I've started to try to build my own profile for my 16-50mm Sony lens and running into a few minor problems. The ALPC documentation recommends making a wide range of picture sets to build a profile, including combinations of Focal Length, Aperture and Focus distance.
Now as far as I know, nobody knows how to extract focus distance information from Sony EXIF, not even Phil Harvey, the Mighty Master of EXIF and author of EXIFTOOL.exe. So, if I make a profile for my 16-50 including varioussets for various focus distances, how on earth is the Adobe Lens Profile Creator going to know how to separate photos shot with same focal length/aperture but with different focus distance? And if it *can* do this, how will Lightroom know which focus distance group to use for any particular picture?
Does Adobe know the secret of extracting focus distance from Sony EXIF or would I be better off just shooting one picture set for each FL/Aperture combination and not worry about focus distance sets?
If I shoot just one focus distance for each FL/Aperture combination, I suppose I should use a distance typical to my style of shooting.
This brings me onto my other problem - my typical shooting distance for the 16-50 lens is indoors.but sometimes outdoors. Indoors, I can just about cobble together a big enough calibration grid for 16mm at 3m shooting distance but outdoors, I'd need billboard sized charts! How have people overcome this or do lenses not really change their barrel/pincushion distortion profiles very much as focus distance varies. It would be great and convenient to profile a lens up close to an A4 sized calibration chart but I'm not convinced that will give a very good profile unless all my shooting is also of close up objects.
How do people deal with calibrating wide lenses?
Thanks for your help here!
I cannot answer the question on the Sony EXIF focus distance question. I am not the expert in that area. Even without EXIF support, Adobe Lens Profile Creator will be able to estimate the distance of the camera based on the 3D geometry.
For wide angle lenses, you don't need a super big charts. You can cover more complete field of view by taking more overlapping shots. Also because of the extreme distortion of wide angle lens, print a chart with a bigger checker square sizes. It will help the grid detection.
Knowing the target size, show allow the LPC to calculate the distance to the target, so the profiles, themselves, could know what the focus-distance was. That does not mean that Adobe knows the distance from any arbitrary photo, though, so the correction of the distortion in LR or ACR may not be able to use the information in the lens profile.
As of LR3.2 Subject Distance is not a visible metadata field, but the following forum message describes how to put it back in to LR 3, with the idea that if LR can show the Subject Distance of an arbitrary photo with your lens, then it can use it to determine which lens-profile segment most closely matches: