Merge to Panorama Work Flow?

Question

Hello,
When one uses the "Merge to Panorama in Photoshop" command, a 16-bit TIFF is created and brought back into LR.  Is it better to make image corrections to all of the RAW files BEFORE using this command? Alternatively, is there just as much latitude with a 16-bit TIFF as there is with a RAW file for image corrections?
What do you folks do?
Thanks in advance.
Dave

Answer 1

Neither Evening nor Kelby (LR2 editions, my LR3 editions are elsewhere) mention masking with respect to fill light and recovery.  They imply that it's an algorithmic swizzle on the conversion from raw to pixels. If that's the case, I suspect they would work OK if you were preprocessing before a panorama and made sure that the fill light and recovery values were used consistently across all segments of the panorama.
Yeah that is not right. The Shadow fill, recovery and clarity are done using automatically generated masks. In principle, you do not want to apply them before merging panoramas as you might get lighting issues in areas that in one shot might be in a large shadow area and in the next shot might be only a smaller area that is hardly affected by the shadow fill slider. This is probably not a big deal when you only apply small amounts of these local contrast HDR-like adjustments but when you do large amounts you are going to want to do it after the panorama merge. The way to do this is to develop all source images very flat using only standard (i.e. no shadow fill, recovery or clarity) sliders, making sure you do not clip highlights and blacks more than you need (generally you are shooting for no clipping) and that you keep contrast low. Make sure to apply all adjustments equally to all images. Then merge to panorama and bring the result back into Lightroom and then do all the adjustments including shadow fills, recovery, clarity etc.
P.S. It is best to not use PS for panoramas in my opinion. It is very inflexible and fails at them quite badly if you have anything but the simplest of image sets.
I'm actually going to go a step further.  I've got a shot in mind that requires a person to be in three different parts of the picture.  It will be shot through the door into a two-car garage (person on left, middle, and right) and I may need HDR to get the exterior and interior to both be OK..  The camera won't move between the three setups.  I'll HDR each setup individuall, then slice the HDRed setups, but I'll slice the picture into three parts and use panorama software to glue them together.  On the other hand, I've got enough slaveable strobes that I may be able to just just blast a lot of light in the garage door (maybe mount a couple of strobes above the roll-up garage door to light the back) and not have to do HDR.
I've done this sort of thing many times in landscapes. The best results are obtained by stitching similar exposures first into separate panoramas and then combining those. My favorite panorama software hugin does this really well as you can just dump all images in there, align them and have it spit out assembled separated exposure layers that you can have it combine for you using the built-in enfuse (awesome open source software for natural-looking HDR). You can also put the separate assembled exposure as layers in Photoshop and combine using masks or the built in HDR function or import them into dedicated HDR software such as panomatix.
That said, I think you will get the very best results if you light this particular scene correctly yourself. Strobes are usually the way to go if you have them.