gdanmitchell
Offline
• • • • • • •
Upload & Sell: Off
|
(moderators may want to move this to another forum, and I'll gladly continue the conversation there.)
There are actually at least three options, one of which you did not mention—though the third one is easier to use in Photoshop than in Lightroom. (I'm a Photoshop guy, so someone else will have to explain how the "other" method could work in Lightroom.)
Between the two you mention:
Gradiated/gradient/graduated neutral density (GND) filters: In film photography this was pretty much the only effective way to deal with extremely wide dynamic range scenes in which a single exposure would leave you with blown highlights, blocked shadows, or both. It works well for subjects in which the division between light and dark areas is relatively linear—such as the horizon—and less well for subjects where that is not the case. In the latter there are some complicated ways to make things better by using more than one GND and/or by manually moving the filter during exposure.
Gradient filters in Lightroom, ACR, and other software: If the dynamic range of the original scene was extremely large, while you might make things better with this software adjustment, you are still working with the data of the actual capture, and truly blown out digital capture of highlights is not really recoverable in post and you can run into noise and banding issues if you try to use the software to recover (or create?) detail from areas that are all black or nearly black. On the other hand, the gradient filter can work very, very well for less extreme exposure balance issues and is really no different that other masking-related techniques (dodging/burning, curves adjustments, etc.) that are almost always applied in post.
Exposure blending is the technique that you did not mention. For scenes with extreme dynamic contrasts I prefer (YMMV) this approach over the other two in almost all of my photography. The idea is to capture more than one exposure of the scene, varying the shutter speed (not the aperture!) to get a range of exposures that are optimized for different areas of the subject. In the simplest form you might make one exposure optimized for shadow detail (and allow the highlights to blow out) and a second optimized for highlight details (allowing the shadow areas to go all the way to black). Then you align/layer the images in your post-production software and take the best portions of each. Most often this is done by creating masked adjustment layers. I won't go into all the details here, but this can be very effective. (By the way, this is not the same process as what is often called HDR.) The following is a photograph made this way, from three exposures that were exposure-bracketed and focus bracketed.
http://gdanmitchell.com/gallery/d/3300-2/SabrinaBasinLupinePeaks20090807.jpg
Advantages of this approach include:
1. No need to carry additional filters in the field.
2. The exposure difference between the light and dark areas can be anything, rather than being limited to the stop difference—.5 stops, 5 stops, etc
3. The boundaries between areas of various brightness are completely fluid rather than being limited to straight lines.
4. Partial blends of both layers are possible in any area of the image by means of controlling opacity.
5. You can reconsider the blending at any time in post.
6. etc.
Issues that you have to deal with—and there are effective ways to do so—including things like possible motion of the subject between exposures, the need to keep the camera position the same, and so on.
HDR: This is also, of course, the option of using the multiple exposure approach in conjunction with software-based methods of constructing a composite image based on algorithms that analyze the image and attempt to combine the light parts into a whole.
Dan
Edited on Apr 10, 2014 at 11:23 AM · View previous versions
|
