare the linearity of two specific cameras: my old ST-9 CCD camera vs my new QHY268M CMOS camera. The difference in linearity curves is so stark that there really isn't a comparison to make. I've put the plots onto Google Drive at https://drive.google.com/file/d/1Qfad_X5idDOGKvNlOaJW5v8kmAslQcuW/view?usp=drive_link.
What those plots show is the ST-9 ADU response vs flux curve varies over roughly a 30% band as you go from very low light levels to near saturation. The equivalent plot for the QHY268M showed less than 1.1% variation as you go from dark to near-saturation. And the difference is quite visible in my photometry. With the ST-9, when I plot all my Landolt standard field residuals, there's a distinct error trend that correlates with photometry SNR. With the QHY268M, the residuals are somewhat lower, but -- even better -- the residuals don't show that correlation with SNR.
Replacing the old CCD camera with a modern CMOS was the smartest move I've made in a long time.
Postscript: I've been re-reading what I originally wrote here, and realize I need to be more precise with my terminology, because I'm using two different ways to quantify nonlinearity. If you plot actual ADU vs "perfect ADU from a perfectly linear sensor", then the ST-9's nonlinearity amounts to variation in about a 7% band. However, what messes up photometry isn't that ADU delta directly; instead, it's the variation in sensor gain associated with the differences in ADU. (This is because we're performing differential photometry, where we assume a constant gain across all the pixels measuring our comparison star(s) and the pixels measuring our target stars.) The gain is directly related to the slope of the ADU vs flux curve. The 30% band that I quoted is the variation in the system gain (slope of the ADU vs flux -- actually, the inverse of that slope, since gain is measured in electrons per ADU) that results from the nonlinearity.
It doesn't change the takeaway message: The QHY268M sensor in my particular camera is far more linear than the sensor in my particular ST-9 camera.
- Mark