Qkslvr
Pooh-Bah
Reged: 06/23/06
Posts: 1316
Loc: NE Ohio, US
|
|
Quote:
Q,
.....
Bottom line is if you can shoot 5 minute subs it's easier and based on what I have seen, results in a smoother, deeper image.
JMO's
Strgazr27, I agree. I think if you can do 5 min sub, you're probably a lot better off than doing 15 sec subs. But if you can only do 15 sec subs, you still might be able to get good results.
--------------------
Mike
Onyx 80ED/N8/CG-5/40D
|
russ_watters
Pooh-Bah
Reged: 11/24/04
Posts: 1340
Loc: Trappe, PA
|
|
Two things for the two separate discussions going on...
1. My personal experience with my EQ-6 is that I can't reliably do unguided exposures above about 30 seconds at my focal length of about 1200mm. I've done as high as 1 min, having to discard more than half.
2. Regarding stacking vs longer exposures. For a lot of people and a lot of targets, this is an irrelevant discussion. I'm 30 miles from Philadelphia and I can't do 5 minute luminance subs in any part of the sky. So it doesn't matter if a single 10 minute is equivalent of 3 5 minute subs.
Now regarding averaging vs summing, the issue is similar. There are few targets out there that summing will be useful on, even in dark skies. The dynamic range is too large and if you sum the exposures, you'll blow-out the brigter areas.
So the simplest answer to this issue is use the longest exposures you can reasonably take and stack them as an average. All those complexities are interesting, but not terribly relevant to most people.
--------------------
Equipment: Orion Atlas 11, ED80, DSI-C, DSI III Pro, Dell Inspiron Laptop.
www.russsscope.net
|
Chuck Anstey
scholastic sledgehammer
Reged: 01/17/05
Posts: 960
|
|
Quote:
I've been reading this thread with intense interest. May I ask I side question?
How does one go about using factors like LP to estimate a good exposure lenght for your subs? Does everybody just have to experiment until they figure it out, or is there a more quantitative way to determine how long your exposures should be?
My paper has all the formulas you need and I put them in an Excel spreadsheet but here is a quick and dirty way to find the values.
First I shoot what I want for a sub-frame exposure, recently 10 minutes for M51 at a dark site, to verify the object can take the exposure without blooming or saturating from LP. I then measured the background signal minus bias. For my 10m frame it was 800 ADU - 100 ADU bias. That leaves 700 ADU LP signal. Noise equals sqrt(LP signal) / sqrt(camera gain) = sqrt(700)/sqrt(1.3) = 23.2 ADU RMS. Now I want to drive the final stacked noise to around 6 ADU. That takes (noise per frame / final noise)^2 = (23.2 / 6)^2 = ~16 frames. I had plenty of time to shoot 16 10-minute frames so that is what I did. If I didn't then I would have used a shorter exposure and fewer frames.
If you do not know the gain of your camera then you can simply use the sqrt(LP signal). It should be guaranteed to result in at least the number of frames you need to drive the final noise below whatever value you set (start with 6). If you find the number of frames too small then increase your exposure time if you can.
The number of frames from the formula is around the maximum number of frames where adding more frames no longer improves the final picture. If you don't believe it then shoot at least twice as many and compare the result of stacking just the number the formula recommends with stacking all the frames. If you see a visible difference then lower the final noise value so it recommends the larger number of frames and do the experiment again. Try to find the final noise value that no longer seems to improve the stacked picture. It isn't going to be 1.
Edit - Addendum
The above assumes the exposure is long enough that LP noise is the dominant noise and not camera read noise. If that is not true then you will have to stack more frames than the formula gives and the new formula is at least #frames = (read noise / final noise)^2.
Edited by Chuck Anstey (03/23/07 03:34 PM)
|
Chuck Anstey
scholastic sledgehammer
Reged: 01/17/05
Posts: 960
|
|
Qkslvr,
You are on the right track but you made two minor errors that make a big difference in the outcome.
Quote:
If 15 sec doesn't capture but 1 or 2 photons, and your read error is equal or greater than that, you're right.
Read error doesn't matter here, only the target noise issue. Read noise is constant per frame so you can reduce it by sqrt(# frames). That is why it is recommended people stack 20+ dark frames to create a master dark. The problem is that 1 or 2 photons have a rather large noise explained by Binomial distribution for small sample sizes.
Quote:
If each sub captures 50+/-5 photons, and has +/-2 adu's read error, 10 sub's will have 500+/-50 in signal and +/-20 adu read error for a total of 500+/-70 vs a single sub with 500+/-52 A difference of 18 adu of error.
You are correct that noise goes up when summing but not linearly as you show. Noise goes up by the square root of the sum of the squares. So for your case New Noise = sqrt ((5^2 + 2^2) + (5^2 + 2^2) + ... 10 times). As long as LP noise is stronger than read noise, as is the case in your example, stacking or shooting longer are effectively identical in SNR for all but the faintest signals, which the longer exposure has a big advantage.
For all but the faintest signals, SNR is identical (within 10%) whether stacking or shooting longer for the same total time as long as camera noise is < ~24% of the total noise. The 5% rule is overkill.
Russ,
Quote:
2. Regarding stacking vs longer exposures. For a lot of people and a lot of targets, this is an irrelevant discussion. I'm 30 miles from Philadelphia and I can't do 5 minute luminance subs in any part of the sky. So it doesn't matter if a single 10 minute is equivalent of 3 5 minute subs.
I disagree here because if nothing else you should come away with the understanding that there is a limit to how many frames can be stacked that show an improvement and that in general you should be trying to get that many frames. If you are limited to 30 seconds subs and that you find 20 subs is all you need to hit the wall of diminishing returns then you know not to waste time trying for 120 frames. Also you know that you should always try for 20 frames for each object that can take a 30 second exposure.
I live 30 miles from downtown Atlanta and I can shoot between 6 and 10 minutes but I have an SBIG camera and a CGE and AP1200 mount so guiding and dynamic range are not an issue. Six minutes is about my true limit when shooting toward Atlanta because it takes about 60 frames to reduce the large amount of noise. At 10 minutes it is closer to 150. For that I go to a dark site and can do it in just 16 frames.
|
Jared
Post Laureate
Reged: 10/11/05
Posts: 4554
Loc: Piedmont, California, U.S.
|
|
Quote:
So John, using your logic.. from a dark site:
240 exposures at 15 seconds per exposure has the same overall outcome as 60 exposures of 1 minute each?
CF
Even though I'm not John, I'll take the liberty of replying...
It wouldn't quite have the same results for two reasons:
1) Each subexposure has readout noise as well as ordinary dark current, etc., so the fewer the subexposures the better. That being said, at some fairly moderate exposure duration other forms of noise are MUCH larger than readout noise, so while 240 15s exposures may not be a great idea, the difference between five minute sub-exposures and ten minute sub-exposures might be pretty small if you get the same total exposure time
2) In a perfect world you would be able to "stretch" the averaged subexposures to achieve any intensity you want, but in fact the bit depth puts a limit on how much stretching you can do without posterization. You can always add the subexposures instead of averaging or median combining them (you get the same s/n ratio benefits), but then you run the risk of blowing out highlights in areas of the picture. Usually, 15s subexposures would be too short to let you effectively use averaging/median combine methods on anything but the brightest targets
If cameras had larger bit depths, and if read noise were negligible, then the answer really would be "yes"!
--------------------
- Jared Willson
- TV 60iS refractor
- 80mm SV/LOMO refractor
- A-P 130 EDFS refractor
- Planewave 12.5" CDK
- A-P 900 mount
- A-P Mach1 GTO mount
- Takahashi Teegul SP travel mount
- SBIG STL-11000 C1 camera with AO-L
- Some heavily light polluted skies
|
Jared
Post Laureate
Reged: 10/11/05
Posts: 4554
Loc: Piedmont, California, U.S.
|
|
Quote:
So the simplest answer to this issue is use the longest exposures you can reasonably take and stack them as an average. All those complexities are interesting, but not terribly relevant to most people.
Absolutely correct! The theoretical examples of perfect subexposure length probably don't apply unless you have both very good sky conditions and an extremely accurate tracking system. The original poster was asking about unguided imaging with an EQ-6. All by itself, that will become the limiting factor.
Choose the longest subexposures that consistently (60% of the time?) yield nice round star images. Depending on the periodic error of your mount, how carefully you have balanced it, how carefully you drift aligned, and the focal length of your scope that could be as little as 10 seconds or as much as five minutes. That's probably the only factor that really matters for unguided imaging.
--------------------
- Jared Willson
- TV 60iS refractor
- 80mm SV/LOMO refractor
- A-P 130 EDFS refractor
- Planewave 12.5" CDK
- A-P 900 mount
- A-P Mach1 GTO mount
- Takahashi Teegul SP travel mount
- SBIG STL-11000 C1 camera with AO-L
- Some heavily light polluted skies
|
ribuck
sage
Reged: 11/01/05
Posts: 329
Loc: Newcastle, England
|
|
Thanks everyone, this has been a really interest thread and never expect so much interested in it. Whilst i'm a noob and dont fully understand the theory behind it all, i'm basically taking away form this that there is no set hard and fast rule and that it all depends on your setup.
As it stands, i will be buying the EQ6 Pro and will probably purchase either the new WO 110mm doublet with a F5.9 or the Skywatcher Pro 100mm ED2, but that has a F9 triplet, and just think that the long focal length will make it a lot harder to image with seeing that i can only afford a smaller 1/3 size ccd.
I guess i'll just aim for 2min subs and see what i can get with that until i can afford a guide scope and guide camera.
Thanks all, it's been a really interesting thread and has opened my eyes.
|
Chuck Anstey
scholastic sledgehammer
Reged: 01/17/05
Posts: 960
|
|
Quote:
The theoretical examples of perfect subexposure length probably don't apply unless you have both very good sky conditions and an extremely accurate tracking system. The original poster was asking about unguided imaging with an EQ-6. All by itself, that will become the limiting factor.
While unguided will be the limiting factor in exposure length, I still think knowledge of exposure limitations, how stacking works, what it can accomplish and what can be expected from it are very important. Just the knowledge that there is a point of diminishing returns from additional frames and how to find it are important and that deeper requires longer. Ideally one wants to optimize their time under the stars so it is important to know that shooting 60 frames won't really be any better than 20 frames or that shooting 25 frames leaves details on the table and that 50 frames is pretty much the limit. No sense spending forever trying to get those faint spiral arms if the exposure length won't accomplish it.
"Optimal" is a gray area that has some range but you still ought to know where it is and what it means even if you consciously shoot outside of it. I think you are selling yourself and your equipment short if you totally ignore it and shoot by "feel".
|
ribuck
sage
Reged: 11/01/05
Posts: 329
Loc: Newcastle, England
|
|
Yes Chuck, I must admit i totally agree with you. I found myself thinking that I need to read up on an awful lot more before i even consider buying my new toys.
So i'm going to have a good read through the forums to understand how to calculate exposure times properly.
This has always been a question of mine when i see peoples stunning images where they have taken different lengths of luminance exposures to the RGB, i never understood how they came up with these numbers.
Even after a year this is still a steep learning curve and i' not going to plunge in blindly like i did last time, which was a very costly mistake.
|
jgraham
Postmaster
Reged: 12/02/04
Posts: 11575
Loc: Dayton, Ohio
|
|
Clownfish: I've been scratching my head over that one (240x15sec = 60x60sec?) and my quick answer is there's no way I could consider them the same, if they are you'd have to accept that any combination of a*b=constant are the same, but mathematically I can imagine how they might be. The right answer is to dive into the math (and Chuck may have already done the hard work for us). I know that in the extremes they're not the same because of the way different source of noise work, but I don't have feel for where the gray area is between what you'd encounter from a practical stand point and when you start running into the limits of the models. This is complicated by the issue of two images that are mathematically different (different S/N, different resolution ine luminance) but visually identical (they 'look' the same). If I recall right we can distinguish between 60 and 80 levels of gray, anything with more levels that that 'looks' continuous. So if two images are scaled so black is black, white is white, and the luminance and S/N is greater than about 80/1 (after processing), the two images will 'look' identical.
As someone pointed out, from a practical standpoint for some of us this is pretty moot. I expose my indiviudal frames as long as my tracking and sky glow permits (generally 30sec) or until an object is short of saturation, then average as many frames as I have patience for (about 25-100 depending on the circumstances). This method isn't very scientific, but it gives all of the illusion of working. 
-John
--------------------
-John
The best advice on imaging I've ever been given... don't forget to look! 
|
Chuck Anstey
scholastic sledgehammer
Reged: 01/17/05
Posts: 960
|
|
John,
The gray area is the signal strength that has equal SNR. 240x15sec = 60x60sec for strong signals in the FOV. Think about planetary imaging. There the signals are extremely strong and the exposure is extremly short. The SNR is as high as it can be for each individual frame and stacking is the only way to increase it. But at the other end, no matter how many shots are stacked, you cannot see the moons of the planet. One longer shot that blows out the planet can easily show multiple moons with no stacking required. I think this is the most prevalent misunderstanding when talking about SNR. It matters what signal strength you are talking about and there is no one SNR for the whole FOV. The length of the sub-exposure moves the slider where signals < slider point are below the sqrt(# frames) stacking rule. It also controls the upper slider where signals > upper slider = blown out or partially clipped.
John, I do have one question. What do you mean by "and sky glow permits"? How are you measuring that sky glow is impacting the end result?
|
ClownFish
Post Laureate
Reged: 04/26/05
Posts: 6254
Loc: Baghdad, Iraq
|
|
I can easily guide my LXD75 for 30 to 45 minutes now - so I hope all I will need when I go digital some day is a nice dark sky - that way I can reduce the number of subs and keep it simple. Right now I use only one exposure, and hardly process at all. Then again, digital is so much more critical to guiding errors than film, so I may not be able to get the same quality I get now with that mount - I do not know.
This digital stuff sounds a bit depressing and expensive! I think I'll ask for a posting to Namibia after South Africa - that should be dark enough to keep the LP noise lever down, which would help!!
CF
--------------------
Learn all about Polar Alignment and Manual Guiding on my website at www.PetesAstrophotography.com! Or visit my Foreign Service Blog!
|
jgraham
Postmaster
Reged: 12/02/04
Posts: 11575
Loc: Dayton, Ohio
|
|
Chuck; as I lengthen the exposure the black point (essentially the background skyglow) slides to higher and higher values, consuming dynamic range. Depending on how much skyglow I'm dealing with, going longer than 30sec just doesn't buy me much, particularly if it starts to shove the signal (the stuff above the skyglow) into the nonlinear range of the CCD. In my mental equation I'm trying to balance signal (exposure), noise (number of subs), and tracking (exposure), and the solution converges at about 30 seconds (on my LXD75 SN6). On the rare (and I mean rare) nights when the air is dry, calm, and dark and the fates align to give me good tracking I'll move up to 45 sec subs. Also, when I do wide field work (using various short focal length optics) I often move up to 45-60 sec subs since the tracking issue is much reduced or goes away completely.
Clownfish; for a given field of view the guiding issues are the same. One neat thing about small CCDs is to get a wide field you need a small telescope. Using a 35mm f/4 achromat (the objective from an old pair of binoculars) I get the same field of view as a full 35mm frame as an 8" f/5 (or anything else with a 40" focal length).
I shot film for many, many years (everything from 5x7 and 4x5 sheet film, 620 and 120 large format roll film, and 35mm with my trusty OM-1) and there are trade-offs between the different formats depending on what you're trying to do. From what I've seen of your work (beautiful sharp wide fields, dark skies) you'll probably be looking at large CCD arrays, and that can get pricey. If I had the time to get out under dark skies I'd probably go that way myself (I'm kinda eyeing a DSLR one of these days) but for now I need to stay close to home and the equipemnt I have is a good match to what I can do from my back yard.
-John
--------------------
-John
The best advice on imaging I've ever been given... don't forget to look!
|
Andrew Welsh
Carpal Tunnel
Reged: 03/28/06
Posts: 2571
Loc: Rochester, NY
|
|
I can say from my experience shooting M100 in my LP'd skies, that I can get 7 min of exposure while using an IDAS-LPS filter. Any longer and I will wash out the image. However I can only squeeze so much detail out of the image. And I did 100 subs!
Given these facts, how can I get any more detail out of the image? Wait for perfect transparency? Go to a dark site? Get another 100 subs? If I were at a dark sky site I can expose for way more than 7 min before the sky fog starts taking over the image... so does the adage that for each mag drop in sky darkness you have to increase the integration time 2.5x to get the same SNR for the dark site? In other words, 1 hour at a mag 7 site is the same as 2.5 hours at a mag 6 and 5 hours at a mag 5 site?
--------------------
LX200 8" classic, f/10, Meade eq. wedge, .63x FF/FR
Canon 40D (LifePixel clear glass mod) and 5DMkII, unmodified
Canon EF 200/2L IS, 400/5.6L, 100/2.8 Macro, 50/1.4, 85/1.8, 35L, 24L, 17-40L and Peleng 8mm fisheye
Orion Apex 102mm (4") Mak-Cass
Pimped out with accessories and bling
My DSLR Astrophotography Webpage and photo bucket with full equipment list
|
Charlie Hein
Postmaster
Reged: 11/02/03
Posts: 11211
Loc: 26.06.08N, +80.23.08W
|
|
Quote:
Hi all,
I'm looking to buy a skyscan EQ6 Pro which I believe will track very well for about 2 mins, before needing to be autoguided.
I was wondering if it's possible to get some decent images from 100-120 second exposure lenghts, as i cant afford to buy all the extra kit required for guiding right now.
I use an Orion 80ED on my Atlas with either a Canon 300D or an SBIG ST2000 XCM, and I can consistently get three or four minute unguided exposures out of this rig. Have a look at my gallery here for several examples - YMMV, but I think that three minutes should be achievable for you if your rig (scope, camera) is similar to mine.
Charlie
--------------------
Weston CSC:
|
Qkslvr
Pooh-Bah
Reged: 06/23/06
Posts: 1316
Loc: NE Ohio, US
|
|
Quote:
...
Given these facts, how can I get any more detail out of the image?
Andrew, This is where you can add some subs together to get more photons to work with. Which you don't get from Stretching, which would be like adding the same sub to itself.
And since the addition is done in software, you just need to select a format that has large numbers for pixel values, I know some of the formats support 32 and 64 bit floating point numbers. You can add a lot of sub without clipping with a 32 bit floating point number.
--------------------
Mike
Onyx 80ED/N8/CG-5/40D
|
Qkslvr
Pooh-Bah
Reged: 06/23/06
Posts: 1316
Loc: NE Ohio, US
|
|
Quote:
Qkslvr,
You are on the right track but you made two minor errors that make a big difference in the outcome.
Quote:
If 15 sec doesn't capture but 1 or 2 photons, and your read error is equal or greater than that, you're right.
Read error doesn't matter here, only the target noise issue. Read noise is constant per frame so you can reduce it by sqrt(# frames). That is why it is recommended people stack 20+ dark frames to create a master dark. The problem is that 1 or 2 photons have a rather large noise explained by Binomial distribution for small sample sizes.
Quote:
If each sub captures 50+/-5 photons, and has +/-2 adu's read error, 10 sub's will have 500+/-50 in signal and +/-20 adu read error for a total of 500+/-70 vs a single sub with 500+/-52 A difference of 18 adu of error.
You are correct that noise goes up when summing but not linearly as you show. Noise goes up by the square root of the sum of the squares. So for your case New Noise = sqrt ((5^2 + 2^2) + (5^2 + 2^2) + ... 10 times). As long as LP noise is stronger than read noise, as is the case in your example, stacking or shooting longer are effectively identical in SNR for all but the faintest signals, which the longer exposure has a big advantage.
For all but the faintest signals, SNR is identical (within 10%) whether stacking or shooting longer for the same total time as long as camera noise is < ~24% of the total noise. The 5% rule is overkill.
Thanks for fixing that Chuck!
I'm stuck at 30 Sec exposures until I get a cable to switch my 300D, Plus I'm not sure how much longer I can go unguided. Worse still is I can only go to about 1.5 minutes to half histogram at ISO1600 with my skys.
--------------------
Mike
Onyx 80ED/N8/CG-5/40D
|
ClownFish
Post Laureate
Reged: 04/26/05
Posts: 6254
Loc: Baghdad, Iraq
|
|
Quote:
Clownfish; for a given field of view the guiding issues are the same
John, I know what you are saying here, but what I want to describe (and other may or may not find interesting) is how the fast response of digital imagers require a more acurate mount when shooting long exposures than with film. Here's an example:
When I am exposing with film, I can make huge guiding corrections that would have easily ruined a digital sub. I can bump the scope so far hat the guidestar is knocked of the field of view, and then get it back within a few seconds and everything si still fine. If I get a large PE error every 30 seconds, or gusts of wind a couple of times a minute, it will have little to no impact on the final image. With a digital imager those subs would have to be tossed. Guiding a few subs is not hard, but trying to grab 50 to 100 while hand guiding would be a back breaker. Autoguiding doesn't apeal to me - as a photographer.
So in the end, my single 30 minute film exposure is all that I need. As long as my focus is dead on, I can pretty much be comfortable that the image was a success. With digital, you either have to take a lot more subs to catch enough good ones, or buy a better mount.
About me getting a large format CCD array.. that will probably never happen. The most I can hope for is a modified DSLR. The price of the large fomat CCD cameras (SBIG) are just too prohibitive for me. I would be very happy with a used medium format film camera and play with that baby for a few summers!
CF
--------------------
Learn all about Polar Alignment and Manual Guiding on my website at www.PetesAstrophotography.com! Or visit my Foreign Service Blog!
|
russ_watters
Pooh-Bah
Reged: 11/24/04
Posts: 1340
Loc: Trappe, PA
|
|
Quote:
Russ,
I disagree here because if nothing else you should come away with the understanding that there is a limit to how many frames can be stacked that show an improvement and that in general you should be trying to get that many frames. If you are limited to 30 seconds subs and that you find 20 subs is all you need to hit the wall of diminishing returns then you know not to waste time trying for 120 frames. Also you know that you should always try for 20 frames for each object that can take a 30 second exposure.
I didn't say anything about how many frames to stack, just how long the subs should be. But as you point out in the next post, there isn't really a "wall of diminishing returns", but a curve. So that makes the number of frames to be a mixture of other factors and still not a calculation. It is a matter of trial and error, personal taste, and time. The number of subs in my recent M-63 image was based on how many I could take in one night!
--------------------
Equipment: Orion Atlas 11, ED80, DSI-C, DSI III Pro, Dell Inspiron Laptop.
www.russsscope.net
|
russ_watters
Pooh-Bah
Reged: 11/24/04
Posts: 1340
Loc: Trappe, PA
|
|
Quote:
This has always been a question of mine when i see peoples stunning images where they have taken different lengths of luminance exposures to the RGB, i never understood how they came up with these numbers.
I'm less than a year into it and only have one galaxy photo I would consider "great" and another "good", but the criteria for this M63 image really was just a combination of sky limit (though it turned out, it was unusually transparent, so I probably could have gone a little longer if I trusted my guiding) and hours available in the night. With roughly 75% of my exposures being keep-able, I ended up with about 60x2min lumnance 25x2min red, and 20x3min, green and blue - roughly 5 hours of exposure over the course of about 6 hours of night (before going to bed later than I should have...). I choose 2 min for luminance partly for detail over brightness, as I'm still having some tracking issues (I think my scope is flexing).
http://www.russsscope.net/images/m63-3-18-07.jpg
--------------------
Equipment: Orion Atlas 11, ED80, DSI-C, DSI III Pro, Dell Inspiron Laptop.
www.russsscope.net
|
Previous
Index
Next
Threaded
[Re: 
Edit
Reply
Quote
