* * * * *
Experiments in HDR
I downloaded and installed some software [1] (trial version—it fully works
but leaves a visible watermark in the resulting image) so I could do some
experiments with HDR (High Dynamic Range) [2] images.
First image: from my desk (at Casa New Jersey) looking out the window. I took
a total of nine images from total overexposure (shown below) to total
underexposure, then averaged those images using the installed software.
[Overexposed shot] [3]
Overexposed shot, to get detail on the desk.
[Normal exposure] [4]
Normal exposure—desk is too dark and the scene outside is too light.
[Underexposed shot] [5]
Underexposed shot—you can see outside just fine, but inside is too dark to
make out any detail what so ever.
[Average of 9 photos, auto settings] [6]
The result of averaging the luminosity of the nine photos I took, using the
automatic settings. Inside is fine, but outside is ever-so-slightly washed
out.
[Average of 9 photos, mildly aggressive settings] [7]
This time, I used some more aggressive settings. The outside is a bit darker
(but still lighter than on the overexposed shot), but the inside is a bit
darker too.
[Average of most underexposed and overexposed photo] [8]
This time, I just averaged the two extreme photos and the result is a bit
better than the other two (and notice the missing watermark) although the
outside is a bit flat in this image.
The second experiment was taken completely outdoors. Again, I took nine shots
(of the neightbor's mailbox across the street) and took the average of them
(just the automatic settings) and the average of the two extreme photos.
[Outside overexposed shot] [9]
Outside overexposed shot. Note the total lack of detail in the sky (this was
taken around 8:00 pm Eastern by the way).
[Outside normal exposure] [10]
Outside normal exposure. You can now make out some detail in the sky, but the
foreground is dark.
[Outside underexposed shot] [11]
Outside underexposed shot. More detail in the sky (althought it's getting a
bit dark) and the foreground is nearly devoid of any detail whatsoever.
[Outside average of all 9 shots] [12]
The average of all nine outside shots. The foreground is a tad dark (compared
to the next shot) and the sky, while the correct color, is a bit blurred—or
rather, the clouds are a bit blurred, due to movement as I was taking the
nine shots. There's also a noticible bit of halo around the two trees.
[Outside average of the two extreme shots] [13]
The average of the two extreme shots. It's easier to see the foreground in
this one, but overall, the image is a bit flatter than the previous one, and
the color of the sky is a bit odd, although you can see great detail in the
clouds. The halos around the two trees are less visible in this shot as well.
These were very hasitly done experiments and I'm sure I can do better as I
play around with the software. And it's also nice that the software can do
this “dodge-n-burn” automagically.
[1]
http://www.hdrsoft.com/
[2]
gopher://gopher.conman.org/0Phlog:2006/06/25.1
[3]
gopher://gopher.conman.org/IPhlog:2006/06/26/overexpose.jpg
[4]
gopher://gopher.conman.org/IPhlog:2006/06/26/normalexpose.jpg
[5]
gopher://gopher.conman.org/IPhlog:2006/06/26/underexpose.jpg
[6]
gopher://gopher.conman.org/IPhlog:2006/06/26/ave-9-auto.jpg
[7]
gopher://gopher.conman.org/IPhlog:2006/06/26/ave-9-intense.jpg
[8]
gopher://gopher.conman.org/IPhlog:2006/06/26/ave-2.jpg
[9]
gopher://gopher.conman.org/IPhlog:2006/06/26/os-overexpose.jpg
[10]
gopher://gopher.conman.org/IPhlog:2006/06/26/os-normalexpose.jpg
[11]
gopher://gopher.conman.org/IPhlog:2006/06/26/os-underexpose.jpg
[12]
gopher://gopher.conman.org/IPhlog:2006/06/26/os-ave-9.jpg
[13]
gopher://gopher.conman.org/IPhlog:2006/06/26/os-ave-2.jpg
Email author at
[email protected]
