I can’t resist shooting a total lunar eclipse, but getting good detail on one is not easy. This was the 8th one I’ve shot, and I did one since but it wasn’t as deep into the earth’s shadow and there was too much dynamic range between the darkest and lightest parts, due to it just grazing the earth’s shadow. I’ve never seen one that was centered in the shadow, and it would be exceedingly dim at the darkest, requiring a tracker for long enough exposure.
The smattering of stars here shows that I was able to get a short enough exposure for motion blur not to be a factor. This one was not very high in the sky, which caused atmospheric turbulence to more of a factor than I would have liked.
Specific Feedback Requested
All comments welcome!
Technical Details
Is this a composite: No
Canon 1DX2, Canon 600mm f/4 at f/4, 1/20 sec, ISO 6400. Mirror lockup and remote release. I shot a series at different exposures but this one was best. I had stacked 20 and aligned them then reduced noise using a smart object in median mode, but I decided to choose one and re-process with the slightly newer software and got a slightly better result, much more easily.
I chose the TK Linear profile for the least contrast/color bias and just did a few very basic moves in LR’s Develop module – Exposure + 0.75 and Vibrance +32. Then into PS for Topaz Denoise (Low Light) and a touch of Detail Extractor + Tonal Contrast. About a 50% crop.
I didn’t get the partial sequence as this was the moon 2 hours before totality, and I’ve done it enough times before that I didn’t bother as it came out.
This may well be a case of monitor calibration differences but I did a lights mask using a midtones 2 with some tweaking to give a bit more definition to the craters and seas. No color adjustments and in fact, toned down what the midtone correction created. It’s just a thought.
@chris10, I’m just starting to stumble around with the TK masks but I’ll give it a try. I see you pulled my initial post which was the wrong version – it was the NR-stacked one before I did the newer processing, but not a major difference. I had caught the error and replaced it just after you commented.
@chris10, I tried a midtone 2 mask and couldn’t get a result I liked – no idea how you tweaked the mask or what adjustment you output it to. But I tried a Lights 2, increased its contrast slightly with a curves and output it to a curves. That let me bring out more detail in the lightest areas and I brushed a large soft black brush over the darker parts of the moon to limit it more to the lighter area. Then I did a subtle warming in the color channels of the curve as it came out a bit too pink. Posted alongside the original. Thanks for a nudge to dig into the TK masks a bit more. It will be a journey, but a very good one, I’m sure!
Diane, the upside to having a fine image such as this it can always be tweaked for various moods and thoughts such as Chris’ modification. Regardless of the look, the original is excellent and provides you with base for other ideas overall.
Thanks everyone! My goal with most images is to pull out as much realistic detail and color as the camera was able to collect, which may be beyond what the eye could see in dim astro subjects. The next eclipse I’ll hope to do a better HDR.
If I still had my sturdy Orion Sirius I could add the 2X, track the moon and expose for 30 sec or longer at ISO 100 and f/8 (if I did the math right) and bring out more of the “truth” in the subject, so I have to try to compensate as best I can. Fortunately with today’s software, that is getting more possible.
@chris10, it wouldn’t be nearly accurate enough. A star tracker can be set to follow the movement of the moon and the sun as well as the stars – all different. If it is properly leveled and aligned to center its movement around the area near Polaris that is the center of apparent rotation of the sky, it will follow a programmed object to within a pixel over an hour. Just that alignment is a precise operation. Here’s an iPhone screen grab of the current position of Polaris relative to the center of rotation. (And there are more precise ways than this polar scope.) The green dot is Polaris and by the time I get this posted it will have moved noticeably.
For the long lens used for an eclipse, that sort of precision would be excellent. For a 1200mm lens, a point on the moon would move quite a number of pixels in a second. Some where I have it noted but won’t take the time to dig it up now.
Diane, the details in both posts look great. The subtle reduction in brightness of the second does let more detail show in the brighter areas. The scattering of stars looks good as well. Yes, you can use various luminosity masks for contrast adjustment, including on the stars.
