Perhaps I'm not understanding what is supposed to go on here. I thought the purpose of the HDR function was to use the differently-exposed images to create one with the best of each ...
When I choose a few versions of a file to Merge to HDR, PS seems mostly to add them (a la Screen Mode). IOW, shadows get enriched, but correctly-exposed parts get blown way out.
Yeah, you've been fogged by the "photographer's folly". HDR is a file format with extended range. Imagine a single staircase that stretches from the 1st floor to the 2nd floor. In 8-bit (bpp) there are 16,777,216 steps (256^3), in 16-bit (bpp) there are 65,536^3 steps but the staircase (intensity value range) still only stretches from the 1st floor to the 2nd floor. With HDR you have 32-bit (or the much more rare "16-bit half") but you're actually extending the range beyond the 2nd floor - way beyond. As our monitors can not display all of this range in a single image you're usually presented with an exposure slider or an interactive mouse-over that allows you to view a segment of the full range as an exposure level. This should look just like a normal photo and not have the various tonal ranges remapped into a single image. The "photographer's folly" comes from misunderstanding this and confusing HDR with two techniques called "tone mapping" and "exposure blending" where such often require the construction of an interim HDR image (HDRI).
Here's a tone mapped 8bpp image (jpg) followed by a normal jpg (the middle exposure of 3), followed by what a real HDR looks like displayed on a typical monitor (notice the exposure slider that appears in the HDR window and also how the HDR at that exposure looks just like the middle exposure JPG): http://tesselator.gpmod.com/Images/_Equipment_n_Tutorials/Compare_Three.jpg
The other mistake commonly made stems from the belief that one NEEDS multiple exposures in order to achieve this tone-mapping effect. Multiple exposures can be beneficial if there is too much clipping going no or if the tonal range you're trying to compress is too great, but it's usually not necessary if you exposed the image correctly in the first place. Here's tone-mapping achieved from a single camera jpeg taken with a Minolta A2 that shows some of the intensity ranges we normally think are not available from a jpg - let alone a 12 , 14, or 15 bit camera RAW file: http://tesselator.gpmod.com/Images/_Equipment_n_Tutorials/JPEG_Lattitude.jpg
Here's more on the PS specific "Merge To HDR" function that covers most of this in maximum verbosity. http://www.luminous-landscape.com/tutorials/hdr.shtml
Brooke,
Bifurcator is correct. It seems that many people are trying HDR whan a better understanding of PS would be sufficient. However it can be very effective for some images.
Still - It sounds like you are not converting the 32 bit HDR file to a 16 bit version. Nothing should be blown out if you are completing the entire process. I do not usually use PS for HDR, but when you convert the 32 bit file it gives you a variety of choices to do the conversion.
Well PS is probably displaying a default exposure "section" that shows some of the highlights as blown out - like the flashlight lens in my example image. Just move the slider in the image window to display the HDR without the blown out areas. I notice often that when PS loads or assembles an HDRI it's default (initially presented) exposure is too bright and not centered to what I would expect it to be given the exposures used. HDRShop (the original HDR tool) displays it initially more to my expectations. But it's just a slider move to change that (in PS).
When you adjust that slider you're modifying the defaults PS (and other apps) will choose for an initial display of the image - it's saved with the image. Converting to LFR (Low Fixed Range) in PS will additionally display some sliders for tuning the exposure. This is not tone-mapping or exposure blending though and doesn't attempt to compress (or "remap") any of the tonal ranges present in the HDRI.
I've seen that problem before. I forget what it is and how to fix it. It doesn't happen to me. When i Merge_To_HDR in PS using those two images I get this:
Just to clarify your process a bit: Are you using two differently developed versions of the same original image (2 differently developed TIFFs from the same raw original, for example) or are you using two different original images?
Eyeball wrote:
Are you using two differently developed versions of the same original image (2 differently developed TIFFs from the same raw original, for example) or are you using two different original images?
Two differently developed TIFFs from the same RAW original.
That might be part of the problem. Photoshop doesn't really like it when you try to "cheat" like that. In fact, Photoshop yells at me when I try to do that. Do you not get that message?
Besides that, I think the general consensus is that that kind of technique doesn't really buy you a whole lot. You are going to a lot of trouble to essentially do tone-mapping, as Bifurcator mentioned, but you are not getting the advantage of the additional tones that starting with differently exposed originals would get you.
There is nothing wrong trying to do what you're doing but it seems to me that it can be accomplished more simply by using a combination of Camera Raw and the Shadow/Highlights control in PS.
I'm telling myself that I remember that being the cause of your problem. Using the same image like that makes it go all poster-art and stuff. But I dunno if I'm remembering right or not.