..needed to search a bit, but i mean that one: http://graphics.stanford.edu/papers/hdrp/hasinoff-hdrplus-sigasia16.pdf
because it could run as a preprocess which would make it easier to suffer through the couple of seconds of processing time that it'd probably need. -jo On Thu, Apr 27, 2017 at 2:54 PM, johannes hanika <hana...@gmail.com> wrote: > great, we've got a flamewar! let me join the fun: > > i'm very much unconvinced by their examples re: lens blur or generic > sharpening. it has the typical fourier artifacts (even though the > ringing seems surprisingly well balanced in their examples. but i can > still see it). i think our current local contrast tools would do a > similar job. > > the motion compensation looks nice though. i'd probably rather > implement the multi-frame merging technique that morphs multiple short > exposures to match up and then creates a hdr from that. > > cheers, > jo > > On Thu, Apr 27, 2017 at 2:24 PM, Moritz Mœller (The Ritz) > <virtualr...@gmail.com> wrote: >> On April 27, 2017 2:03:47 PM Roman Lebedev <lebedev...@gmail.com> wrote: >> >>> Are there any actual use-cases for this algorithm, specifically in >>> darktable? >>> >> >> Did you bother opening the webpage (relaunch.piccureplus.com) the OP gave >> and looking at their example section? >> >> Do you think we don't have landscape, wildlife, insect, etc., etc. >> photographers using DT? >> >> Or just ordinary people who have that odd holiday shot with great >> composition, light and everyone smiling but being oh-so-slightly out of >> focus? >> >>> I'd imagine 95+% of blurred images/images with motion blur >>> where that is not intentional, would be deleted and not processed.. >> >> >> And where did that number 95%+ come from? Some bodily on orifice? I refrain >> from making an educated guess here from which -- to avoid foul language. ;) >> Just because you can't imagine it, Roman, doesn't mean it doesn't exist. :) >> >> To give you an example, to the contrary, based on actual numbers: >> >> If you do low/available light (no flash) photography of people that move >> (wedding photography, social dance photography, parties, events etc.) you >> will usually have a ton of images that could be great if they only could >> have a bit of blur (motion and/or focus) removed. Because even with fast >> glass and high ISOs you will need exposure times that are too slow for the >> motion depicted. >> >> In my case, I photograph tango and swing dance events. >> I shoot with a f/0.95 lens on an A7II body. There are no lenses of this >> speed that have autofocus. And even if they were, the focussing would be too >> slow. By the time the camera had hunted the focus down, the moment were >> gone. >> >> So I need to track focus and motion (of dancers) manually, at the same time. >> Usually 10% of the images have either sharp focus or no motion blur (both >> absent are less than 1%!). >> >> Aka: at /least/ nine out of ten of these images would benefit from this sort >> of deconvolution magic. >> >> Beers, >> >> .mm >> >> >> >> >> ___________________________________________________________________________ >> darktable developer mailing list >> to unsubscribe send a mail to darktable-dev+unsubscr...@lists.darktable.org >> ___________________________________________________________________________ darktable developer mailing list to unsubscribe send a mail to darktable-dev+unsubscr...@lists.darktable.org
