mikegreenwald wrote:
Facing a subject such as flowing water, wildlife in motion, sports, etc the lighting may vary from very dark to very bright. Normal HDR can't be used because of subject motion. My practice is to underexpose by 1 EV to minimize washout of the bright. Then, in PP I create additional images at plus & minus 2 EV from the original. Then use the three images in what's called pseudo HDR, to create a final image. Usually it works, but sometimes the range of light is just too great to get both shadow and highlight detail.
Is there a better way?
Facing a subject such as flowing water, wildlife i... (
show quote)
I frequently do something similar...
I'd suggest you
not deliberately underexpose the image. That will tend to increase digital noise and is usually unnecessary. Take the image "properly" exposed and use that as your "middle" or main image file... then create one at +1/2 stop... +1 stop at most... and another at - 2 or -3. You might only need one or the other.... depending upon what you're trying to accomplish (i.e., whether you are trying to darken or brighten the background). You can "pull" exposures back down in post-processing a whole lot better than you can "push" it up from underexposure. In other words, in digital images slightly blown-out highlights are generally more "recoverable" than too-dark shadows.
For a more "normal", less HDR-looking image, use narrower range of difference between only two or three shots. For a more exaggerated HDR effect, use wider range of difference and possibly more shots (4, 5 or more).
Here are a couple examples, both of which I was trying to avoid "HDR exaggeration". (Both finished images are slightly over-saturated due to the printing process that was going to be used.)
This first example is done with single shot that was shot using best possible normal exposure for the primary subject, then double-processed: one for the shaded indoor subject, the second about - 2 or -3 stops to recover some of the very overexposed, sunlit, outdoor background. Exposure, color temperature and tint were all adjusted differently, then portions of each image were manually combined:
+ = Above, left image is the normal exposure for the subject.... center version is -2 or -3 stops to recover background... and the far right image is the result manually combining the two. Here's another example that's more subtle... an outdoor and brightly backlit subject, normally exposed (but exceeding the dynamic range of the camera). Here I wanted to both recover some highlights and to open up the shadows a bit. For this I started out the same way with a single image, then made three versions... one about +1 stop, one without any adjustment to exposure and the third -1 stop. Not a wide range of difference, because my goal was a "normal, unexaggerated" look. There also was no need to tweak the color temp and tint in this case, since the entire scene is sunlit. And here I automatically combined using Adobe Photoshop Photomerge HDR (stripping the three images together manually would be near impossible!):
+ + = Far left image is +1 stop in post-processing, next version of image is 0 adjustment, third shot is -1 stop in PP... and far right image is the final "HDR" composite of the three. I know with this second example it's difficult to see the difference at Internet sizes and resolutions, plus even the best graphic quality backlit LCD computer screen "clips" both highlight and shadow detail... The end result is intended to be rather subtle improvement over the original image and it's more noticeable when prints are compared.
Note: While I refer to "stops" above, I'm making my adjustments in Lightroom using its "virtual copy" feature, very convenient for this sort of thing. Lightroom adjustments aren't indicated in actual stops, so I'm estimating.
So, basically I'm doing same as you. Yes, it might be possible in some cases to take two shots, as previously suggested... one with the moving subject and a second of just the background.
But I think the main mistake people make is "fear of blowing out highlights" that causes them to underexpose, which makes for a lot more noise in their images.
There is probably some variation from camera to camera, but I've found highlights in digital images are a lot more recoverable than most people think. The top example above, the outdoor light was probably 4 or 5 stops brighter than the indoor that I was exposing for... so the outdoor sunlit background ended up extremely overexposed.... yet quite a bit of it was recoverable. Also, often you can't fully see the highlight and shadow details on a computer screen, but will see them both in any prints made from the image. In other words the dynamic range of a backlit LCD computer monitor screen is less than that of the print.
EDIT: a later post reminded me of another important thing.... Yes, I also shoot RAW almost exclusively. RAW files have much more latitude for adjustments such as shown above, and both those were originally shot RAW and processed in 16 bit mode, then reduced to 8-bit JPEGs as a final step. (Usually I shoot only RAW, sometimes RAW + JPEG if images are needed immediately, there's no time for post-processing... Virtually never shoot JPEG only.)