Astro Picts.

[tkerr]

Here are a couple of my astrophotos.


The Pleiades (Messier 45) In Taurus.
Camera: Canon EOS 50D.
Lens: Celestron C80ED f/7.5 Refractor telescope (600mm Focal Length).
ISO 800
Stack of 45 x 180 second exposures.


http://api.photoshop.com/v1.0/accounts/e308b554104a49c9973e1e35131b193c/assets/34e5e4026af84fdb8aaf31950a1f205f/renditions/1024.jpg?md=1281563418000





Messier 31 The Andromeda Galaxy
Camera: Canon EOS 50D
Lens: Canon EF 100-400mm L IS USM, at 300mm f/5.6
ISO 800
Stack of 25 x 180 second exposures.
http://api.photoshop.com/v1.0/accounts/e308b554104a49c9973e1e35131b193c/assets/b4ad97d662754c29aa6d72ffcbe1481a/renditions/1024.jpg?md=1268685577000



Telescope mount used was the Celestron CG5-GT Computerized mount.
Autoguiding with Stark Labs PHD Autoguiding software.
Image stacking with Deep Sky Stacker
PP with Photoshop CS5

[PaulM]

Hi Tim,


In laymans ( read I

[tkerr]

Generally, it's usually a matter of the Aperture of the objective lens.
A majority of astroimages are done with a telescope, but not all. I guess it depends on what you are trying to accomplish. A camera lens is usually smaller and will give you a wider field of view. Using the 100-400mm gives me more flexibility than using a fixed focal length telescope. But the telescope has a larger aperture and longer focal length which will have greater light grasp and resolving power.(Greater Detail), and increased image scale.


BTW, you don't need a telescope mount to take pictures of the night sky. You will if you want to do something like these two pictures, but there are other things you can do also.


[View:http://www.youtube.com/watch?v=6JvzlMhZavc&hd=1]

[Baker]

What program/software do you use to make your time lapse?


Here is a shot I took a couple of weeks ago. I believe it is 70 thirty second exposures stacked together.





http://farm2.static.flickr.com/1405/5104623653_fe1f92da6f_b.jpg (http://www.flickr.com/photos/chrishinytzke/5104623653/)


IMG_1634-IMG_1697 (http://www.flickr.com/photos/chrishinytzke/5104623653/) by Chris-Baker (http://www.flickr.com/people/chrishinytzke/), on Flickr

[Jon Ruyle]

Nice shots of the andromeda galaxy and pleadies. I like the color in m31. I haven't been able to get nice color like that in my pics.



Generally, it's usually a matter of the Aperture of the objective lens.


Exactly. Many people get hung up on focal length, but with a large, high pixel density CCD, there is a lot of latitude for adjusting effective focal length. (Though admittedly, for a very wide angle picture, one needs a wide angle lens).


An interesting question, though, is which is preferable when there is overlap? Eg, would an 800mm f/5.6 L be preferable to say, a 140mm AP refractor? I don't know the answer . Most of us don't have both, so we make do with what we have :)


I took a couple of pics last night with my little scope. I'll post them when I get them processed.

[tkerr]

What program/software do you use to make your time lapse?
<div style="clear: both;"]</div>





Currently I use Pinnacle Studio HD Ultimate 14. when I first started creating time lapse I would just use Windows Movie Maker. But I didn't like the rendering and limitations so I had to look for something else.


How did you stack you star trails image, Photoshop Or Startrails?

[Jon Ruyle]

Here is one of the pictures I took last night from my yard in Wonder Valley, CA:


/resized-image.ashx/__size/550x0/__key/CommunityServer-Discussions-Components-Files/15/6523.m81-wonder-valley.jpg


5DII (long exposure NR on) + 9.25" Edge HD (2350mm @ f/10) + Astrophysics Mach 1 GTO mount


10 exposures, 4 min each, stacked with Deep Sky Stacker.


I had some trouble with tracking: though autoguiding, I experienced a slow steady drift. Don't know if it was differential flexure, mirror flop, or what, but 4 min was about as much as I could do. I never had such problems with my refractor.


Seeing was lousy, but it was a nice dark night. I looked at galaxies through the big telescope while the small one was taking pictures. My idea of a nice evening!

[tkerr]

Here is one of the pictures I took last night from my yard in Wonder Valley, CA:


/resized-image.ashx/__size/550x0/__key/CommunityServer-Discussions-Components-Files/15/6523.m81-wonder-valley.jpg


5DII (long exposure NR on) + 9.25" Edge HD (2350mm @ f/10) + Astrophysics Mach 1 GTO mount


10 exposures, 4 min each, stacked with Deep Sky Stacker.


I had some trouble with tracking: though autoguiding, I experienced a slow steady drift. Don't know if it was differential flexure, mirror flop, or what, but 4 min was about as much as I could do. I never had such problems with my refractor.


Seeing was lousy, but it was a nice dark night. I looked at galaxies through the big telescope while the small one was taking pictures. My idea of a nice evening!









<div style="clear: both;"]</div>








Very nice, A lot of detail in the galaxy. However, I wonder why you are using the Long Exposure Noise Reduction rather than shooting dark frames. Using the cameras long exposure noise reduction often removes data that you might otherwise want, i.e. small faint stars and some of the fainter detail in the outer reaches of the galaxy and surrounding areas.
Additionally, on my monitor it looks like the dark points are clipped a little too much. Space is dark, but it's not pitch black.
I can see the tracking error, but all in all even considering what else I mentioned it's a very nice image, the resolution is great.


How do you like that New Edge HD?


Didn't you shoot any darks, Flats or Bias?
Even if you don't shoot darks or flats, I feel bias frames are a minimum necessity for image calibration and stacking.

[doggiedoc]

Really Awesome Jon!


Paul

[Jon Ruyle]

Thanks, Paul.


Tim-when yous say the dark points are clipped to much, do you mean I subtracted too much or not enough? On my monitor, space looks totally black.


As for darks and flats, I can't either to work right with DSS- maybe you can help.


When I dark frames in DSS (instead of in camera NR), I get little streaks all over the picture. I think it has to do with the fact that I'm getting a little drift (separate issue), but I don't know. I haven't heard of other people having the problem. Of course for the darks I use the same ISO and exposure time and make sure the temperature is close.


Whenever I've tried flats the thing always overcompensates (gives me a bright edges and a dim center). Maybe I'm doing something wrong.


My refractor has almost no vignetting, so I never needed flats. The 9.25HD has comparatively a ton of vignetting, so flats are more important now. If I am to be serious about this, I suppose I need to figure all these problems out- either by getting it to work with DSS or using a different program.


Other than that, the jury is still out on the 9.25. I only get the drift I was talking about with the 9.25, so if it is caused by mirror flop, that means thge clutches aren't working very well, which would be pretty annoying. But the problem seems to uniform to be mirror flop. If it is differential flexure, it is still a strike against the 9.25 because it is so hard to set up an off axis guider with it. I might go with one of the inexpensive RC's if I had it to do over again, but no doubt they have their own problems.


The 9.25 *does* have pinpoint stars to the edge of the field, I'll give it that.

[tkerr]

Thanks, Paul.


Tim-when yous say the dark points are clipped to much, do you mean I subtracted too much or not enough? On my monitor, space looks totally black.


I mean you adjusted it so that the dark points of you picture are too dark.


Here is a pretty good overview of using DSS and post processing.
</div>


[/quote]


Flats can be a bit tricky to shoot and get the right exposure. How are you setting the camera up to shoot them?


Flats should be brighter in the center and darken as you get closer to the outer edges. The transition will be more pronounced if you have serious vignetting. The trick is to get the exposure so the center where it is brightest is not blown out(over-exposed). You will probably have to take some test shots keeping a close eye on the histogram. You want to expose to the right, but not too far to the point that the histogram is jammed against the right side.


Maybe this will help give you a better idea. &gt;&gt;&gt; Flat Field Calibration (http://astrochat.co.uk/forum/viewtopic.php?t=13241]Astro Image Tweaking.






As for darks and flats, I can't either to work right with DSS- maybe you can help.


When I dark frames in DSS (instead of in camera NR), I get little streaks all over the picture. I think it has to do with the fact that I'm getting a little drift (separate issue), but I don't know. I haven't heard of other people having the problem. Of course for the darks I use the same ISO and exposure time and make sure the temperature is close.





Make sure you are shooting your dark frames the same exposure as you shoot your light frames.
Dark frames don't correct for all the noise! Some of the random noise as well as hot or cold pixels, but there is also what some might refer to as read noise or fixed pattern noise (aka Bias or Offset) that is generated by the camera internal processor reading the data collected from the sensor. This is why you need to shoot bias frames. The ugly streaks you see all over are probably the bias error.


[quote]


Whenever I've tried flats the thing always overcompensates (gives me a bright edges and a dim center). Maybe I'm doing something wrong.


My refractor has almost no vignetting, so I never needed flats.
<div style="clear: both;)


Additionally, Flats are not just used to correct for vignetting, but for other optical anomalies such as dust spots on your sensor, and artifacts caused by pixel to pixel variations.

[Jon Ruyle]

I just put a white t-shirt over the lens and shoot in av mode. I have a set for each iso setting. I try to expose them fairly bright, but make sure there are no overexposed points.


I think I understand the basic idea, but somehow it just doesn

[tkerr]

I thought you were saying space in the picture was not black enough. I agree that it would look better if I subtracted less background light, but until I fix the vignetting problem I have to get rid of all the skyglow (or I get a little glow in the middle and black on the edges- it looks terrible)



<div style="clear: both;"]</div>





That's just another reason it is so important to use flats.


You also selectively adjust isolated areas of your image using select by color range with layers and layer mask. That is if you are using Photoshop CS*. That way you can adjust around the areas you don't want your adjustments to affect and get it to blend real well..

[tkerr]

Here is a couple more from my collection.


Both of these were shot with my EOS Rebel XT/350D
ISO 800 through my Celestron C80 ED refractor


Monkey Head Nebula
30 x 120 seconds
http://api.photoshop.com/v1.0/accounts/e308b554104a49c9973e1e35131b193c/assets/be805116d3084404aa4966bab5e75be7/renditions/1024.jpg?md=1268685031000


The Rosette Nebula Complex
46 x 180 Seconds
http://api.photoshop.com/v1.0/accounts/e308b554104a49c9973e1e35131b193c/assets/73ed909e56de47d1a671c4a290ab8d52/renditions/1024.jpg?md=1268685021000

[Jon Ruyle]

Yeah, I agree that the vignetting is a problem that must be solved.


Both of those are wonderful, Tim. Makes me wonder why bother with autouiding :)


I

[tkerr]

Yeah, I agree that the vignetting is a problem that must be solved.


Both of those are wonderful, Tim. Makes me wonder why bother with autouiding :)


I'm guessing you used all types of ancillary shots for these.


The 80ED has about the same aperture as your 100-400. Any thoughts on how they compare?


(I realize that the 80 ED is a bit longer, but not too much, and I would think the rebel has plenty of pixels to crop a bit)
<div style="clear: both;"]</div>





With my mount and the amount of weight I have on it Autoguiding for long exposures is critical. I am using a Celestron C6R-GT (CG5).


Although the 80ED is only a few mm larger than the 100-400 nothing compares to a prime lens, which is, in effect, what I am doing when attaching the camera to the telescope. And the focal length does make quite a bit of difference in the Image scale and field of view.

A lot depends on what kind of shot I am after, and the object or object within it, that will determine which lens or which telescope I use.

[Jon Ruyle]

You aren

[Jon Ruyle]

Okay, it looks like DSS does not allow the use of flats without dark frames or vice versa. I have no idea why that is.


That might explain why when I was using the refractor and used dark frames (but no flats because vignetting was not a problem), it did not work. By the time I got the 9.25" I had given up on dark frames (the in-camera solution works okay) and tried to use flats with no darks, and again it didn

[tkerr]

You aren't autoguiding those wide angle, 3 minute exposures are you?


I just meant- why bother with long exposures when stacking short exposures gives such good results.
<div style="clear: both;"]</div>





Yes I do use the autoguiding even for wide angle exposure that long. The CG5 GT mount isn't that accurate, especially when you overload it like I do. And no mater what camera and lens I use it does make a difference in the stars


Why bother with long exposures? Increased image data / SNR.
Stacking multiple exposure gives you a better SNR, but it doesn't necessarily increase your exposure / saturation. Stacking multiple longer exposure will result in a better image than stacking multiple shorter exposures or a single very long exposure.


An assumption that many people make is that if you stack 40 x 30 second exposures that it will equal a 20 minute exposure. When in fact what you will have is a cleaner looking 30 second exposure instead. Don't get me wrong, more exposure will gain more detail than a single longer exposure, but only because you are improving the signal to noise ratio.


This thread can explain it better than I can here.
SNR and Total Signal ("http://www.cloudynights.com/ubbthreads/showflat.php/Cat/1,2,3,4,5,8/Number/3904172/Main/3902239)

[Jon Ruyle]

You sure get nice results with your overloaded mount :)


I use my mount nowhere near capacity- if I polar align carefully, I can do a 15 minute unguided exposure with an 800mm lens and the stars will look round when viewed 1-1.



This thread can explain it better than I can here.


I'm not sure what the guy in that thread means by "doing the math". There is no math. The only differences between the sum of 10 60-second exposures and a single 10 minute exposure are read noise and clipped highlights. As long as you take a long enough exposure so that read noise is small and a short enough exposure that you don't clip highlights, there is no real advantage either way.



An assumption that many people make is that if you stack 40 x 30 second exposures that it will equal a 20 minute exposure.


I'm one of those people, apparently. I believe that the only differences are the two I mentioned above. Then again maybe read noise would be a factor in a bunch of 30 second exposures.


I think there is a lot of confusion about this- I've heard people say you need a single long exposure to get dim details, or who knows what else. On the other hand, maybe I am the one who is confused :)

[tkerr]

There is a huge difference in a Celestron CG5 GT Mount and an AP Mach 1 GTO. Even if you were to push the load limits it would still track better than mine.


In that thread they briefly touched on the mathematical formula, I don't worry about the math, instead I prefer practical application and seeing with my own eyes.


Signal remains a constant from one exposure to the next, but the noise in each exposure is random, and when you stack multiple images the noise cancels itself out.


Increasing the SNR you're also increasing the Dynamic range which might make it appear, or make you think that you have increased exposure length, when instead you're only revealing those fainter details because there is less noise blocking them. Does that make any sense?


Anyways, I was one of those who thought multiple exposure added up to a greater exposure length, E.g 10 x 60 seconds would equal to 10 minutes exposure. I was wrong, through trial and error I have learned differently.


The best thing I can suggest doing is to experiment; on the same object if you can. You can even do it in your home with your camera and camera lens taking pictures of a gray card or something. For Example take 40 x 30 second exposures to stack and process, and also take 20 x 60 second exposures to stack and process, and then compare the differences.


I think were a lot of the confusion comes from is from SNR and people equate the Signal to an accumulated exposure length.


If you really want to get those faint details you will need longer exposures. I try to push my exposure to the maximum limited by the skyglow, not the noise level evident in each exposure due to the high ISO. That will clean itself up with enough exposures/light frames, and darks, flats and bias frames.

[Jon Ruyle]

There is a huge difference in a Celestron CG5 GT Mount and an AP Mach 1 GTO. Even if you were to push the load limits it would still track better than mine.


Sure- well, there had better be a difference, anyway. [:)]



Increasing the SNR you're also increasing the Dynamic range which might make it appear, or make you think that you have increased exposure length, when instead you're only revealing those fainter details because there is less noise blocking them. Does that make any sense?


Sure it does. In some sense, SNR is all that matters. That and quantization noise.



Anyways, I was one of those who thought multiple exposure added up to a greater exposure length, E.g 10 x 60 seconds would equal to 10 minutes exposure. I was wrong, through trial and error I have learned differently.


If there is a difference, it is due to read noise, and maybe thermal noise. With exposures that short, I believe it. But if you compare a single 100 minute exposure to 10 10-minute exposures, I doubt you would see a difference. (Assuming you overexpose in the 100 minute exposure).



If you really want to get those faint details you will need longer exposures.


Even read noise can be overcome with large samples. There is no limit to what you can resolve with 1 minute exposures, provided you take enough of them.


I agree with you- noise level apparent in a single frame is not important.

[PaulM]

Hi Tim - thanks for the answer:









Generally, it's usually a matter of the Aperture of the objective lens.
A majority of astroimages are done with a telescope, but not all. I guess it depends on what you are trying to accomplish. A camera lens is usually smaller and will give you a wider field of view. Using the 100-400mm gives me more flexibility than using a fixed focal length telescope. But the telescope has a larger aperture and longer focal length which will have greater light grasp and resolving power.(Greater Detail), and increased image scale.


BTW, you don't need a telescope mount to take pictures of the night sky. You will if you want to do something like these two pictures, but there are other things you can do also.



I've been out all weekend and only now have the opportunity to reply. I was referring specifically about capturing magnified images of galaxies, nubulae etc. I remain a little confused about two things though.






Generally, it's usually a matter of the Aperture of the objective lens.



<div style="CLEAR: both"]</div>



However a Canon 600mm f/4 hasa bigger aperture than the telescope you used, which was 600mm f/7.5 if I read it right. I have no idea which is more expensive. The canon lens is not cheap.


Then you said:



BTW, you don't need a telescope mount to take pictures of the night sky. You will if you want to do something like these two pictures, but there are other things you can do also.


but that was my main point of surprise because the second shot was not from a telescope but instead a Canon 100-400mm f/5.6, yet it shows what I typically thought required a long focal length.


In my limited exposure to telescopes I've encountered the term "light bucket" which was explained to me as "not a particularly long focal length but very large aperture to capture the light". My expectation would be a focal length not dissimilar to a telephoto camera lens but a monster aperture. Conversely if it were simply this description then I don't know why people wouldn't use these "light bucket" style telescopes more often for land based photography, yet I don't believe they do...


I understand your reference to the telescopes being prime lenses when compared to a camera's zoom lens. I'd expect these long+stacked exposures are sensitive to distotions that are lesser in prime lenses.


Great video. Nice alternate use of your obvious skill at time-lapsefor stacking!


Paul

[Jon Ruyle]

However a Canon 600mm f/4 hasa bigger aperture than the telescope you used, which was 600mm f/7.5 if I read it right. I have no idea which is more expensive. The canon lens is not cheap.


I'll bet the 600 f/4 would do far better (due to its larger aperture) than the 80mm Celestron ED, but it is much more expensive and much heavier (thus requiring a far more expensive mount). A more balanced comparison would be between a fast 6" refractor and the 600 f/4, or between a 5.5" refractor and the 800 f/5.6.



In my limited exposure to telescopes I've encountered the term "light bucket"


The term "light bucket" originally referred to a large aperture telescope of not-so-great optical quality- usually a large dobsonian. Today people use the term for any big dob (I've never heard anyone refer to anything but a dob as a light bucket). The term, as far as I know, has nothing to do with focal length. But today's big dobs are getting *fast*. f/3.3 in apertures of 20" and up is no longer unusual.

[tkerr]

However a Canon 600mm f/4 hasa bigger aperture than the telescope you used, which was 600mm f/7.5 if I read it right. I have no idea which is more expensive. The canon lens is not cheap.


Then you said:



BTW, you don't need a telescope mount to take pictures of the night sky. You will if you want to do something like these two pictures, but there are other things you can do also.


but that was my main point of surprise because the second shot was not from a telescope but instead a Canon 100-400mm f/5.6, yet it shows what I typically thought required a long focal length.
<div style="clear: both;"]</div>





Let me clarify that comment. I should have said you don't need a Telescope or a expensive Mount to take pictures of the night sky with a DSLR.


Those images are cropped down a little which is probably making it look more like an image from a longer focal length. They are about 80% of the original image.
Additionally, Pleiades, and Andromeda are pretty large Deep Sky Objects, you don't need a lot of focal length for those.





The Canon 100-400mm lens is three times the cost of my C80ED. I started using it out of curiosity to see what it could do, and it gives me the versatility that a fixed focal length refractor doesn't.
A Canon 600mm f/4 would most likely cost more than all my telescope and current camera equipment together.
The things I could do with such a lens. [^o)]

[Baker]

How did you stack you star trails image, Photoshop Or Startrails?






Actually I used ImageStacker for that shot

[tkerr]

How did you stack you star trails image, Photoshop Or Startrails?
<div style="clear: both;"]</div>





Actually I used ImageStacker for that shot
<div style="clear: both;"]</div>





I think I've got that hiding somewhere, on my USB Hard Drive. I remember downloading it but never got around to trying it out.