A clear sky over Flagstaff

i-865789a605dd2cf60f1a962e7f6e5406-flagstaffsky_usno1.jpg

An absolutely stunning nightscape taken ten kilometers from Flagstaff (Arizona) just three weeks ago: the San Francisco Peaks covered with lenticular cloud with the Milky Way behind. All a testament to the status of the city as the first International Dark Sky City. Having spent plenty of time in Flagstaff, I can attest to the spectacular skies at night, but obviously I’ve never seen anything like this. APOD has a (much) bigger version which you really need to see.

More like this

I was stunned when I went way north of Phoenix on clear night, the light pollution of any city does not let you have a real idea of what the stars look like in a natural sky and it is impressive, thanks for reminding me of that.

Wow. Maybe I should have chosen NAU instead of ASU after all...

In Australia, if you're ever out bush at night, some of those unlit country roads have so many stars above them its almost scary.

Pretty and spectacular. I didn't see any info on how the picture was taken, but I didn't look too far.

It's very pretty, but it looks, close up, like a montage done in PhotoShop. The edge of the hills is too sharp, and the resolution of the lower half is higher than the resolution of the higher half.

By John S. Wilkins (not verified) on 16 Apr 2008 #permalink

I also have to say that I think it is a composite. For the stars to be that bright, the shutter would have to have been open for many seconds if not a minute or more. That is long enough for the rotation of the Earth to be visible, either in streaking the stars or the landscape, based on whether or not the camera was held steady with respect to the sky or the land. Looking at the hi-res image shows no streaking in the lights in the buildings in the lower left, nor in any of the stars.

The cloud appears to be genuine. If you look closely at its coverage you can see the treeline appearing on the ridge of the SF peaks, complete with the silhouettes of treetops, but where the cloud covers the peak no such "sawtooth" is visible.

Taking that observation to the night sky horizon, you see a similar sawtooth effect, suggesting that we're seeing the silhouettes of treetops overcutting the sky. While it can certainly be argued that this is a shopped image, it's extremely sophisticated in execution if so.

Looking at the technical quality of the image exposes a high degree of pixel noise, a characteristic of an image which has been taken with a digital camera whose sensor is operating in "high ISO" mode.

This has the effect of simulating "fast" film -- but as with fast film, there's a trade-off. For high ISO films, the compromise lies in larger silver halide crystals in the photosensitive layers, resulting in vastly increased visual grain; for CCDs the compromise is electronic noise.

So it's entirely feasible that this image would be taken with an exposure of ~30 seconds, which would yield only some blurring due to rotation in the sky, though another possibility is that it's an HDR exposure. This would combine image data from multiple photographs to find the optimal mix of brightness, detail and contrast.

In any event, I question the suggestion that the picture has been shopped. (FTR, in my capacity as graphics wonk at the local hospital I also do quite a lot of photography and extensive image modification with Photoshop. I also live about 2 hours west of Flagstaff, and have never seen it represented so beautifully.)

Quick followup. A closer examination of the dimmer stars does show some lateral distortion, suggesting motion blur.

Additionally there is what appears to be either a meteor or satellite "streak" above the cloud formation, which would suggest the starfield was not taken from orbit a la HST or by similar means.

The stars also appear to be blurred in a semicircular pattern -- that is, the stars on the left side of the image seem to be blurred in streaks upward and to the right; the stars on the right appear streaked downward and to the left; and the stars in the center appear more or less horizontally blurred. This is consistent with what one would expect from a legitimate terrestrial image that had some small distortion due to planetary rotation in it.

Again, if this is shopped, it's very subtly done (it would probably have been much easier to just go out and take the photo), and as the publication date for this image was not 1 April I think it's safe to presume it genuine -- possibly not even HDR, as I originally posited. Certainly a tripod and a Nikon D40DX could do something similar.

Well, this obviously is a fake.

If the stars are THAT bright in this image, then there is a long exposure mode needed to capture the nebula, and even with sensor image noise taken into account, you will find star tracking OR ground tracking. Believe me, I do this type of exposure and run into that very problem despite attempts at pushing the "ISO" to a high level. Aurora photography is a similar situation and star trails even with very wide angle lenses greater than 30 seconds are a problem.

Next, there is no fade as one gets to the horizon. Yes, it is clear out there, but once the overhead thinnest part of the atmosphere is taken into account, the brightness of the stars near the horizon are pretty weird. The portion of the image to the right that appears to be the North American Nebula also is so well exposed that the time factor is coming into effect again, and some blurring of either the sky or the ground must be occuring if this is a sky-tracking shot. Then, IF the sky is that dark, how the hell is there enough light for the ground or the sides of the SF peaks to be so nicely exposed? Similarly, where is there enough light to show the detail on the mountain, and also the cloud. Speaking of the cloud, how is it that the whole cloud is so evenly lighted by reflection of ground light and there isn't a fall off in backscatter sky glow.

If anyone has spend even a little time in the dark parts of the country, this just plain looks fake.

Let's say that for dramatic purposes the creater of this image did an HDR image. Then it is photoshopped and is an artistic representation of what it might look like, not what really is.

I am all for preserving dark skies and argue for it all the time, but I still will hold to the gut instinct that this is a prepared image, showing a nice dark sky (quite lovely) but the ground portion was not taken at the same time.

Now that I've seen the picture, I'll study some more and ask the professional sky photographers about it and either apologize for thinking it was fake if I'm proven wrong, or furnish further insight as to other parts that raise suspicion. I don't know exactly where it was taken, but I'd wonder if the date/time information could be evaluated to see if the sky is in the correct orientation to be visible as this when the picture was taken.

Here's a suggestion ... why doesn't someone who thinks it is a fake contact the photographer and find out. The details are on the APOD link.

By John Lynch (not verified) on 16 Apr 2008 #permalink

I live in a reasonably dark area in Nova Scotia. After Hurricane Juan it was even darker for several days, as the power was out for a hundred or so km. in every direction, including the area's biggest light emitter, Halifax. It was certainly a good sky-watching opportunity, as we had clear dry weather, but it certainly never approached that photo for sheer density of visible stars. Are Texas skies really like that, to the naked eye?

Whoops! I meant, of course, Arizona skies.

Bee: In the right parts of the state, the skies are strikingly illuminated, but TTBOMK no skies appear like this to the unaided eye. The best the human eye can manage is about 3000 stars (winter skies, northern hemisphere); the rest are too dim to see. A longer-term exposure shows more, of course.

The best I've managed, ever, was in the 1980s (less light pollution/general pollution than now). There was one night when I was able to resolve the largest/most distant of the Galileans, and there was another when M31 was visible as a faint light smudge, though only in peripheral vision.

However, it's not at all unusual for the Milky Way to be visible any time, even in town (small town, not Flagstaff).

Dale: Your suggestion that an HDR image is somehow less than honest is one of semantics and perspective. HDR images are all taken within a few minutes of one another, all depict the same scene, and are intended to capture the full dynamic range of human perception, rather than the narrow and limited perspective allowable by only one exposure and setting. HDR imagery is no more dishonest than is a 3D photograph -- or a movie capture.

As for this:

it is photoshopped and is an artistic representation of what it might look like, not what really is

I challenge you to show me any aspect of human interaction with the world that is not, essentially, "artistic representation" (interpretation, sensory overlay, etc.). I doubly challenge you to show me any visual capture that isn't, on one level at least, "artistic representation". The idea that any photograph captures "what really is" -- or that any photograph can do so, or has ever actually done so -- is naive.

Note from the creator:

The image is a composite of 2 images taken in succession (one 90 seconds of the sky, tracked to follow the stars, and one 4 minutes in exposure taken 5 minutes later, untracked). I have no qualms about saying it is a composite, but object to the term "fake". Some of the speculations above are false, some are OK, some are editorial. I suppose the root of the argument is whether or not any type of hand editing is appropriate. With two images of differing exposure times, and with a slight change of shape in the cloud between them, hand editing was necessary to make the two blend and be in register. The point is they ARE IN REGISTER, that is represent the same area of the earth at essentially the same time. So the image is more of a MOSAIC than a COMPOSITE. Are panoramic mosaics fake? I suppose it depends on your interpretation of digital photography versus illustration. The photons collected by the sensor of the camera are turned into electrons and then read out to a computer, how real is that? Significant color balancing, contrast enhancement, saturation adjustment, and other adjustments were made to the raw 12 bit RGB files, converted to 8 bit RGB tiff, converted to JPG. How much is actaually left of the orginal photons?. Does it really matter? The photographer/processer (me) tried to make it look like the real thing to the best of my (limited) ability and within the limits of existing technology. I needed to take two images to do this, and do some hand editing with various editing tools. I did not spend hours and hours trying to make every pixel blend with every other, and I am sure a few artifacts of my editing can be seen on the high resolution image, but enough time to make an effective image. I think that falls within the bounds of creative photography/imaging.

I posted an explanation on another thread which I cut and paste to here.

I have no qualms in stating that the image in question is a "composite" or "mosaic."
However, two images were obtained in as rapid a succession as possible
from the same location at the same time. It would be better to have two cameras and take the shots truly simultaneously, but I can't afford that luxury. As many astrophotographers know, the stars move and the land doesn't. Today's modern color digital cameras and fast lenses ALMOST allow a photographer to get outstanding depth in the sky with an untracked mount and a very short exposure. Almost, but not quite. They also allow existing light imaging of the land under the night sky, but with a fairly long exposure (minutes). By placing the camera on a tracking mount for the sky, then repositioning the camera for the land and turning tracking off and making a much longer exposure to get more signal, both land and sky can be depicted without star trails or blurring of the land. The whole process takes about 5-10 minutes in the field. While purists will consider any type of composite "faked", and a certain amount of "hand work" is necessary to make the two images blend properly, I make no apologies for this method whatsoever. The method overcomes the limits of the technology to get a result that depicts the real world. Nor do I imply that the image is what the eye sees, it is what the camera sees, obviously, but it depicts the scene in a manner that extends the human sense of sight without giving wrong information or a misleading idea of what the place is like. I believe the image communicates the spirit of the place and the night. As an artist, I can hope to do no better.

Dan Duriscoe

By Dan Duriscoe (not verified) on 17 Apr 2008 #permalink

I don't have a problem with composites, or long exposures. I even like this picture. The problem I have with the whole thing is that it was presented as an image from the Flagstaff area, which was described as an International Dark Sky City (which I am completely behind 100%).

However, the implication which comes along with the image is that if "your" town adopted these same light policies, you could see this out "your" window. I think that is false advertising, a bit like showing a muscled hunk in a swim wear ad. "Buy these shorts and you too can look like this!" Well, no.

This image seems to say, "Adopt these policies and your skies can look like this." No, they won't. It would be great if they did. I think that combating light pollution is an important enough issue that we can promote and defend it on it's own merits, not trumped up ones.

@ Andrew

I have to admit I'm unsure that there's a problem here. You'd have to be fairly naive (imho) to imagine that the human eye alone would see the sky as depicted in the photo. And - more importantly - the fact that Dan could take the photo at all is a testament to the dark skies program in Flagstaff.

By John Lynch (not verified) on 17 Apr 2008 #permalink

Well, call me naive since I didn't even consider the idea until I read John S. Wilkins's comment. Then I went back, looked at it more closely, looked at the hi-res version, compared it to similar images of stars taken with long exposures, and came to the conclusion that this was a composite, not an image of the sky as it actually appears.

And I don't think I've done something wrong by pointing that out.