Re: reg 8mm dupe negs

From: Jeff Kreines (email suppressed)
Date: Sun Feb 04 2007 - 20:55:54 PST

On Feb 4, 2007, at 10:10 PM, amanda christie wrote:

> okay... that's making sense slowly to me... just
> trying to grasp it...
> if the perfs are the same size... and you can run
> unsplit 8 through a 16 projector... then how can the
> distance be different? wouldn't it create a huge
> difference over the length of hundreds of feet of
> film? not sure if i'm wording my question right....

2 x .15 = .3000 -- 2 8mm perfs to 1 16mm perf.
.2994 is only a tiny bit closer than .3000 -- .0006" -- but that does
matter in printing.
> i would show it as 16mm... or actually, i'm not
> looking to show the print... you see, i want to work
> with these images thrugh various optical printing
> techniques to create new images... but i don't want to
> damage the original kodachrome... that's why i want a
> dupe 16mm neg... so that i can make 16mm prints that i
> can work with (using both an oxberry optical printer,
> and my analysis projector for rear projection since it
> shows the whole film width edgecode and all) without
> risk to damaging the original.
> these films were never meant to be shown as 8mm
> format... that's why i didn't have them split.

But the lack of reversal print stocks (ideally, ECO in this case)
makes this messy and you face either significant generation loss of
the pain of working from a negative rather than positive.

I'd work from the originals, and just be careful. Film needn't be
damaged -- priceless originals are optically printed all the time.
Keep the gate and machine clean, and be careful.
> 16mm to 16mm....
> my options are contact print, or optically print.
> i would assume that optical printing would have more
> loss of quality due to the intermediary lens....
> correct me if i'm wrong though.

Not if the lens is good. Contrast may increase, and if the lamphouse
is specular grain may appear to be increased. But go optical -- try it.

For info on FrameWorks, contact Pip Chodorov at <email suppressed>.