I just spent the last couple of weeks at my field research site in Ecuador, and on our first night up high, we had a (very!) rare clear night. I was able to situate myself to capture a section of the Milky Way spreading across the seracs of the steep icefall that marks the glacier terminus to capture this scene. This location is less than 1 kilometer north of the equator (and at elevation of around 4650 m (15,250 ft), so the stars in general and the Milky Way in particular present a bit differently than we are used to in the mid-latitudes. Clear nights like this at altitude result in a lot of radiational heat loss, so the cold night air shuts down the melting process for a while - a brief reprieve from the onslaught of the tropical sun that otherwise wears away at the ice, clouds or no.
Specific Feedback Requested
I was really excited to have the opportunity to capture a Milky Way scene over this tropical glacier, but after some of the most intricate and challenging post-processing I’ve yet attempted, I just feel like this image is a near miss. I really struggled with how to make the transition from starry sky to sharp ice features look as appealing as possible, and it just looks a little fake to me (see technical details below). It’s also a bit of a different perspective view on a glacier than many will be familiar with, which perhaps also makes it looks less natural, despite this being the scene that actually existed in reality.
Technical Details
17 mm @ f 2.8; 10 seconds (multiple exposures); ISO 1600.
The night sky is a product of nine different 10-second exposures processed in Sequator to reduce noise. The ice and cliff silhouette are from a single 30 second exposure, with the product of the two processes blended into one base scene. A few areas where the ice/sky edge didn’t perfectly overlap were cleaned up using a very narrow clone stamp. Some initial tone adjustments were made in Lightroom, including increasing exposure on the ice itself by around 2.8 stops. I also brought the white balance on all constituent images down to around 3800 K (from 5200 K in camera) to create a cold feeling that accentuates the cold of the glacial ice. Additional processing in PS required a total of 13 layers, 7 of which were local dodge or burn operations applied using luminosity masks. Others were curve, level and saturation adjustments applied using luminosity masks. I also applied an Orton Effect at 60% opacity for the ice, then ran the ice through Topaz DeNoise AI. I also cropped about 1/6th of the right side of the image tp create a 4x5 image ratio.
Hopefully this isn’t too much information! What, if anything, would you have done differently?
Overall it is pretty dark… I would have gone 20 sec. at f2.8 for the Milky way and would have probably used iso3200… There is a lot more detail that could be brought out…
The main thing I would have done differently is frozen to death! Other than that, not much. Most MW images have more color brought out but interestingly, this is how it actually looks – we don’t see the color that a camera can record.
I think this is dramatic and the clarity of the high-altitude air is wonderful. You have avoided the often-seen super-dark vignetting in the upper corners, and I like the unusual (to me) color of the ice. Most higher-latitude MW images show the warmer tones of the galactic center but that appears to be behind the ice here.
There always seems to be a cutout effect when the FG is exposed differently, and for us lowlanders that is usually mitigated by light pollution or airglow toward the base of the sky, so the purity of your environment has caused an interesting but natural effect.
Hey Jeff,
Nice attempt here. I’ve always found it hard to make sense of a de-noised foreground and a sharp night sky. It makes for an automatic detraction for me and when my own work has it I just start over, lol. Here’s what I mean:
Second, I would try photographing that foreground for a longer exposure, like ISO 1600 at 3 or 4 minutes and then you’ll have to use some tasteful de-noise to get it right and to remove hot pixels, but it will return a much cleaner result and more detail in the FG.
What a scene! The ice is really cool (pun intended).
Take anything I say with a grain of salt; I’m still learning night images myself.
As long as you are blending in a separate image for the land portion, I would have taken one when the light was brighter. My experience so far is that it is really hard to get good land detail when the conditions are best for the Milky Way. My full-frame is much better than my crop-factor was, but I still find the images soft and lacking detail. My work around, so far, has been to shoot the land portion while there is still light in the sky (after the sun has set so it is non-directional). I then leave the tripod in place, and come back to shoot the Milky Way. I have to darken the land image to match the actual conditions when the Milky Way was photographed, but having detail to darken is much easier than not having detail to brighten.
A second bonus is that it is often much easier to build a mask of the land vs. the sky in the image taken at twilight. I find that if I move the Milky Way image slightly down when blending, it gets rid of any overlap with the land from the Milky Way image and leaves a much more natural transition between the land and sky. (Hopefully I wrote that in a way that makes at least a little sense?)
Fabulous night sky image. Those stars are amazing, as is the MW of course.
I’m not able to offer any advice on what to do differently, since I’m pretty inexperienced with night time images - oh, and a -Zero experience at 15k on Equator!
From a layman’s perspective, yeah, a little more light (a touch) on the rock formations and maybe not so blue? But then I wasn’t there to experience.
You’re absolutely right, Matt. The de-noise transition is much too harsh, and that’s something I’ll need to address. The foreground is actually blended from a 30 second exposure whereas the others are shorter exposures processed through Sequator (irony points for using Sequator to processes an image from the Equator).
@Lon_Overacker@Diane_Miller@John_Williams@Dan_Kearl . Thanks for the feedback on this. I’ll keep poking at this image with some of the suggestions you’ve raised. Letting it stew for a while is probably a good decision, too.
I’m late on this one, but I like it a lot so I wish to share my thoughts anyway.
I can imagine that the transition between sky and FG evokes some comments, and I can image that photographing the FG at twilight makes life easier. I never tried anything comparable, so I suppose that this is good guidance.
But I love the overall feeling in this image and the way you balanced the FG with the sky. I have seen so many night scenes that looked like a daylight landscape with a night sky glued in, completely unnatural. Too much detail in the FG or a too bright FG are very far from our own observation, so this should be handled with care (or tasteful, as @Matt_Payne commented). Looking forward to any of your succeeding attempts.
I’m also very late here and can’t add much that hasn’t already been said but I have to say that the sky portion is very nicely detailed and not overly processed like so many milky way images are. I prefer a much more natural scene like you’ve attained here which I guess is why I don’t care for most milky way images that have the life processed out of them and just don’t look real at all. As for the foreground, I agree with others that either a much longer exposure or an exposure just after sunset when you can get detail without a ton of noise are the way to go. If you get another opportunity at a clear night while you are there I think it would be worth the effort to try this one again. Otherwise, just enjoy this one.
Hey Jeff! Actually, second what Matt is saying here. And if it were up to me, I would even brighten the glacier some to allow me to find it a little bit more easily. But again that might make it look a little bit more on the fake side than you were going for.
Thanks for adding your helpful comments @Han_Schutten, @David_Haynes, @Nelson_Stegall and @David_Johnston. You inspired me to go back and take another crack at processing, starting from scratch. After a good five hours of working and reworking, Version 2 I think represents the limits of my current skills with LR, PS, and Topaz tools. I think the new version addresses the worst of the noise discrepancy that @Matt_Payne rightly flagged, but I still wish I could remove a few more of the issues. Pixel peeping is not very satisfying on this one.
