The Milky Way / State 2

[Diane Miller]

Here's a slightly different view of the FG trees, and slightly different processing of the base image.

Specs are not that different from my post from yesterday. I hesitate to bring out more contrast because this already exceeds what the eye can see, even in the clean air at high altitudes. And because even the best sensors of today, in a single exposure with a decent wide-angle lens, capture what is basically an ugly image, with noise and edge distortion. Bringing out contrast and detail brings out the ugliness.

[Don Lacy]

I very strong image Diane the foreground elements lead the viewer nicely to the night sky.

[dankearl]

I like the sky in the first one with the horizon in this one.
Very cool photos.

[Don Nelson]

Diane -- Excelllent. This image is far stronger than the previous one posted. Clear separation between foreground silhouettes and starry dark sky. I like the composition a lot.

[Morkel Erasmus]

This works well, Diane. Just the right amount of "oomph" in the Milky Way. :5
FWIW - using artistic licence I like to work with a "fluorescent" White Balance in these shots, adds some nice colour to the sky.


I hesitate to bring out more contrast because this already exceeds what the eye can see, even in the clean air at high altitudes. And because even the best sensors of today, in a single exposure with a decent wide-angle lens, capture what is basically an ugly image, with noise and edge distortion. Bringing out contrast and detail brings out the ugliness.

Well, the point for me in using the modern high ISO sensors is to capture more of what the eye can't see, I mean it's not unnatural in the sense that it is THERE and photographing the night sky like this shows folks the magnificence of the heavens. And I disagree that it's an ugly image out of camera. The point of an ultra wide angle lens is that it brings with it edge distortion. The lens you use will make a big difference indeed - the 14-24mm I use is amazing in its distortion control and it's corner sharpness. :t3

[Diane Miller]

Thanks, everyone!

I have no problem in going beyond what I can see. Morkel, I don't know that lens -- are you a Nikon user? The new Canon 16-35 f/2.8 II wide open is awful in the corners. I'm thinking of getting the Zeiss 15mm f/2.8 Distagon T* ZE.

The point of using an ultrawide for me isn't to get edge distortion, not when it is coma and chromatic aberration, and general smeariness. The problem is, to shoot the dim light of the night sky, you need to shoot wide open, or trade off to ISOs higher than 3200., where noise blends with the smallest stars. I'm looking at other methods, such as those astrophotographers use.

[Morkel Erasmus]

Diane, it's the Nikkor 14-24mm f2.8
As sharp as a 14mm prime and 24mm prime f2.8 at both ends and in between :bg3:

[Martin Dunn]

Stunning photo Diane.
The colours and textures are mesmerizing.

I've just taken a look at your previous post.
IMHO this one is way better. Draws me right in.

[Diane Miller]

Thanks, everyone!

[Roger Clark]

Of your 3 images of the scene, I like this one the best. How are you brightening your image in post processing? If you use something like levels, you clip the brighter stars turning them only white. I only use curves and keep the end points fixed so I do not change any clipping. That helps maintain star colors.

Roger

[Diane Miller]

I never use Levels, only Curves, and try to be very careful about clipping. (I don't move the end points except in very low contrast images where there is no information anywhere near them, but have not done that with the star images.) The Milky Way in this image (the tree is composited) was with the 24-70 at 24 and f/2.8, ISO 3200, 20 sec. The RAW file was processed in LR 5.whatever-it-is-now. Camera Calib. was Adobe Std, but going back to a copy of the RAW file, changing to Camera Faithful only makes a very subtle difference in the background sky becoming slightly darker -- no apparent change in star color. The Curve (in LR) is linear by my default. Sometimes I'll change it to medium contrast, but in everyday shooting I'm usually more concerned with reducing contrast. Luminance NR upped 25 (I don't usually do that). No HSL, Sat or Vibrance adjustments. Slight boost in Highlights and drop in Shadows, and slight Contrast boost. Temp 4000, Tint -6.

Going back to a virtual copy as imported, the only color I can find in the stars is some slight haloing that looks like CA. Playing with adjustments didn't result in a noticeable difference in the stars at 1:1.

In PS, processing was just some subtle Structure with Viveza on the sky, then denoise, and then the trees were composited.

Maybe the stars are just too small to register color, even at 24mm? The WB doesn't seem to be affecting them, only the dark sky and the "dust" in the Milky Way.

I have gotten a very nice range of star colors in several recent star trails, taken from our rural property in a semi-dark area on clear nights. I shoot them at ISO 200, 16-35 lens between 16 and 24, f/6.3, 3 minute exposures stacked.

I may not have time to check in till Monday evening -- I appreciate all the help and I'll keep reading your material!

[Roger Clark]

Maybe the stars are just too small to register color, even at 24mm? The WB doesn't seem to be affecting them, only the dark sky and the "dust" in the Milky Way.

I have gotten a very nice range of star colors in several recent star trails, taken from our rural property in a semi-dark area on clear nights. I shoot them at ISO 200, 16-35 lens between 16 and 24, f/6.3, 3 minute exposures stacked.

I may not have time to check in till Monday evening -- I appreciate all the help and I'll keep reading your material!

Hi Diane,
The star colors come out better on star trails because the trails are not as bright as a single exposure wider open and higher ISO.

Things to try: reduce ISO to 1600 to give more dynamic range. Remember, changing ISO does not change the amount of light gathered.

In raw conversion, make sure to do no increases in exposure or brightness (not sure what the lightroom sliders are called); in PS-ACR the exposure and brightness sliders should be set to zero. Then all brightness increases with curves in post processing.

Roger