Yes, mixing filters can be great. There is a real science to it, but you can
also play around. As for testing on the cheap...
What I did when doing more precise colour projects... I bought a really good
little light table... Geppe makes a portable 4x5 inch light table (more of a
tablet) that is colour balanced to 9600 kelvins (dalylight) and runs on 6 AAA
batteries (it also has an AC adaptor). I think it was about $100 when I bought
it back in 2005. yes... film projectors don't run at 9600 K.... however, I
found that it was nice to have a light table that was a known value as a
starting point.... (it also comes in handy when doing expanded cinema
performances when looking for loops on the fly)
I also had an analysis projector that could project one frame at a time without
burning the film.
I could look at single frames as long as I wanted and make notes.
So then, when I did my colour filter tests on the optical printer... I would
just do one frame per filter... I would do a series of frames, making a change
at each frame... followed by 5 black frames, and then move on to the next
series of tests. (I was working with superimposing various layers of film,
through various colour correction and ND filters... so finding just the right
balance involved over 9 different filters per frame... lots of variables!
I kept notes of what I was testing... I printed out an excel spread sheet and
left columns and rows blank, but big enough to write in for the following
categories... it was essentially a table with the following columns:
# of frames : shot # : description of shot : layer : ND filter : CC filter :
Colour Separation (CS) filter (Red, Green, or Blue) : other filter : aperture :
This table was then my guide when I got my film back from the lab.
Once I got the film back from the lab, I would look at it with the table, and
use a sharpie to write directly on the film itself... not all of the details,
The black frames between test sequences were invaluable for knowing which
sequenc is which... sometimes the differences were incredibly subtle.
Because I was only doing one frame for each test, my roll of film that went to
the lab was only about 5 - 10 feet.... the tricky thing is, that you really
need to leave enough head and tail on the film so the lab doesn't just cut your
test off when loading their machines (I learned that the hard way)... to be
safe... give them about 15 feet at the beginning and the end... and also write
"SAVE ENDS" I remember I could easily spend 5 days or more prepping just a 10
foot test strip... then when I got it back, I would analyse the results both on
the light table, and also projected with the analysis projector.
If you don't have a light table or analysis projector, you would have to shoot
A LOT more film.... because you would want to see it projected, and you need
time to look at it and process it... 1 frame, even 24 frames would go by wayyyy
too fast... if you're judging your tests on a regular projector, you'll want to
give yourself enough time to look and assess each test... it's up to you to
decide how much time you need to see it... then put a section of black between
each test so that you know when one ends and the next begins.
Hope that's helpful... best of luck, and most importantly, have fun with it!
Amanda Dawn Christie
On 2014-12-27, at 7:14 PM, Jeff Kreines wrote:
> With the filters Scott suggests, shoot a series of single frames with
> different filter combinations and exposures — 5 feet of film will give you a
> lot of tests cheaply, especially if your lab is friendly about running tests.
>> On Dec 27, 2014, at 4:57 PM, Scott Dorsey <klu...@panix.com> wrote:
>> Sure, you could use CC filters on the light source. Ask your local camera
>> store for a CC filter kit like people with B&W enlargers used to use for
>> printing still color work.
>> But you could also just let it go blue, and correct it in the final print
>> when you print the interneg. You won't have as much range, but if you are
>> starting with a blue positive original you probably don't have much color
>> range in the original anyway.
>> FrameWorks mailing list
> Jeff Kreines
> FrameWorks mailing list
FrameWorks mailing list