Agile DSLR Astrophoto Imaging & Processing
Posted 08 August 2012 - 07:09 AM
With the advent of newer DSLR cameras (for example, Canon 60Da, and I don't mean to make the case for it, just using it as a good example), I am feeling rapid imaging/processing is becoming possible without overly long integration times, and without overly exhaustive & complicated processing techniques.
What are your thoughts about this changing dynamics of DSLR imaging? And I don't mean to devaluate long integration times and extensive post-processing in anyway; I am just wondering given the right mix (the camera, the scope, sky conditions, etc.), it may be possible to reach "relatively close” to the perfect image quality that once took hour upon hours of imaging/post-processing?
When it comes to agility, DSLRs are already a step ahead of monochrome CCDs where they cut down individual channel times. I am just thinking why not make DSLR astrophotography even faster, with the newer technology we have at our disposal.
All schools of thought are welcome, for or against this concept; if you happen to be for it, we could expand it later into imaging/processing techniques that may help achieve this goal. Your thoughts?
Posted 08 August 2012 - 11:19 AM
Posted 08 August 2012 - 12:41 PM
We are still regulated to long intergration times to get every bit of starlight we can to make a noise free image after you process all of the sub frames, darks, flats and such. Software had automated some of this and has made this easier but you still have to pay your dues getting the images.
The right mix of equipment is something we all strive for. The perfect scope, sky and camera is something most people never obtain.
I live on the East coast. I suffer from humid summers, light pollution and the jetstream. So I will never have the best sky.
The perfect scope is out of my budget so I work with what I have. Swaping out different cameras does not make my imaging time shorter. I then have to take the hard drive full of images and process them.
You will need to define the term "faster".
Posted 08 August 2012 - 01:28 PM
Posted 08 August 2012 - 05:08 PM
Posted 10 August 2012 - 04:57 PM
Posted 10 August 2012 - 05:17 PM
I would have to agree about with you keeping things simple but I've taken it to the extreme of shooting single exposures of 3 mins with a modded 5DII and an FSQED at F3.6 no flats just auto dark and although the guys that shoot one object all night then process it all day get a better result for show piece stuff I don't think it’s worth it, the really faint stuff is another matter.
You got it; this IS the kind of concept I had in mind for this discussion knowing techniques will vary depending upon the kind of scope/camera one has. Since I don't tend to expose too many frames, I mostly end up taking auto darks. Your image looks great for 3mins.
Posted 10 August 2012 - 11:37 PM
Posted 11 August 2012 - 02:42 AM
The other night I shot a lot of the objects in Robert Reeves and Verinbergs books and ACTUALY HAD FUN IMAGING ! remember when this hoby was fun and not just for the computer boffins .
If any thing my 3 min shots were better than they did with thier Schmidt cameras.
I think I'm becoming some sort of a ludite but isn't having fun the reason were in this hobby ?
More power to people that can spend hour after hour on one object but for now I enjoy doing my " one shot wonders "
this one an 8min shot of the Key Hole with 12" RC
Posted 11 August 2012 - 06:47 AM
Posted 11 August 2012 - 09:02 AM
Posted 11 August 2012 - 09:06 AM
Posted 11 August 2012 - 11:21 AM
In terms of "faster capturing" the newer camera's do allow for higher ISO settings which increases the gain as in a webcam. But the real items that would allow for faster recording of faint light with lower noise levels would be larger aperture telescopes with fast f/ratios. The larger the scope/F/ratio, the faster ability to go deep in short order. I sometimes image through a Meade 16" F/10 system but I incorporate a focal reducer to speed up imaging by 4x with a much larger field. I took a single image at this setup of a comet at ISO 3200 and I did have to use some noise reduction applications to get it comparable to a multi-stacked image. For a single one minute exposure it came out very good.
I think it is photographically fast optical systems combining aperture with ever lower noise DSLR's, in good sky conditions that is the core of the principle here. However, I do not think the term "agile" describes the process at all. (The OP may be deriving it from a software development methodology emphasizing iteration and collaboration of which fast turn around is only a component) I would suggest a cut-off, that if it takes more than 60 seconds to generate a good quality image it is taking too long. In my view, there should be no or little post-processing and no agility in techniques required. I think what we are aiming for is "express" or even "instant" astrophotography, and it should be one-shot, fast, and deliver heavy doses of fun and instant gratification straight out of the camera such as my 15 second exposures in an earlier thread.
Of course this works for bright eye-candy objects in dark skies. As soon as you are chasing faint objects and going deep, exposure times will increase again but thats not the point. The point is to have fun in generating great results of those objects that do lie within the exposure time threshold - straight out of the camera.
One other aspect of cameras like the Canon 60Da is that it in terms of apparent resolution (for a given telescope/camera combination) it only has 4nm pixels. Its bayer quadruplet has equivalent resolution to most mono CCD cameras in the 7nm to 9nm pixel range such as my SBIG STL 6303e CCD camera. (If not in QE or depth etc)
Posted 11 August 2012 - 11:30 AM
Posted 11 August 2012 - 11:38 AM
Chris, your 15 seconds images are outstanding to say the least, but another point which needs to be factored in is the amount of equipment to gel with being an instant astrophotographer. For me, obtaining an image with minimal equipment in the fastest possible times is what I would describe as instant gratification. I have seen some images with just 10 seconds exposures wide field which baffles me still. I wish I had those type of skies, another thing to be factored in.
I would agree that minimal/simple equipment or at least easy portability is a requirement. Most of don't live in very dark skies (I certainly don't) In my situation I am going to argue that a simple 24" reflector is minimal equipment! (I can transport, setup, and align it, much more quickly than a standard astrophotography rig with scope, equatorial mount, counterweights, pier, camera, guider, computer, laptop, cables, power, yawn...)
Posted 11 August 2012 - 02:28 PM
I would suggest a cut-off, that if it takes more than 60 seconds to generate a good quality image it is taking too long. In my view, there should be no or little post-processing and no agility in techniques required.
What I am intending from "Agile DSLR Astrophoto Imaging & Processing" is NOT necessarily one-shot imaging. Since imaging agility will vary given the scope/camera/sky conditions, the idea is to produce decent images in less time. That less time, for example, may be 10 min (combined) for 24” scope, 30 min (combined) for 10” scope, and may be 45 min (combined) for 5” scope, etc. Note: These times are just examples, not actual recommendations. So time scale will vary a little given the variables, but concept remains the same, NOT having to do hours upon hours of exposing and everyone finding their own imaging agility parameter (shorter the better).
And yes, for processing agility, I am suggesting little/quick post processing (basic stacking, basic processing, etc.) but WITHOUT major/time consuming post processing techniques.
Keep your ideas coming...
Posted 11 August 2012 - 03:16 PM
How about DeepSkyStacker Live?... Is that what you have in mind?
Thanks for your input; concept again I have in mind is everyone using tools they like/prefer but doing little post-processing, not just to save time but to avoid time consuming tweaking/altering of the original/combined image (I see some folks doing processing to the point image starts looking synthetic). I, for example, use ImagesPlus, but use it to quickly combine, do basic stretching/smoothing and get done in few minutes.
Posted 11 August 2012 - 04:44 PM
Posted 11 August 2012 - 05:11 PM
I spend less than 10 minutes on post-processing and I always finish my image(s) before going to bed. I usually toss the subs as soon as I stack to save space on my laptop.
I would call it quite 'Agile' post-processing; that's partly the idea.
Posted 11 August 2012 - 09:16 PM
I find the tracking on the mount is the limiting factor ( unguided ) 3mins on the AP600E , if I have to resort to autoguiding the camera noise is the limit which I thought was 10 min but will have to experiment.
Maybe if we call it Minimalist or Fast Track imaging would cover what we are trying to do.
As some have pointed out I have the advantage of dark skies down here , but like most things in life , Its what you do with what you've got .
Number one thing is to have Fun !
This shot M83 10min exposure and about 10min post processing.
Posted 12 August 2012 - 01:10 AM
Maybe if we call it Minimalist or Fast Track imaging would cover what we are trying to do.
Or we could include what you are suggesting into the definition.
Great looking image of M83; I presume it is single 10 min exposure?
Posted 12 August 2012 - 02:05 AM
I took this photo with a Pentax K-5 (which doesn't have the aggressive IR filter Canon and Nikon are in love with), and a C11 using hyperstar.
The frames are 30 second subs, and you can see a lot of the dim hydrogen clouds are coming through. For this fall, I still want to go to longer exposures to get cleaner data on the really deep exposure portion.
Because of that, I have to disagree with part of the supposition of short exposures being a 90% solution. The longer exposures still have less fixed pattern noise, and the cameras really only reward low f/# imaging trains.
Posted 12 August 2012 - 05:00 AM
But the big one is available optics to go to lower f/#s.
...I have to disagree with part of the supposition of short exposures being a 90% solution. The longer exposures still have less fixed pattern noise, and the cameras really only reward low f/# imaging trains.
Welcome to the discussion Rich; I don't think anyone is suggesting short exposures (< 1min) per se if I understood you correctly; although folks with really large apertures (e.g., 24”) have tried sub-1min exposures and I think rightly so.
Partly the idea is to reduce overall integration time, regardless of individual exposures which are going to vary given the equipment and type of the object. So yes, longer exposures with shorter overall integration time are part of the ‘Agile’ equation just as shorter exposures with shorter overall integration time are.
Lower f# is a good point you mention; I'll let folks chime in on that.
Posted 12 August 2012 - 08:39 AM
What you are proposing fits that existing model precisely: short subs, low total integration times, minimal processing and a "natural" look.
Whatever the purpose of this thread, I think you should revise one of the opening statements. The idea that, "one is going to capture almost 95% of the features of most DOSs [sic] in about 30 min combined exposure" is simply absurd.