Night sky photography

[Diane Miller]

This is a continuation of Roger Clark's thread in Landscape: San Juan Mountains, Colorado Nightscape.

I would not have thought that there would be less banding noise at ISOs below 1600, for the Canon 5D3. Excellent to know.

I've read your lens information on your web site and am still grappling with how to determine clear aperture. (I will read it a second time!) I wonder if you could do some lens recommendations (sometime, when you have time). My lenses I have now are the Canon 16-35 II and the 24-70 original. Both are horrible in the corners unless stopped down quite a bit. I'm thinking of renting the Zeiss 15mm f/2.8 Distagon T* ZE which is supposed to be excellent in the corners at wide apertures, but maybe you could save me the trouble... I have no idea how to determine its clear aperture.

A friend had a Canon 24mm and said it was horrible in the corners, with very bad coma apparent with stars, but it may have been the f/2.8.

I'd rather get (or borrow) a good tracker than stack, all else being equal. (I'm in the process of befriending some local astronomy buffs.) Good to know the Astrotrac works with a Wimberley -- that I already have. The Astrotrac seems a better investment than the Polarie, as I'd like to be able to use it with my 300, or even 600.

I'm thinking ahead to the solar eclipse in 2017, and I will do whatever it takes to be in the center of the shadow. I'd like to track to be able to stack exposures, although with the dynamic range that can be brought out in ACR / LR Process 2012, maybe I don't need to think about that. Thoughts?? (By a succession of amazing events, we made the eclipse in the Caribbean in 1998 and it was an incredible thing to witness. Unfortunately I had film equipment that pales in comparison to what I can do today.)

[Roger Clark]

Hi Diane,



I've read your lens information on your web site and am still grappling with how to determine clear aperture.

As f-ratio is lens focal length / lens diameter, so the clear aperture (lens diameter) is focal length / f-ratio.

Examples:
15 mm /f2.8: clear aperture = 15/2.8 = 5.36 mm diameter.
24 mm f/2: clear aperture = 24/2 = 12 mm.
35 mm f/1.4: clear aperture = 35/1.4 = 25 mm.

The amount of light collected is proportional to the area of the lens, so:

24 mm f/2 versus 15mm f/2.8: 12*12 /(5.36*5.36) = 5.0, thus the 24 mm f/2 collects 5 times the light from the subject as the 15 mm f/2.8.

35 mm f/1.4 versus 15mm f/2.8: 25*25 /(5.36*5.36) = 21.8, thus the 35 mm f/1.4 collects 21.8 times the light from the subject as the 15 mm f/2.8.



(I will read it a second time!) I wonder if you could do some lens recommendations (sometime, when you have time). My lenses I have now are the Canon 16-35 II and the 24-70 original. Both are horrible in the corners unless stopped down quite a bit. I'm thinking of renting the Zeiss 15mm f/2.8 Distagon T* ZE which is supposed to be excellent in the corners at wide apertures, but maybe you could save me the trouble...

A friend had a Canon 24mm and said it was horrible in the corners, with very bad coma apparent with stars, but it may have been the f/2.8.


I assume you mean the canon 24 mm f/1.4. Yes, it has pretty poor star images in the corners at f/1.4. Mind you I tend to image with the 1D4 so a 1.3x crop. The star images at 1.4 in the corners of the 1d4 are pretty bad, but get much better at f/2., though still far from perfect. I use the 24 f/1.4 at f/2, and is the lens I used for the image in the landscapes forum. It is still a great night lens. So it is a trade between a lot of light fast versus longer exposures. My newest night lens is the new sigma 35 mm f/1.4 art lens. At f/1.4, it has better star images in the corners and better sharpness in the center than the canon 24 f/1.4 at f/2.

That image, for those interested is here:
http://www.birdphotographers.net/forums/showthread.php/113887
This image, by the way makes a great 16x20-inch print.

I'll be doing most of my nightscapes with the sigma 35 f/1.4 for the next few months. it is a significant boost in light gathering over the 24 at f/2 and makes a big difference in the star and Milky Way.




I'd rather get (or borrow) a good tracker than stack, all else being equal. (I'm in the process of befriending some local astronomy buffs.) Good to know the Astrotrac works with a Wimberley -- that I already have. The Astrotrac seems a better investment than the Polarie, as I'd like to be able to use it with my 300, or even 600.

The astrotrac is great but pricey. Mind you a 300 f/2.8 is pushing things because the magnification is pretty high at that point and as the system rotates to follow the stars, differential flexure, even with very large ball heads, will cause drift. A 200 f/2.8 works well with a 1D series camera in my experience. It is hit or miss with the 300 f/2.8 due to differential flexure. I do not believe the astrotrack is sturdy enough for tracking with a 600 mm f/4 lens. For that you need a much sturdier equatorial mount. I use a Losmandy G11 mount, which, with counterweights and tripod weighs around 100 pounds for big lenses and telescopes.



I'm thinking ahead to the solar eclipse in 2017, and I will do whatever it takes to be in the center of the shadow. I'd like to track to be able to stack exposures, although with the dynamic range that can be brought out in ACR / LR Process 2012, maybe I don't need to think about that. Thoughts?? (By a succession of amazing events, we made the eclipse in the Caribbean in 1998 and it was an incredible thing to witness. Unfortunately I had film equipment that pales in comparison to what I can do today.)

For a total solar eclipse, most exposure times would be pretty short, but to keep a 600 mm f/4 centered on the sun during the entirety of totality will require a very good mount, like the Losmandy G11 (about $2000).

Roger

[Diane Miller]

OK, clear aperture just soaked in, I think. In layman's terms, it means the star dots will be bigger on the sensor at longer focal lengths, thus each one capturing more light? (Sort of.) So, the practicality of that is what? Does it mean that I can use a slightly lower ISO and smaller aperture, thus getting a darker and more noise-free background, and still retain enough brightness in the stars? Or is contrast not part of the equation? I think there was a link where you show stars captured on different sensors. I've misplaced the link but will find it. There's a lot to digest in all this.

Of course, the longer the focal length, the longer the star trails will be for a given exposure time, but I'll concede that we're talking about tracking with sufficient accuracy to take motion out of the picture. The Losmandy G11 is cheaper than the Zeiss 15mm!

[Roger Clark]

OK, clear aperture just soaked in, I think. In layman's terms, it means the star dots will be bigger on the sensor at longer focal lengths, thus each one capturing more light? (Sort of.)

Hi Diane,
Larger clear aperture is to collect more light. Focal length gives more separation of stars and more detail on objects. Star image size is a function of aberrations and diffraction. Brighter stars are larger due mainly to lens flare.




So, the practicality of that is what? Does it mean that I can use a slightly lower ISO and smaller aperture, thus getting a darker and more noise-free background, and still retain enough brightness in the stars? Or is contrast not part of the equation?

The practicality is a larger aperture diameter will collect more light from the subject. If you stop down, you are decreasing the clear aperture diameter. The larger the aperture, the shorter you can make the exposure times and still collect enough light to make a nice image. ISO does not change the amount of light you collect. ISO is a post sensor gain. For night images, you need the gain (iso) high enough to keep the noise from the sensor (read noise + noise from dark current, etc) above the downstream electronics noise. On a 5D3 I would not go below ISO 1600 because of this, and in particular the bandinh noise.



I think there was a link where you show stars captured on different sensors. I've misplaced the link but will find it. There's a lot to digest in all this.

My relevant articles on this topic are:
http://www.clarkvision.com/articles/nightscapes

and the series on exposure:
http://www.clarkvision.com/articles/index.html#exposure
In part 2, see Figures 5a and 5b for comparisons of night scenes with 15 and 35 mm lenses at f/2.8. There I show that the light from the subject is not the same from the two lenses even at the same f/ratio (not to be confused with light density in the focal plane). Again, clear aperture wins; the 35 mm f/2.8 lens has a larger clear aperture than a 15 mm f/2.8.



Of course, the longer the focal length, the longer the star trails will be for a given exposure time, but I'll concede that we're talking about tracking with sufficient accuracy to take motion out of the picture. The Losmandy G11 is cheaper than the Zeiss 15mm!

Table 1 of the nightscapes article, last column gives the rating by lens for best light gathering with exposures limited to keep stars round. Shortest focal length is not the best answer. Top lenses include 24 mm f/1.4 (Canon at f/2), and 35 mm f/1.4. A 600 mm f/4 will collect a lot more light from the subject than a 24 mm f/1.4, but of course the 600 has a small field of view. The key in night photography is to collect as much light as possible as quickly as possible. That means use the largest clear aperture lens that fits your subject in the field of view and ignore the f-ratio.

Roger