Post by Martin HartAs I have mentioned before, each six perf frame is being scanned at 2K.
I don't know how that relates to standard 35mm format which is merely 3
perfs for 1.85, but it's twice as much information, obviously. Then the
three 35mm scans are married together at full resolution. From that all
the necessary resolutions are derived, including DVD, Blu-ray, and
theatrical rez, whatever that is.
Doing something like that effectively gives you slightly better resolution
than a 4K scan. That's not as good as the originals, not even as good as a
top-notch 35mm print, but it's respectable and usable on a big screen.
Post by Martin HartI've seen some of the footage and it looks pretty good. Of course a
contact print off pristine negatives would be optimum but, believe me,
these are not pristine negatives. Scratches and dirt are a problem but
the proprietary scanning system can actually see through dirt with a
separate infra-red pass and scratches are removed through the scanning
software as well. Each frame is scanned four times, red, green, blue and
IR.
If this is a Kinetta, it doesn't see through dirt. What happens with the
IR pass is that it _only_ sees dirt, so you get a frame that shows only the
defects. This allows you to go through with software that replaces the
dirty pixels with the same color as adjacent ones. It works remarkably well
on 35mm although it can have some bizarre artifacts on 16mm reversal originals
because the scratches and dirt are much larger compared with the frame.
The image dyes are transparent to IR so the image itself does not show up on
the IR pass. Unfortunately this doesn't work with B&W film where the silver
image shows up on the IR pass.
Post by Martin HartThe biggest difficulty is the negative fading which, for reasons I don't
fully understand, varies not only from panel to panel but also within
the panels themselves.
It's because the fading is caused by light and air and residual junk. If
the washing is not completely even, the fading may not be even. If the air
flow is not completely even, the fading may not be even. I see prints where
three reels were packed on cores together and the two on the sides fade a
little more than the one in the middle.
Post by Martin HartPanel matching is difficult enough when you have
a perfectly stored and seldom run negative like HTWWW, but when you've
got variations all over the place it becomes a real battle of nerves to
try to pull off a seamless look to the film. In fact, it's impossible
but those panel joins were always there and they'll probably wind up
looking better than ever. I do know that corrections are being made to
focus errors which cause the same object to appear in two adjoining
panels like the famous three-legged woman at La Scala in TIC.
I think leaving the focus errors there is a good thing... some of the
artifacts of the original are what make the original interesting and
unique.
Is the dustbusting and color correction being done on the composited frames
or on the original frames? If the color correction can be done on the
original frames then it should be relatively easy to go out and do a
filmout from the original frames... giving you three strips of 35mm that
have been color-corrected and dustbusted and are ready for showing at the
Dome.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."