I've been experimenting with HDR processing and I am alarmed at the differences I am seeing when the same two images are processed in two different ways.
1. In Bridge CS4 I open the two images in ACR 5.3 and, without making any adjustments (normally I would tweak the shadow in the dark one and the highlight in the light one, but bear with me.) I choose Select All and (holding down the Shift Key - I'm PC) click on Open Objects.
I now drag the dark one into the light one and I've got an image I can work with.
2. Alternatively, in Bridge CS4 I highlight the two images in the folder and choose Tools/Photoshop/Merge to HDR....
To say that the merged image looks awful (I accept this may be specific to the way PS can handle these specific images) in the Merge to HDR window that opens is an understatement. I get very weird streaking in certain parts of the merged image - it looks like oil on water. Adjusting the Set White Point Preview makes no difference. The coloured streaking is still there.
Anyway, I click OK and the file opens in PS. It looks flat and dull in a way that the approach in 1. above does not produce.
Oddly, even though the streaking looks purple(ish), the info reading for it is very close to neutral - 27, 27, 28.
One last thing, the Histograms for both methods could not be more different. 1. is normal. 2. is hugely biased towards the shadow area.
Do others use 2.?
Any recommendations on good tutorials?
Despite what I've said, the ability to get sub-255 readings from a light shade in a room while getting the shadows under a sofa in the same shot reading 20-30 is simply fantastic! ;-)
D.
