wrote:
I haven’t used HDR (CS-1 user) but have done it by hand (Fuji-S2 Pro, locked f-stop with floating shutter speed, Bogen 3021 tripod, cable release, raw to TIFF).
My thought is that a 2 EV difference is too much. When I hand do it, I use 0.33 or 0.5 step intervals depending on the scene and free space on the CF cards (normally carry 4 gig available when I go out).
I suggest you try shooting a test scene at 2, 1, and 0.5 EV steps and see if there is an improvement as the difference between pictures narrows.
I’ve seen some remarkable work others have done with HDR and will be tring it out on the CS-2 trial soon.
Bill
The following is an answer that gave to another current message in this newsgroup on Jan 6 07:
———————
You have made a few conceptual mistakes about HDR that will usually screw up HDR photography. The first is using Photoshop’s Merge to HDR. It isn’t very good or very easy to learn with. You will do much better if you spend the money on Photomatix Pro and learn on it. It does much more too.
A High Dynamic Range picture is one that has values outside the capabilities of most digital hardware and software. It takes 32 bit pictures to be able to save these. i.e. They have highlights and shadows that your camera, monitor, printer or 16 bit files will never be able to represent. For most practical uses, these are useless the way they are; you can’t display them anywhere.
The reason for HDR software is to squeeze the dynamic range of that scene down into a 16 or 8 bit format that you can actually use. The two most common ways to do that are to blend a bunch of exposures together to get averages of all of them OR to use Tone Mapping to figure out the best parts to use to make into values inside the range you want. OK, that is very simplified, but hopefully you get the idea.
The number one problem that most HDR beginners have with HDR photography is that they are trying to work with scenes that don’t need HDR. They take pictures of scenes that can fully or successfully be captured in their cameras in one exposure. Then they run them through HDR to flatten them out into very flat and very ugly pictures.
There are fewer scenes out there that need HDR than you think. A dark, dense forest with a bright shaft of sunlight coming down inside would be one. The sun stabbing through the dark clouds of a heavy storm would be another. Interior shots where you need to properly expose the lights inside, the shadows inside, and the view outside the bright daylight windows would be another. (Inside a church works for this.) Daylight inside a canyon with your view including outside the canyon is almost always an HDR scene.
I very much doubt that a typical daylight coastal scene would need HDR at all. You may have bright sun reflecting off water, but you probably don’t have any large sections of dark shadows that need detail in them. You probably really don’t need any detail off the specular reflection from the sun either. So, don’t try to capture detail where you don’t need it.
The next biggest mistake to not bracket widely enough. A one stop bracketing is almost never enough. Your camera’s natural dynamic range is much wider than that. You need to bracket enough to properly expose everything in the scene. An exposure that is at the meter’s setting will probably capture 4 to 6 stops of the scene on most digital cameras. One stop of bracketing will be well inside this range and a waste of your time at the camera and at the computer. Start experimenting with 2 to 3 stops apart for bracketing.
The next most common mistake is using the HDR software to squeeze the dynamic range down too much. Remember what makes most pictures good – you want as much dynamic range as possible. (There are plenty of exceptions to this, of course.) You don’t want the picture as flat as possible. You want the picture squeezed down just to the point that the highlights and shadows just fit inside the range that your equipment can handle. sRGB at 8 bit is a good standard to work from as most equipment can work well with that in a color managed system. Still that leaves a pretty white highlight and a pretty dark shadow.
If you do the above correctly, you have the principles in line for figuring out the software. You WILL have to do a lot of experimenting with the software. None of it is intuitive and there is precious little written in the form of tutorials that is useful. Frankly, I have never got a good output from Photoshop’s HDR. I don’t know anyone who has. That’s why I and many people have bought Photomatix Pro. It still takes a lot of experimenting, but it’s a bit more logical – after awhile. It also has more flexibility and tells you more useful information throughout the process.
A few other notes I’ve learned from doing way too much HDR photography…
* Blending is more "realistic" than Tone Mapping, but tends to make flat, dull interior pictures. Blending has worked best for me for exterior pictures, but not always.
* Tone Mapping has usually worked best for me on interior photos.
* Tone Mapping can shift colors a bit though. This is particularly true if you have different colored lighting throughout the scene. (Which you probably will if there is daylight outside the windows.)
* Use Tone Mapping as lightly as possible or many scenes get a very surreal look to them. (Well, unless you like that look.)
* Expose 2 or 3 stops apart from the darkest part of the scene to the lightest. For me that was usually 4 to 7 exposures. Judge this from the screen and not the viewfinder or histogram.
* Shoot in manual mode and change the speed; so you don’t affect the DoF and focus.
* Use a good tripod.
If you need HDR and you’ve spent the time to learn it, it is a great tool. Have fun!
Clyde
