[Tig] Emulating Print Transfer Functions

Rob Lingelbach caltech
Sat Mar 1 01:13:01 GMT 2003


On Fri, Feb 28, 2003 at 04:52:05PM -0800, S. T. Nottingham III wrote:

> Now I admit, if that
> harsh/hard look is what the director wants, then you could go that route.
> However, in more than twenty years of transferring feature material for
> video release, every time the issue of print transfer reared it's ugly head,
> we were able to prove decisively that an IP yielded much more satisfactory
> results.

it is exactly because the director wants the look of the print:
extremely black blacks, and the ability to have the highlight
detail preserved in a curve that the print already provides.

I think we're just not defining the argument here, that we're not
saying print makes a better transfer, it makes a different
transfer with loss of detail, but with addition of some other
sometimes desirable effects.  Another effect I hadn't mentioned
is that if it's a print done well from a clean negative dirt in
transfer is obviated.  (i know these days with PTRs etc. dirt
isn't very much a factor anyway.)

And when I was saying sky detail I meant the kind of sky with
clouds \:   ..to reiterate, it is available (the cloud detail) in
the negative but one must apply a curve to the response to get
it, and it becomes somewhat easier to work with a print that has
the curve built-in.  But it is very much the artifacts and edgy
inky grainy look you can get from a print that can be sought
after by director/DPs and colorists.

it's not to say that negative isn't the best medium for detail,
it is, but when detail isn't desirable, but rather a print effect
is, then print works great.

-- 
Rob Lingelbach    unix system administrator    caltech at calarts.edu
Simultaneity is impossible.  --Einstein





More information about the Tig mailing list