Moon w/ Sigma 300-800

[Steven Kersting]

Just for you Roger, I hadn't forgotten. :S3:
Nikon D3, Sigma 300-800 f/5.6 @ 1/125, f/8, 800mm, +1/3 EV, edited in LR.
http://skguitar.com/photos/Moon/800.jpg

Link to image at full resolution: SK13677-2.jpg (http://skguitar.com/photos/Moon/_SK13677-2.jpg)


As this was done as a result of a discussion on the sharpness of zooms and benefit of TC's I have done several more shots/edits. I will be adding them over time as per forum guidelines (1 per 24hrs).
In order to make this as "fair" as possible all images were edited only in lightroom 3 and all were given the exact same sharpening treatment (arguably excessive). I made no special attempts to fine tune any of the images or to optimize quality upon resize/output.
This isn't for "critique" as such, it is just to show what is achievable with my gear in different combinations with minimal fuss, and the benefits/ differences of using TC's.
(I was actually pleasantly surprised by some of the results I got)

[Dave Mills]

Hi Steve, love the sharpness and the detail. I also like the way you handled the image even with that amount of neg space. Roman has the same lense and one would think that with the lenses range sharpness might suffer. Definately not so. His images are as sharp as my eye can discern. I just wish the the lense was lighter...

[Roger Clark]

Hi Steve,

Nice image at this scale. How about posting a crop with 1:1 pixels (1 camera pixel = 1 output pixel)?

Roger

[Steven Kersting]

Hi Steve, love the sharpness and the detail. I also like the way you handled the image even with that amount of neg space. Roman has the same lense and one would think that with the lenses range sharpness might suffer. Definately not so. His images are as sharp as my eye can discern. I just wish the the lense was lighter...

No kidding! I often use this thing mounted on a shoulder stock and carry it up to 4 miles...If a shot is going to take more than 30 seconds or so I have to add the monopod and I always end up just a bit wore out.

[Steven Kersting]

Hi Steve,

Nice image at this scale. How about posting a crop with 1:1 pixels (1 camera pixel = 1 output pixel)?

Roger

Not sure what you mean Roger. Above is a link to the original file un-resized...too big for posting here.

But here is a 1:1 crop of just a 1024x section.
http://skguitar.com/photos/Moon/800cropped1_1.jpg

BTW, is this a first quarter or last quarter moon, and is one better than another for getting detail?

[Dave Mills]

Hi Steve, very impressed with the detail....I almost feel I'm getting ready to land on the planet.

[Steven Kersting]

Hi Steve, very impressed with the detail....I almost feel I'm getting ready to land on the planet.

Thank you. I have another version which I have "optimized" to my taste....I'll get to that eventually.

[Roman Kurywczak]

Hey Steven,
I use that lens often.......so no surprise to me how sharp it is! As this was just for more discussion....will leave all the negative space discussion alone:w3. Repost is pretty amazing in detail!

[Roger Clark]

Not sure what you mean Roger. Above is a link to the original file un-resized...too big for posting here.

But here is a 1:1 crop of just a 1024x section.

BTW, is this a first quarter or last quarter moon, and is one better than another for getting detail?

Hi Steve,

Yes the 1:1 image is what I meant. The image is a few days before first quarter. Each day shows different detail, so no one day is better than another in my opinion. Closer to first quarter (when the lunar disk appears half lit) is when there are more craters in view for these short focal lengths.

Your image, however is way over sharpened. The edge of the Moon, for example, shows a detached bright line due to ringing in the sharpening step. That ringing also distorts detail in the image.

I dug up an image from July 16, 2010 that is almost identical phase angle as your image. I made the image with a 1D4, 500 mm f/4 + 2x TC (so 1000 mm compared to your 800 mm). I've attached a section of that image on yours and then rescaled that section to the same sampling as your image (I did no sharpening after that resampling). Your image seems to lack many of the small craters. I'm not sure why, perhaps it is the processing? Also, the Moon is reddish, so I kept the color (daylight color balance). Do any of the small craters show in your original image?

Side note: The reddish color was poorly understood until the last decade or so. The reddish color is mainly caused by nano-sized metallic iron particles in the lunar rocks and soils. Such nano-particles cause Rayleigh absorption (in contrast to Rayleigh scattering which cause our sky to appear blue).

Roger

[Roman Kurywczak]

Could that also be due to sharpening method you use Roger?

[Roger Clark]

Could that also be due to sharpening method you use Roger?

Roman,
No because many craters show in the original out of camera jpeg. For example, I'm attaching a portion of the out of camera jpeg with no sharpening and craters pointed out that are not on steve's image. And these are large craters. Crater A is in one of the mare with no other craters around. Craters at B are a series that are only faintly seen as hints of shading in Steve's image. The 1D4 out of camera jpeg shows many smaller craters too.

This illustrates why, in my opinion, the Moon is a great test target. It has higher contrast than a simple bar chart test target, and has essentially infinite detail for Earth-bound observers, and everyone can compare the same subject no matter where on the Earth they are (a little more challenging at the poles), and throughout many lifetimes.

Note the slight color fringing on the out of camera jpeg below, which also adds to image softness. That is a small amount of chromatic aberration that I fixed in the raw converted image above.

Roger

[Steven Kersting]

Interesting Roger. I accept the over-sharpening comment. Since all images were done identically and in LR I chose one setting which was not ideal for any image but works ok for all. I could certainly have done better.

I don't know why the small crater indicated isn't showing....it's simply not there. It does show at 2400mm (800 w/ 2x+1.4x TC's, part of what I was surprised by (but not thrilled by)). I live in an area with a pretty good amount of light pollution and I didn't travel to try to avoid it. Would that be a likely cause?

[Roger Clark]

Interesting Roger. I accept the over-sharpening comment. Since all images were done identically and in LR I chose one setting which was not ideal for any image but works ok for all. I could certainly have done better.

I don't know why the small crater indicated isn't showing....it's simply not there. It does show at 2400mm (800 w/ 2x+1.4x TC's, part of what I was surprised by (but not thrilled by)). I live in an area with a pretty good amount of light pollution and I didn't travel to try to avoid it. Would that be a likely cause?

Hi Steve,

Regarding light pollution, it is not a factor. But atmospheric turbulence could be, although it would show as a soft edge on the lunar disk and in the (ideally) sharp shadow boundaries.

There are 3 things that could contribute to the disappearing craters:
1) atmospheric turbulence,
2) the crater fell on the intersection of 4 pixels so didn't get recorded well,
3) lens is not very sharp.

Atmospheric turbulence would show in soft edges, but so would #3. #2 is a possible factor and I have seen this in images. Evaluating which needs an image less modified by sharpening. Your stacked TCs image that shows the crater would be helpful to evaluate the cause(s). #2 seems less likely, because it isn't just this one crater as many small craters seem to be missing. I would not expect a zoom lens to be as sharp as a super telephoto fixed focal length lens, so #3 is probably a factor.

Did you use mirror lockup?

Roger

[Steven Kersting]

I realized there's one other factor...
I manually focused using live view and I recently became in need of either bifocals or removing my glasses for anything inside of 2 ft....I haven't quite adjusted to it yet. (part of the reason I tend to over sharpen in post). Getting old sucks!

I used mirror lockup and generally gave 6-10 seconds for the camera to settle before firing....recomposed and refocused between shots using liveview (separate steps w/ nikon)....It started to generate sensor heat artifacts, but nothing major.

[WIlliam Maroldo]

I don't understand why the moon is so small in the image. I've shot the moon at 500mm and it seems that it filled more of the frame. I am not impressed at all by the image quality. If it is only the lens it would not be happy with the resolution. Using TCs won't do anything but make the resolution worse. Of course we can't rule out other variables, like atmosphere.
Since there are so many variables involved in taking a good lunar image, why would this be considered an appropriate test target? In any valid scientific test all variables must remain constant, except for one. I can't see how a valid result can be obtained when your target is moving roughly 1 km per second and atmospheric changes are so unpredictable.
A few questions for Roger. 1) at 100,000 miles it would seem even a wide open aperture would give the necessary 1,080 mile DOF 2) When attempting my own shots of the moon I've noticed it moves surprising quickly across the sky, at least enough to need faster shutter-speeds than I would have guessed. Is it possible that 1/125 second is too slow without some sort of tracking system?
regards~Bill

[Roger Clark]

I don't understand why the moon is so small in the image. I've shot the moon at 500mm and it seems that it filled more of the frame.

The original image was make on a full-frame sensor at 800 mm, so the field of view was 2.58 x 1.72 degrees. The Moon should be relatively small in the frame.



I am not impressed at all by the image quality. If it is only the lens it would not be happy with the resolution. Using TCs won't do anything but make the resolution worse. Of course we can't rule out other variables, like atmosphere.
Since there are so many variables involved in taking a good lunar image, why would this be considered an appropriate test target?

I disagree about TC's unless the lens is really bad, which I do not believe it is. The Nikon D3 has 8.46 micron pixels, so at 800 mm, one pixel subtends 2,2 arc-seconds, and the blur filter smears the resolution to about 4 arc-seconds. This is generally not a problem for the atmosphere. (see responses below.)



In any valid scientific test all variables must remain constant, except for one. I can't see how a valid result can be obtained when your target is moving roughly 1 km per second and atmospheric changes are so unpredictable.

Velocity in km/second is irrelevant when so far away. The key is angular velocity. The moon moves at a maximum about 15 arc-seconds/second. With 2.2 arc-second pixels and a need to have smear less than 1/2 pixel, then shutter speec should by faster than 1.1/15 ~ 1/14 second. As the image in question was made at 1/125 second, the Moon moved only 0.12 pixel. The Moon is generally bright enough to use fast enough shutter speeds to avoid motion blur.



A few questions for Roger. 1) at 100,000 miles it would seem even a wide open aperture would give the necessary 1,080 mile DOF 2) When attempting my own shots of the moon I've noticed it moves surprising quickly across the sky, at least enough to need faster shutter-speeds than I would have guessed. Is it possible that 1/125 second is too slow without some sort of tracking system?
regards~Bill

See above.

Other comments. When photographing birds at the beach, or wildlife at a distance, the atmospheric turbulence is often greater than that found when imaging the Moon when the Moon is high in the sky. I've seen lenses that can make a nice image of a bar chart, but lousy images in the real world, where higher dynamic range will reveal flare and chromatic aberration. The Moon provides a good target to test more of a lens than a bar chart does, and it is there for everyone to use. At supertelephoto focal lengths, other variables are less important. The disadvantage is you have to wait for the Moon to be up in a good clear sky. In my opinion, it is a good training target as well as a lens testing target. The only tougher test of a lens is a star test.

Roger

[WIlliam Maroldo]

Thanks Roger for the quick reply! I was comparing moon size size with how it looked with with a ASP-C sensorand if it was with a full-frame sensor it makes sense. I appreciate you explanation of the the minimum shutter-speed necessary...I always wondered.
What about DOF (my last question)? regards~Bill

[Steven Kersting]

Roger, what's a star test? I'm always interested in a challenge.

I agree that moon shoots can be a very good training tool as at long focal lengths it requires all of the techniques to get a sharp image and it seems it will always fool auto exposure with any settings. (regardless of my posted image).
For me it also exposed a possible issue with my eyes and live-view.

Since RAW capture sharpening is subjective, and output sharpening is "automatic" in many softwares, would you suggest that Jpeg fine with "default" settings is a more accurate test?

[Roger Clark]

What about DOF (my last question)? regards~Bill

at 100,000 miles it would seem even a wide open aperture would give the necessary 1,080 mile DOF


The distance to the Moon is about 240,000 miles, and the moon is about 2140 miles. The depth of the scene would then be 1070 miles. The focus difference from the center of the disk to the limb with a 1000 mm focal length lens would be about 0.000017 microns, or over 20,000 times smaller than the wavelength of visible light. So depth of field is not a problem at any aperture. Somebody can check my math--it's been a long day,:w3

Roger

[Roger Clark]

Roger, what's a star test? I'm always interested in a challenge.

Hi Steven,
A star test is simply photograph stars. On a fixed tripod, point at the celestial pole (the star Polaris in the northern hemisphere) and make an exposure. For example, a 30 second exposure with your 800 mm lens, evn in the city will show many stars.
For example, here are some images of Polaris with my 500 mm lens:
http://www.clarkvision.com/articles/cropfactor/
then examine the stars near the edge of the frame. On my 500 wide open, stars are not slightly oval.

Here is another example with a 28mm lens at f/2.8:
http://www.clarkvision.com/galleries/gallery.africa_2011/web/ndutu.stars.c02.22.2011.img_6385.e-900.html
Even in the small web image, you can see the stars at the edge of the frame are not round, and at 100% view, the stars are really poor. I need a better wide angle lens, but I've not seen one with significantly better performance.

You can also use distant city lights as a star test.



I agree that moon shoots can be a very good training tool as at long focal lengths it requires all of the techniques to get a sharp image and it seems it will always fool auto exposure with any settings. (regardless of my posted image).
For me it also exposed a possible issue with my eyes and live-view.

Since RAW capture sharpening is subjective, and output sharpening is "automatic" in many softwares, would you suggest that Jpeg fine with "default" settings is a more accurate test?

I agree about exposure; one really needs to use manual exposure unless the Moon fills the frame. For uniformity with one camera, the in-camera jpeg is certainly unbiased, but I think you can achieve the same with a raw converter as long as the same settings are used on comparison images. As long as you don't produce noticeable ringing, I think anything one can do to produce a great image is fair game.

Also try imaging the Moon hand held like you would birds in flight. The Moon is illuminated by the sun so exposures are similar to daytime terrestrial subjects.
Here is one of my attempts at 1000 mm:
http://www.clarkvision.com/galleries/gallery.astrophoto-1/web/moon.rnclark.handheld.c10.25.2007.jz3f6583f-8s-800.html

Roger

[WIlliam Maroldo]

Thanks again Roger. I thought that had to be the case. I tried to verify the DOF at 240,000 miles and I've been having trouble with using 1,267,200,000 feet with my handy DOF calculator!:w3:w3
regards~Bill

[Roger Clark]

Thanks again Roger. I thought that had to be the case. I tried to verify the DOF at 240,000 miles and I've been having trouble with using 1,267,200,000 feet with my handy DOF calculator!:w3:w3
regards~Bill

Yeah, I was down to the last digit on my calculator.

Roger