[Tig] compositors doing grading

[Rob Lingelbach]

> Here in Sweden, the trend for so long has been fix it in online -
> lots of
> passes, reference stills or clips from the tk session, the the Flame
> guys
> put together everything.  This was not always the case, but more
> often than
> not.

what do you mean by "online" in this case chris?

> bloomed quite nicely.  Compositing and online is a perfect
> training/development area for would-be colourists.  Now I'm not say
> that all
> flame/smoke people are great colourists - but the best flame people
> give the
> best colourists a run for their money.

hm, I'd say the best training ground for colorists is production.  
They see the
scenes at the source, and understand about lighting, the most
important aspect
of color/grading.   Flame/smoke people are second-hand artists if
looked at from
the perspective of the production (as are colorists).

>> Regarding prints from the labs: there are certain tendencies the
>> lab
>> prints have that the DI shows to poor advantage.   Example 1,2,3:
>> I've been to 3 labs with Truelight LUTs for each one, printing at
>> each lab.  Every print looks (relatively, with some color
>> problems) the
>> same, except that *there is way too much contrast in the image*.
>> Now we can assume that going "down a generation" is going to lose
>> detail.
>
> Sounds a bit strange Rob.  Luts are one specific series of colour
> transformations.
>..........
> 0 copy should start with those aims.
>
> If your prints are coming out with too much contrast, something in
> your
> setup is wrong.  Plain and simple.  Now, I've heard certain
> colourists
> maintain that .04 dens +/- in RGB is unacceptable - but that's just
> the way
> it goes.  Your chem can be locked the fuck down and you couldn't get
> better.

hmm,  now I'm thinking the labs are intentionally building in the
contrast, because
I've been to three now and the results are exactly the same. 
However, let it be said that
the bulb in the DI  JVC projector had aged beyond its time, and
perhaps this will produce
fruit in further tests.

> Are you printing the neg at your shop?  Common colourist error is
> fooling
> the truelight lut by feeding it a un/scaled RGB range  and then
> using
> scaling in the truelight lut in order to have a nicer signal on the
> scope.
> The projection looks good, but the prints have to much/little
> contrast.

unfortunately Chris that level of control isn't available.  It's an
18km drive over
rough roads to the lab we have to use, due to the DoP's preference.

>  Arri's aqua
> and
> Filmlight's calibration strip are used similarly in calibrating
> everything
> from your Northlight to the arri to your monitor to whatever...

thanks for the references.  just need to know (am in third world)
where to find these.

>> This is a huge lack of understanding between labs and
>> facilities in
>> certain countries.  they don't seem to understand the idea of
>> calibration.
>
> Who is "they" in this sentence?  Labs have been responsible for
> internal
> calibration for forever.  Or perhaps you mean facilities?

"they" is the antecedent, "labs and facilities."  in concunction. 
And where I am,
as opposed to where you are, may have a lot to do with it. 
Communication is on the
low side.

> So many of the most common issues aren't really colour related in
> terms of
> grading a visually pleasing or technically correct image, but are
> grounded
> more in practice, procedure and working knowledge of both the post
> and lab
> sides of the DI coin.  It's not really related to being a colourist
> or a
> compositor.

but a colorist, well-trained, knows not to touch the secondaries
until a certain point has been reached.  I've seen compositors reach
for them before the density of the shot has been addressed. 
However, the best colorists often haven't a clue how to grade a shot
so as to achieve the color separation necessary for separation-based
compositing.

-- 
Rob Lingelbach
http://www.colorist.org/robhome.html