I carefully read all 3 responses below, and I think the difference in the definition of what different people mean by HDR.
In my view, HDR provides mapping of the large-bit-space (due to the large dynamic range of the scene light) to a smaller-bit-space. What the mapping function, whether it is analytical or not, spatialy uniform or not (i.e. global or local), etc, is a separate question. When a high-contrast scene is photographed with different exposure times, and then exposure fusion is performed (manually or with a software), what you do, is you take subsets (sub-ranges) of that original large-bit-space (large dynamic range), and capture them separately. Then you combine these sub-ranges into one photo, by some non-analytical, spatially non-uniform mapping function - to a smaller bit-space (dynamic range). The difference from the process performed by the software that has "HDR" in its title is just that has a well defined, frequently power-law (or more complicated function) type of scaling (dynamic range compression). I should point out that local burning and dodging done in photo-printing is also tone mapping to the paper which is typically a relatively low-dynamic-range media. As a simple reference that describes what I wrote above in somewhat more detail, I'd offer this Wikipedia page: http://en.wikipedia.org/wiki/Tone_mapping Best regards, Igor Wed Sep 4 17:00:38 EDT 2013 Mark Roberts wrote: > Igor Roshchin wrote: > > >I have a question for John Sessoms: > >John, you wrote for your photo: > >"This is not an HDR image. It is a blend of 4 exposures 2 sec to 45 sec > >using layer masks & blend modes." > >How is it different from HDR? > > HDR combines images into 32-bit-per-color space and then tone maps the > 32-bit image into a viewable 16 or 8-bit image. Wed Sep 4 14:35:16 EDT 2013 Matthew Hunt wrote: > I would say that the difference is that in "HDR imaging" (as the term > is generally used) at some point (prior to tone-mapping) you produce a > scene-referred high-dynamic-range image. This requires understanding > or measuring the quantitative exposure relationship between the input > LDR images, allowing the software to know (for example) that a bright > point in the HDR image is 1,000,000 times brighter in the original > scene than a darker point. > > What John did is more akin to exposure fusion, as performed by the > Enfuse software for example (although it sounds like John blended > images manually). In exposure fusion, the pixels in the output image > are a weighted average of the pixels in the input image (with weights > varying from pixel to pixel), but exposure fusion algorithms don't > need to know the quantitative relationships between the exposures... > they can just prefer to weight pixels that are "well exposed" (i.e. > near the middle of the LDR exposure range). Wed Sep 4 16:52:16 EDT 2013 John wrote: > It's not HDR because I didn't use HDR & it's not tone-mapped. There's > nothing wrong with HDR per se, I just didn't use it. > > Mainly, I didn't use HDR because I couldn't get it to do what I wanted > it to do in this image, i.e. not look like it was HDR. > > It does still look kind of HDRish, and to the extent it does, I'm not > happy with it. -- PDML Pentax-Discuss Mail List PDML@pdml.net http://pdml.net/mailman/listinfo/pdml_pdml.net to UNSUBSCRIBE from the PDML, please visit the link directly above and follow the directions.
