I'm in the market for a cooled astronomy camera and I'm trying to figure out if I should go OSC or not. Problem is I have very limited access to dark sites; in general I get around 3-5 hours of imaging time per month. And based on what I've heard, and also what I've seen from Astrobin images, it seems like mono cameras require more integration time than OSCs. (Possibly a lot more time? Not sure.)
I'm looking at the ASI183mm-Pro for mono, and the ASI294MC-Pro for OSC. (I ruled out the ASI1600mm-Pro because of its microlensing diffraction.) I would use the camera with a Stellarvue SV70T telescope, which is 420mm FL and f/6 without a reducer, and 336mm FL and f/4.8 with a reducer.
My reasons against picking a monochrome camera:
- I image from areas with very low light pollution, so I wouldn't benefit from the monochrome camera reducing the effects of LP. (I image at a bortle 3 site and a Bortle 1.5 site)
- I have very limited time at dark sites, around 3-5 hours of imaging time a month.
- I'm not interested in narrowband so if I had a monochrome camera I'd just image in LRGB anyway.
- OSC is much less complex than mono, which would most likely lead to more integration time. (More time imaging and less time switching filters, refocusing, etc.)
My reasons for picking a monochrome camera:
- I could image in HaRGB, which would be nice
- I could gather Ha data from my backyard (Bortle 8)
- I most likely wouldn't want to upgrade cameras for a long time, as it seems that mono cameras give you a lot of room to grow
What do you all think? Also if I got anything wrong please let me know, I'm still very much a newbie when it comes to dedicated astronomy cameras.
Thank you for your time!
While OSC is fine if you image at dark sites, I once again feel the need to dispute that OSC is "much" less complex than mono. It is less complex, but the difficulties of imaging with a mono camera and a filter wheel are frequently overwrought. Lets just get down to the base case: OSC vs. Mono+RGB (EXcluding L). Three filters. Here are the additional complexities:
1) Buy and set up a filter wheel.
2) Operate the filter wheel.
3) Focus with each filter.
4) Integrate three images instead of one.
Ok, first off, #1 is super easy, and a once-and-done sort of thing. You buy it, you put the filters in, you attach it to your camera. You don't really have to do that again...not, at least, on a regular basis or anything like that. Maybe if you choose to add or change filters (which is an option, an added bonus, with mono...a choice you do not really have with OSC!)
Regarding #2. Unless you hope to image without a computer of any kind...which might be possible with a DSLR, but is not really an option for most AP OSC...then you need software to run your imaging sequences. You need to configure this software, you need to focus with OSC anyway, etc. Adding the operation of a filter wheel is, again, a once-and-done sort of thing. You install the drivers, you configure your imaging program. The only recurring factor here is acquiring data in three separate channels, than one sequence of OSC data. This is not hard, and minimally more complex than operating OSC with the same software.
Regarding #3. This is the one area that might indeed be more complex. You need to focus more often...which is actually pretty easy but wastes time. OR...you spend time, again mostly as a once-and-done sort of thing, upfront to configure your filters properly, including filter offsets. Most filter wheels, filter wheel drivers and imaging programs these days support some kind of "filter offset" focusing feature. This means determining how each filter differs in terms of focus (most scopes are not well corrected enough to focus all light at exactly the same point, so depending on the filter you have to adjust focus a little; this is, in fact, a BENEFIT of mono, in that you CAN control this...you have no control with OSC, since you acquire all channels all at once). There are some well defined procedures that can help you determine these offsets and configure them. Once configured, you can then focus once, periodically, with a particular filter (say G), then let the software/driver handle adjusting for the other filters via offset config. This then allows much easier operation, as you only need to focus about as often as you would with OSC, and it allows channel interleaving during acquisition. Rather than acquire all data on one filter, then another, you cycle through each filter and acquire all three channels simultaneously. More on this in a moment.
Regarding #4. If you can integrate one channel, you can integrate any number of channels. Pre-processing Mono data is actually a simpler process overall than it is for OSC. With OSC, you need at least one extra step, possibly two, in order to get the best results: deconvolution and bayer drizzling. This is on top of the core pre-processing steps: calibrate, register, integrate. Deconvolution is necessary to ensure proper registration of OSC data, bayer drizzing is highly recommended to get the best resolution and overall quality from OSC data (assuming you have enough frames...) Pre-processing mono data is actually pretty simple: Calibrate, Register, Integrate. Calibration can usually be done with a shared master dark, but different flats. There are programs (or scripts for PI) that simplify this process and can automatically match darks and flats to each channel for you. Registration can, and should, usually be done in bulk for all frames at once, to a single common reference frame. Integration is then done three times, one for each channel. In my experience...mono pre-processing actually goes faster than OSC pre-processing. OSC data, once demosaiced, tends to be significantly larger in terms of data and memory footprint, more data must be moved around at once per frame, which can slow things down. I've always been amazed at how quickly I can rip through pre-processing a ton of 20mp mono frames for three or four (or more) channels...vs. pre-processing 22mp OSC frames from my 5D III. The latter usually requires significantly more time to pre-process, and there is also the added step of drizzling to get the best resolution (something simply not necessary with IMX183 data, which is already well sampled most of the time and already at more optimal resolution.)
Processing mono RGB data only really requires one additional step: Combine the separate RGB channels into a single RGB image. Once you have the RGB image? You can process the same way you would process OSC...or, quite possibly, process with even fewer steps. Mono data is very clean, usually more detailed, doesn't suffer from the same kind of color noise issues that OSC cameras often do, and may not need the kind of extra sharpening or detail enhancement steps that may be required to pull out details with OSC data. With dark skies, these differences may be less than with more light polluted skies, but the differences still generally remain.
So the main complexity increase here is filters and a filter wheel, focusing, and acquisition with multiple filters. Most of the filter wheel stuff, as mentioned above, is a once-and-done sort of thing. The ongoing additional complexity is acquiring the data and focusing. For this, I highly recommend interleaving your acquisition across all filters simultaneously, rather than acquiring one filter at a time. This can be done with many acquisition programs these days, including SGP which is what I currently use, and what a lot of imagers use. Nina, Voyager, TSX and others should also support this. With interleaving, you acquire red, then green, then blue, or maybe a couple of each. THEN dither once. THEN periodically, say every 1C temp change, or every 30 minutes, focus. Then acquire all three channels again. This does a few things. One it limits overhead time. If you acquire one frame at a time then dither, then focus, you can easily waste 30-50% of each night. Not recommended. With filter offsets, you do not have to focus between each filter. When acquiring different filters one after the other, there is no need to dither between each...you can dither after each set of filters. This can greatly reduce the amount of time you spend dithering. Instead of say dithering every 2 minutes, you might dither every 6 minutes. Or if you intend to stack a lot of frames (hours of integration), you might even dither every 2 exposures per channel, meaning you could dither once ever 12 minutes (assuming 2m exposures)! If you focus every 30 minutes, then on a per-hour basis, you have 5 dithers and 2 focus routines. Dithers should not be longer than 10 seconds, so that is less than 1 minute total. Focus might take a couple of minutes each time. So you have overhead of about 5 minutes every 60...not bad at all!
There are then the general benefits of mono. Aside from the 100% fill factor, you actually have the ability to tune focus for each channel. Unlike OSC, where you simply have to focus as best as possible for all three channels concurrently, which could well means that one of the channels scatters more, with mono you can optimize focus for each channel, getting the smallest stars and the least scattering/dispersion possible. This improves IQ. You acquire all channels at every pixel on the sensor all the time. While you can use bayer drizzling to offset the spatial cost of a CFA, you have to actually do bayer drizzling. This is an additional pre-processing step that can be fairly time consuming, and is on top of the other additional pre-processing costs for OSC data. Drizzling cannot help with the focus limitations of OSC, though. With mono, you don't have to drizzle if you are already well sampled (and with tiny 2.4 micron pixels, it doesn't take much focal length to get very well sampled data), and you can focus optimally for every channel. Further, with mono, you have the option of doing Ha imaging, and doing it properly, benefiting from the full fill factor that a mono sensor offers. You can also add OIII, SII, NII, and even other channels if you so desire (if you are a planetary nebula imager, multiple narrow band channels including more exotic bands could be useful). Transmission with LRGB filter sets is also usually very high, well over 90% and in some cases over 97%. CFA filters in bayer sensors are often not as high, sometimes much lower. So there could potentially be a transmission cost to OSC as well.
Mono "sounds" scary and overly complex. But a lot of the complexities are things you only really have to endure once, then you can forget about them. I set up a filter wheel once a year ago...and imaged with it over and over and over. I only re-configured it once just a couple of days ago. I could easily use the current configuration for years, across multiple scopes and scope configurations. Previously, I had my filter wheel set up for about another year. It is a very infrequent operation to fiddle with a filter wheel. Some people configure them, put em on scopes in remote observatories, and don't touch them for years on end.
The hardest thing is focus and determining and configuring focus offsets. First, for focus, you should be using motorized focus that supports ASCOM or INDI and can be automated. For determining offsets, there are reliable procedures that can assist, and as long as you test offsets when the scope has reached equilibrium with the ambient temperature, the results are usually pretty repeatable. Pick a filter, say green. Focus (preferably with a reliable autofocus routine, but accurate focus with some kind of focus assist mask will also do.) Log the focus position and filter. Switch to your alternate filter, say red or blue (just one at a time). Focus. Log the focus position and filter. Switch to green, focus again, log. Switch to red, focus again, log. Do this about a dozen times. Average the green and red focus positions, and perhaps discard any obvious outliers. Calculate the difference between the two using green as the reference point (offset may be positive or negative!!) Repeat the process with blue. Set your filter offsets in your filter wheel driver or acquisition software. As long as you use this scope, camera, filter and filter wheels...your offsets should apply. So you don't have to repeat this process. You can now interleave exposures across channels as much as you want, and minimize your overhead, and focus as infrequently as necessary. Focus frequency here should be about the same as with OSC, in fact.
I'll also offer this. I've been imaging with mono for years now. Once you get automation set up...imaging with mono is no more complex and requires no more time than imaging with OSC. It's all automated. You pick a target, slew, center, configure a sequence...and go. From that point on, even though more moving parts are involved, mono doesn't require my attention any more than OSC does. Both are hands off. I can go to bed, and get up in the morning with a pile of data either way (barring weather). The key difference with mono vs. OSC, and this goes for both cooled AP OSC as well as DSLR, is that mono has consistently delivered better quality for me than OSC. I have dark site OSC data, and I really like a lot of the images I've acquired with OSC from dark sites. But I have no question that under the same conditions, the mono camera would have delivered higher resolution data, with lower noise, particularly lower color noise, more and sharper detail, and generally better contrast.