I thought this was an interesting comparison, though by no means totally decisive as to which High Dynamic Range application one should purchase and use. I ran the same three exposures through three different HDR programs plus Photoshop CS3’s Merge to HDR function. I used the default settings on all of these apps, so as to keep the comparison as close to “apples-to-apples” as possible. Obviously, changing the parameters in the various programs could/would/should improve the output. Nonetheless, the default results are telling, though the story may be up for interpretation.
My original RAW exposures are at EV 0, EV+2 & EV-2. These then where processed accordingly:
– Photoshop CS3’s Merge to HDR function: one HDR output
– Unified Color’s HDR PhotoStudio 2: one HDR output
– Essential HDR, version 1: two outputs, one using the Details Revealed mode and the other using the Fast Tone Balancer mode
– Photomatix 3: three outputs using Tone Compressor, Details Enhancer and the non-HDR blending mode of Exposure Fusion
Looking at the original 3 exposures and then the Photomatix HDR outputs (all three!) it seems pretty clear to me that out of the proverbial starting gate Photomatix better positions your images. They have richer colors, more details, believable contrasts and less strange “HDR-like” artifacts. This is not to say that the other apps presented useless images (well, the Photoshop output was just plain terrible), but all of them would require more work in Photoshop to bring them into the same league with the Photomatix images.
Mind you: This was just one test. Other images with different contrast ratios and color schemes should be tried with the other applications (though not with Photoshop’s Merge to HDR…I’m done with that!) before one could definitively say who the winner would be of this HDR shootout. I’ll continue to explore the other apps and perhaps new ones as well. But I must say, for my HDR work, currently I’ll just be clicking on the Photomatix icon.
I don’t get why you have to include the HDR PhotoStudio picture here as a "default" output, except for maybe to thrash the program (though it is is fact quite powerful and can generate a very nice naturally looking picture for your example) or to show your complete misunderstanding of the program’s nature and its capabilities. At the same time you choose to include non-default (processed) outputs from other apps — what kind of fair comparison is it?
Thank you for your comment, Igor. I, in fact, had no such goal to thrash any program. I do think PhotoStudio appears to be powerful…but my point was not to thoroughly test any of the apps, but instead to put the same 3 images through a few HDR apps utilizing what each developer calls "default." That is what I did. None of the other outputs have been optimized, despite what you may believe. I state quite clearly "Obviously, changing the parameters in the various programs could/would/should improve the output" and "Other images with different contrast ratios and color schemes should be tried with the other applications" so that should have left no doubt that image improvements are possible and further testing is desired. I’m sorry if that wasn’t as clear as I meant it to be.
Hi Jeff, thanks for replying. The problem here is that you include a non-processed, the actual HDR look of the picture, which is rendered using a non-HDR output device (display). Therefore it’s not going to look good, especially considering a good dynamic range that you had in the original (you see that the highlights are clearly overblown not getting into the output range). By no mean this look could be called the "default" output. In fact, there’s no such thing as a default output in HDR PhotoStudio, because right after the merge it shows the image in its actual (non-changed) HDR form. This is done in order to preserve the original data and let you examine it (using the Display Brightness slider) *before* you begin to do anything with this data. To make this comparison fair, for example with Photomatix you would have to just show the Photomatix image right after the merge, which as you know will look extremely terrible.
Thanks for the clarification, Igor. I think what threw me was the "two-step" paradigm of both Photomatix and HDR Essentials whereby you file merge then run tonemapping using some default settings the developer has pre-loaded (and — in the case of Photomatix — then further tweak the output via the sliders, if desired). Correct me if I am wrong, but your product fore-gos that second step and instead the tonemapping — or dynamic range mapping as you call it — is performed dynamically via the various controls. If my understanding is now correct, then you are correct and I have mis-stated what that initial HDR PhotoStudio output represented. To be fair, I will post a clarification as well a dynamically-mapped output of the same 3 exposures using HDR PhotoStudio.