[interested_observer]

Evening everyone - I really haven't been shooting much since last October. Yard work (900 feet of Oleanders), weeds, filling a 40 yard dumpster and then a stint in the ER/hospital and some surgery has filled my winter. But photography wise, I've been somewhat busy trying to put together a different work flow - mainly to work around my K1 white dot issue from shooting astro landscapes. If I would have been more diligent (read less lazy), I would be done and could say what worked and what didn't. I now know what I really want to do (I think, kinda of, sorta of), and somewhat sure of what will work and what will not work for me. All of this backs into how I probably need to capture my images going forward.

I've been using this image to test Camera One (which I have given up on - which works very well on the white dots, but fails on pulling out the Milky Way's natural colors), to the point that I'm sort of sick looking at it. I purposely brightened it when I first processed it (to see what's going on - this is a test shot/PP), and since then, I have been working in the back of my mind as to just how I want to approach things. Which brings me to this post.

I ran across this youtube video on the topic of - "CHIAROSCURO — The Dynamic Range Mistake", which struck a cord with me. The essence of the video is, just because you have a camera capable of capturing great dynamic range - the ability to pull detail out of the shadows, should you (in post processing)? What is sinking in, is that just because you have the dynamic range captured, you can pull it out in varying degrees in post processing.

You Tube

You Tube

A couple of months ago, I also came across a discussion on color theory and composition, and how the great landscape painters of the past approached the topic. It all started to jell quite a bit when I watch the video on dynamic range. I always knew that I wanted to have it look like it was taken at night - be somewhat dark, but I also wanted some detail, and I realized that the texture in the trees provided a real nice 3D sense of depth to the image.

I was able to get one night of shooting in back in early Feb, but the location was a complete bust. The light dome from Phoenix was too bright and at 3am I was casting a shadow from 50 miles away. Which in retrospect is what is lighting the trees in the above first image (70 miles from Phoenix). It had not occurred to me previously. The city light dome gives in some locations, while in others - it takes away.

Last October, I drove up to Oak Creek Canyon - at the northern end and shot this panorama. I was way too concerned with the white dots (used LENR), and capturing the Milky Way, but finding that going from 50 second exposures to 70 second exposures with the K1 with astro - captures tremendously more color, definition and contrast in the core (duh). I'm finally discovering the obvious. However, since it was a 2 minute exposure and a 2 minute dark frame, I was finding it very difficult to compose what I was actually capturing with the cars on the switchbacks going down the side of the canyon. I think a few more images stacked with the car lights illuminating the canyon walls, coupled with a hint of pulling up the details, I might start getting what I'm envisioning.

It's finally dawning on me, that I'm trying to "paint" with the camera. I had to take this art appreciation class 50 years ago as a graduation requirement - which really interfered with my engineering classes at the time. Little did I realize that I actually remembered something from that class.

So, the next two weeks the Milky Way is available again (well with out the Moon). I want to return to the windmill location and try some new approaches, along with a few different location that I have shot in the past.

[biz-engineer]

The Dynamic Range Mistake

hum, the dynamic range mistake is to not care about the quality of lighting and count on camera dynamic range to balance tones again in post. The problem of recording images with DR higher than 10 stops (10 zones), is that the recovery by post processing is never as good as if the scene was exposed to soft light in the first place. So, in fact, having sensors that can record 14 stops, is a luxury that can spoil us bad photographers.

---------- Post added 31-03-19 at 09:08 ----------

trying to put together a different work flow - mainly to work around my K1 white dot issue from shooting astro landscapes

The white dot issue is more a pixel peeping problem. The milky way containing billions of stars, I don't really see how a half dozen of white pixels can change the look of an image, no one will tell the difference on the print even printed and looking at 10" close. On top of that, removing dark frame of the same exposure time and ISO eliminates most white dots. For example, we see a lot of backlit portrait shots, it's an original play, it stresses camera dynamic range capabilities, but if we'd ask for what images are preferred by anyone not being "photography aware", they'd mostly prefer the front-lit images.

---------- Post added 31-03-19 at 09:20 ----------

CHIAROSCURO

The Pentax default tone curve is such that it yields good looking images most of the time. I've been pulling shadows on images, and at the same time realized that more details from the shadows did not necessarily make the images look better. The darker shadows are like blurred background or perspectives, they are components of an image that forces the eye to look at the main subject. The problem is "What is the subject?" , in a lot of photographs, we are confused about what the subject is. Because it's the subject that ideally should get proper light exposure relative to the remainder of the scene.

[pixie]

I agree with you a lot. In my view it's all about the presentation of the dark. How do you make an image stand out in its darkness? How do you make it presentable and likable? Our modern cameras have great HDR. I think it's all about post processing, and making the image "likable" to the masses.

[biz-engineer]

I think a few more images stacked with the car lights illuminating the canyon walls, coupled with a hint of pulling up the details, I might start getting what I'm envisioning.

I much prefer your first shot with the windmill. For the second shot, maybe there is no white dot into it, but it's way too dark to appreciate anything, most of the frame is dark and the lit areas hardly show any subject.

The windmill photo is great. How about going there at dusk, making an exposure for the bottom part of the frame at dusk and photographying the upper part of the frame (milkyway) a few hours later, then exposure fusing the two frames together?

---

Last edited by biz-engineer; 03-31-2019 at 12:52 AM.

[GUB]

Great thread and link OP. Totally agree with your progression of thought. The second shot feels real like one is standing there and can reach out and touch the Milky Way.
I think the art of the dark areas is to use the right amount of detail to keep the image in balance. The slashes of light across your foreground is all that is needed.
And the sky shouldn't be a contest of how many stars we can record -- Your first image simply isn't reality and I think when you go out and record our incredible Universe there is no need to pimp up reality.

[Rondec]

It's all a matter of taste, isn't it?

I tend to like shadows that are lightened a bit. It does tend to be tough if most of the image has basically crushed blacks and is total darkness. At the same time, if you lighten the shadows too much, then it feels like some sort of alternate reality as we know at night we can't see things that well.

[D1N0]

I tend to lift shadows for more detail and the increase blacks so the image doesn't look to have flat lighting but still has detail in the shadows.

[clackers]

It's very rare for an actual chiaroscuro painting to have crushed blacks, BTW. There are nearly always 'recovered shadows' in the background.

[interested_observer]

hum, the dynamic range mistake is to not care about the quality of lighting and count on camera dynamic range to balance tones again in post. The problem of recording images with DR higher than 10 stops (10 zones), is that the recovery by post processing is never as good as if the scene was exposed to soft light in the first place. So, in fact, having sensors that can record 14 stops, is a luxury that can spoil us bad photographers.

I didn't say that I didn't care about dynamic range, however - what I'm striving for is something other than a solid black silhouette of the foreground. I really don't want to go the light painting route, either. So, where I'm trying to go is trying to balance the lifting of the details within the shadows a bit to break it up. It all depends on just what is in the foreground that has some light colors that naturally lift themselves to a reasonable extent.

In terms of a correct exposure, the question becomes what is a correct exposure of darkness? When you're out in the desert and unable to see your hand in front of your face, the question is 2, or 3 or 4 or 5 minutes of exposure for the dark foreground, and at what ISO 100, 200, 400 or 800. There really is no way other than to try a couple of exposures and then try to make an educated guess. Yes, the camera has a light meter, but there is no real dark meter.

The white dot issue is more a pixel peeping problem. The milky way containing billions of stars, I don't really see how a half dozen of white pixels can change the look of an image, no one will tell the difference on the print even printed and looking at 10" close. On top of that, removing dark frame of the same exposure time and ISO eliminates most white dots. For example, we see a lot of backlit portrait shots, it's an original play, it stresses camera dynamic range capabilities, but if we'd ask for what images are preferred by anyone not being "photography aware", they'd mostly prefer the front-lit images.

On this topic, I absolutely disagree. I'm very pleased with the sky on the first image. The stars are not the problem. However down in the foreground, across the deep shadows - full size, it appears that someone dumped a salt shaker across the image. It's very obvious without getting anywhere near pixel peeping. The white dots in the deep dark shadows becomes instantly noticeable - especially in printing. I'm finding that the solution is more image and situation aware. In the first image, everything is very static, and it adapts to LENR, other than having double the time for each frame - especially when doing a stitched pano. Just taking a dark frame for subtraction in post is something I'm working on. So, now I'm taking some additional dark frames along the way.

The Pentax default tone curve is such that it yields good looking images most of the time. I've been pulling shadows on images, and at the same time realized that more details from the shadows did not necessarily make the images look better. The darker shadows are like blurred background or perspectives, they are components of an image that forces the eye to look at the main subject. The problem is "What is the subject?" , in a lot of photographs, we are confused about what the subject is. Because it's the subject that ideally should get proper light exposure relative to the remainder of the scene.

I agree. Having said that, there is a large difference between shooting with some light and shooting with no light. Also, with shooting night landscapes, you just go with what is available, and then work with it. There is no real adjusting or "lighting" the scene. With landscapes - in many situations, it's just the landscape. There is light painting, but other than specific items, the area is just too large - as in the second image.

With the second image, the landscape is somewhat active - trying to use the car's headlights driving up and down the canyon to provide the large scale lighting. This was my first time shooting this location at night - and it's VERY different than in the day. With the moving car lights, using LENR just removes any flexibility one has, so shooting periodic dark frames, I think is the only way to go in this situation.



---------- Post added 03-31-19 at 09:47 AM ----------

I agree with you a lot. In my view it's all about the presentation of the dark. How do you make an image stand out in its darkness? How do you make it presentable and likable? Our modern cameras have great HDR. I think it's all about post processing, and making the image "likable" to the masses.

When I was shooting or use my K5IIs, I find the foreground extremely easy and the sky more difficult. The additional sensor size on the K1 makes the sky much easier, while capturing the foreground a bit more challenging. Probably because the K1 offers more options and it becomes more of a question of determining just how to capture what little light that is available.

I much prefer your first shot with the windmill. For the second shot, maybe there is no white dot into it, but it's way too dark to appreciate anything, most of the frame is dark and the lit areas hardly show any subject.

The windmill is more of a "small corner of the world" landscape. I purposefully brightened it, I think a bit too much, as I would like the sky's background to be pulled down a bit more. But, this was a last minute test shot for a 21 frame stitch, that actually turned out a bit too good. Also, the interaction of the foreground elements (bushes and windmill) overlaying the sky, I'm finding is much more complex than what I anticipated. Especially with the astrotracing, which tends to enlarge the blurred branches that increases the complexity of compositing.

The windmill photo is great. How about going there at dusk, making an exposure for the bottom part of the frame at dusk and photographying the upper part of the frame (milkyway) a few hours later, then exposure fusing the two frames together?

I've considered doing that, however in this location to have the MW in this position, it's March and some April. Blue hour is 7pm and the MW rises and is in position at ~4am - so that's at least 8 hour and probably a 9 hour wait. With the interaction of the windmill into the MW, you can't touch the tripod at all.

Great thread and link OP. Totally agree with your progression of thought. The second shot feels real like one is standing there and can reach out and touch the Milky Way.
I think the art of the dark areas is to use the right amount of detail to keep the image in balance. The slashes of light across your foreground is all that is needed.
And the sky shouldn't be a contest of how many stars we can record -- Your first image simply isn't reality and I think when you go out and record our incredible Universe there is no need to pimp up reality.

I'm learning - slowly, much slower than what I anticipated. So many small variables to try to manage and get right. The analogy with painting is becoming very true here. I use to not worry about a lot of this previously. The more I do, the better I do, the more problems I see, that I want to fix.

The cliff faces across the valley (~5 miles away) are multi layered sandstone (white, tan, orange, yellow and reds) and reflect the light very well, but with headlights, I'm finding that the warm sunlight colors tend to be just off white areas that reflect. It's the random event of a car going along at the right time in the right area to light up the shadowed areas.

It's all a matter of taste, isn't it?

It really is. I'm an engineer with no art appreciation at all. Every time I shoot, it's starting all over from scratch, because even if the location is the same, the conditions are always a bit different.

I tend to like shadows that are lightened a bit. It does tend to be tough if most of the image has basically crushed blacks and is total darkness. At the same time, if you lighten the shadows too much, then it feels like some sort of alternate reality as we know at night we can't see things that well.

I'm struggling with the shadows - and hearing various views I find very helpful. Balance - finding the right balance to maintain the feeling of darkness, while trying to bring out a hint of the various details - is very difficult. The key is just enough to add some interest, and to break up the silhouetted blacks.

I've spent about 2 months with Camera One, ver 11 trial and then the ver 12 trial. It does a fantastic job with the white dots (drop dead easy), however in terms of color (and its strong suit is color management), just does not work with the Milky Way - I'm convinced. Every example I find of the MW processed suffers the same problem I'm having, so it's just not me. LR does such a better job of pulling up the colors that are already there, just not really shown straight out of the camera. The difference between the two packages I found to be pretty dramatic. My solution, I now think - is using RawTharapee to remove the white dots, and then dump a TIFF to do my normal processing in LR. I just need to set that up now.



---------- Post added 03-31-19 at 09:51 AM ----------

I tend to lift shadows for more detail and the increase blacks so the image doesn't look to have flat lighting but still has detail in the shadows.

I'm starting to learn that. It's a hard lesson to master.

It's very rare for an actual chiaroscuro painting to have crushed blacks, BTW. There are nearly always 'recovered shadows' in the background.

This is the first time I had heard of chiaroscuro, but it resonated with what I was doing. I do like recovering shadows, but I'm finding that it's all with the overall balance.

[biz-engineer]

In terms of a correct exposure, the question becomes what is a correct exposure of darkness? When you're out in the desert and unable to see your hand in front of your face, the question is 2, or 3 or 4 or 5 minutes of exposure for the dark foreground, and at what ISO 100, 200, 400 or 800.

That's pretty much a standard how much light comes from the cosmos when the sun and the moon are both behind the earth. I use 180 s. at ISO800 or 120 s. at ISO1600, at f2.8. There is a tradeoff between exposure time and ISO depending on how much long exposure noise vs read noise.

---------- Post added 31-03-19 at 20:16 ----------

Having said that, there is a large difference between shooting with some light and shooting with no light.

Have you tried stacking of shorter exposures / higher ISO ? I've seen nice night photos taken with Canon 5DIII, align/stacking a dozen of frames to cope with poor dynamic range. Stacking definitely reduces read noise, allow higher ISO with shorter exposures, less or absence of white dots in the shadows. And you could also use a faster lens (kinda 50mm f1.4 wide open, ISO400), use shorter exposure times and stitch the frames.

---------- Post added 31-03-19 at 20:23 ----------

Generally speaking , good photographs are obtained with enough light to reveal the subject matter. Decreasing the light levels goes against producing good images. How about adding some portable lighting to the windmill area before exposing for the milky-way?

---------- Post added 31-03-19 at 20:28 ----------

I've considered doing that, however in this location to have the MW in this position, it's March and some April. Blue hour is 7pm and the MW rises and is in position at ~4am - so that's at least 8 hour and probably a 9 hour wait.

Then you can expose for the milkyway at 4am , and prepare some coffee / breakfast while waiting for morning twilight before sunrise, then take a second shot before leaving.

[clackers]

This is the first time I had heard of chiaroscuro, but it resonated with what I was doing. I do like recovering shadows, but I'm finding that it's all with the overall balance.

Yeah, check out the works of Caravaggio, one of the pioneers of the style.