HDR Merge in ACR

Contrail Above Farm

Calm Hazy Day, Seneca Lake

A couple more new shots. I hope my new camera is not overstaying its welcome. If you look at the second one on Flickr, you can see some nice detail in the boat.

This last one is just a test shot. It was blazingly bright, with the sun glaring off the walk and the shadows all but black to the eye. I took three quick frames, one at metered exposure, one greatly overexposed, and the last greatly underexposed, and did an HDR merge in Adobe Camera Raw. It's very easy to do—you just open all three pictures at once, right click on any one of them and select "HDR Merge." The interesting thing about this is that I made all three pictures handheld rather than on the tripod. The dialogue box has a box to tick that tells the program to automatically align the images. I did get a few artifacts in this, and, also, when I took the metered exposure and simply processed that, it came pretty close to this result. So the jury's still out on whether that's a technique I'll use occasionally (when called for) or just ignore it and make do with one exposure in high-contrast situations. We'll see. It's at least another option as a backup for a picture I really want.

Mike

Featured Comments:

Michael Matthews: "Is it a lens effect—or has the pool shed developed a list to the right?"

Mike replies: 'Tis a little wopplejawed, yes. (Wopplejawed: Hoosier variant of whomperjawed, hillbilly/Southern for "askew.")

John McMillan: "My very Southern Ma loved the word, but pronounced it 'whoppyjawed.'"

Jim: "For years I shot landscapes only with 8x10" or 4x5" view cameras. Often I would encounter a scene where the range of brightness was too great to reveal texture in both the highlight and shadow areas. Then I would make a pre-exposure through a piece of frosted plexiglass, followed by a regular exposure. Here is one example, taken in bright morning light with deep shadows."

Mike replies: As you know, Jim, but others might not, that's a technique called pre-exposure, meant to counteract the fact that B&W film has an exposure threshold—that is, it has to receive a small amount of image-forming light before it starts forming a latent image. Pre-exposure is meant to get the sheet of film into the reciprocity range so that shadows begin building immediately, allowing you, in theory, to give less exposure, and thus restrain the highlights, for contrast control. (And dappled sunlight in the woods is indeed surprisingly contrasty.) Pre-exposure is described by Ansel Adams in his photography series. I once built a pre-exposure box out of plexi—a black cube, open at opposing ends, with a sheet of translucent white in the middle. You put this over the lens and gave the film a predetermined small shot of non-image-forming light prior to taking the picture.

Personally, when comparing pre-exposed negatives with the same scene that was not pre-exposed, I couldn't reliably tell the difference, so I just dispensed with using the method. I imagined that maybe the film in Ansel's day behaved differently, and the corrective effect was more pronounced. Although it's possible my technique was simply not rigorous enough to achieve the intended effects, or I screwed up somewhere in the implementation, who knows. 

Rob Campbell: "Something that I think I notice: in the last picture, there seems as if something is missing in the tonality of the buildings’ walls. I can’t put my finger on what it is, exactly, more a vague sense of missing contrast than just due to the quality of light on the planks? I have sometimes found this very same thing on some of my own pictures, too. I don’t remember it from my film days. Any ideas?"

Mike replies: Probably just that we're looking at an 800-pixel-wide JPEG in which a whole lot of information has to be thrown away—including, perhaps, the texture cues for the walls. It's also possible that certain tones get compressed too much in the process of adjusting the sliders, with some "micro ranges" of tones becoming a bit too flat or too harsh. Another way of saying this is that the curve becomes discontinuous, with jagged peaks and dropouts. But I don't know for sure. Where I see it most is in the fact that digital just doesn't have enough information in the highlights. It's one of the main reasons (IMO) for using a high-resolution, high-megapixel camera—the highlights can have more information in them, and thus can look more natural.

Eric Brody: "Whether the merge to HDR works well seems to depend a lot on the scene. I often shoot five-shot HDR's at 1–2 stop intervals and combine them. When I look later at the original images, often one of them is quite adequate, so I toss the others along with the HDR. With the dynamic range of today's cameras—mine is pretty good—HDR is often not needed. When it's noticeable, it's UGLY!"

Comments

[kirk]

I never tire of well written posts about the Sigma fp. Fun to read and also one of the few interesting cameras on the market right now. Double win.

[Alex Mercado]

“I hope my new camera is not overstaying its welcome.”

Quite the contrary, I find your new camera, and how you’re working with it, to be rather fascinating.

[Christopher DellaCorte]

Mike:

I am glad you shared this off-the-cuff HDR method with us all. I use it frequently when out and about in the glare and do not want to think about whether I have enough dynamic range to "make adjustments to a single file" in post.

BTW, if you have LightRoom Classic, the photomerge HDR features found in photoshop are built in.

CDC

[Mark Sampson]

Nice work, Mike. It loos like that your new camera is working out for you. Will you be making prints? For me at least, that would be the real test.

[james wilson]

When I was using 4x5 chrome for interior photos I would often 'flash' the film with a low exposure through a diffusion filter to lift the value of shadows or dark materials. Once I put too much diffusion exposure on the shot and it came out very flat. I scanned it and hit auto contrast in PS and it was perfect.

[Bandbox]

"Pre-exposure" takes me back to darkroom days. But instead of using it when exposing the negative in the camera, we would use it for printing very contrasty negatives in the darkroom. It was called "flashing" and was done by giving the printing paper a short exposure (the threshold exposure time determined by test strips) under the enlarger and then placing the negative in the carrier and exposing the print on the flashed paper. It was useful for bringing down extreme highlights to a manageable level for burning in.

Nowdays it's just a few camera settings, some sliders in Lightroom and Photoshop, and maybe a touch of HDR. How I love digital!

[Curmudgeon]

I'm loving the black & white images. AA is one of my favorite photographers.

[Herman Krieger]

For me, the content of a photo is more important than the tonality. For which I'm satisfied with the sort of corrections gotten by dodging and burning in.

[James]

Mike, have you tried the same scene using more steps for your HDR?
The finished results are sometimes better using five or seven exposures.

[James]

Mike, have you tried the same scene using more steps for your HDR?
The finished results are sometimes better using five or seven exposures.

[Aaron]

I don’t think it’s overstaying its welcome. It’s like a photography version of a dramedy—you and the new camera falling in love. Who doesn’t like such a story?

[Moose]

"Where I see it most is in the fact that digital just doesn't have enough information in the highlights. It's one of the main reasons (IMO) for using a high-resolution, high-megapixel camera—the highlights can have more information in them, and thus can look more natural. "

This is untrue at the sensor level; in fact, the reverse is true. A pure Raw file looks very dark, with the top stop occupying over half the histogram.

In the process of raw conversion, application of Gamma correction creates something than looks normal to our eyes by stretching the bottom and compressing the top. Converters such as ACR/LR also apply profiles than may increase or decrease highlight compression.

Conversions to an 8 bit format, such as JPEG, further compresses highlights and loses some tonal data to compression. Reread Ctein's posts about ISO and curves in digital files. Choice of ISO for the reasons he explains and curves with relatively horizontal tops may also lead to highlight compression.

I have for years had my digicams set to -0.7 EV as standard, because I value highlights. Among many other things, I regularly shoot things like white, yellow and red flowers in the sun. Retaining the subtle tonal details of surface texture, etc. is important to me - and possible.

The data is there, if you expose correctly, and process correctly. The fault lies not in the number of pixels.

[Moose]

My experience is like Eric Brody's. I often bracket exposure in high contrast situations, and very seldom use more than one. As he says, HDR turns UGLY really easily. Even when not obviously bad, the rearrangement of tonal values tends to look unnatural.

When I do use more than one exposure, I'm much more likely to create layer masks and adjust brightness/contrast separately for different areas than apply overall HDR

[Moose]

HDR and Whatever has been done to these photos, they seem to my eye to have flat highlights:

[Earl Dunbar]

The 2nd one really personifies the Finger Lakes for me. Love it.

[Bandbox]

Good road shot.