This scene between Raul Julia’s character Sandy Stern and Harrison Ford’s Rusty Sabich was shot inside the National Arts Club, located on Gramercy Park in New York City.  It takes place at the 0:58:28 mark.

            The camera pans Julia left to right and dollies in on him as he enters the restaurant through the bar and joins the waiting Ford at his table.  The rest of the scene takes place there.

            Note the dark mahogany tone of the walls…  An 8’x8′ sheet of white foam core was suspended 12′ above where they sit; it served as a bounce source for a 2500W HMI located on a floor stand several feet from the table.  Despite this being an interior scene – and an easily controllable one in which a tungsten standard would’ve been the obvious choice – Willis once again balanced his lighting to 5600˚K and used an 85 filter behind the lens of the Panavision camera.  I never learned why he chose to do it that way!  Perhaps it had something to do with imbuing the emulsion with an inherently cooler tone through the use of the HMI’s…



  1. The choice of lighting instrument and color balance is fascinating. I love seeing the hand drawn diagram and the note about floating in a net when the actors stand. It’s wonderful to see the process at work like this.

  2. Thanks Richard for all these amazing posts. I feel I’m learning without being on set.
    Regarding the choice of balancing to daylight and using an 85, is it possible that Mr. Willis wanted to match the sunlight coming from the windows with the HMIs instead of putting some CTO on the windows and using tungsten lights? As you said, perhaps he wanted to keep the cool vibe of the HMIs, creating more separation with the tungsten bulbs of the location and having the characters stand out even through color.
    I’ve been at the National Arts Club a couple of times and I remember the park outside, on the north side. I’m not sure what time of the day this scene has been shot, but maybe the trees and the sun were creating shadows, moving the color temperature closer to 7000˚K. It seems to me that the window is facing them directly, so he pushed for the same temperature and used the 85 filter to warm up the look, but keeping the characters more prominent in a pool of cool light surrounded by practicals.

  3. Luigi – Perhaps the diagram and my description didn’t make it clear, but the question as to why Willis chose to go with a daylight balance is due to the fact that the table where the actors sit is so far removed from any hint of light from the outside. There was simply no way possible ambient daylight would’ve had an effect on the photography. And even if it could have, some CTO on the window would’ve easily allowed for a tungsten standard. As it was, there was already a layer of ND3 applied. My conclusion is that he was going for the infinitely small difference in color tone delivered by using the daylight\85 filter combo with the tungsten-balanced stock. We continued to use our standard tungsten printer lights, so if you’ll notice – the look is somewhat on the cooler side…

  4. I have been under the impression tungstens color temp always vary depending on the age of the bulb and actual power output. Perhaps the HMI gave a consistent output to get that white light look.

  5. David – It’s actually the other way around. As long as you monitor them from time to time, tungsten sources are the most consistent of all. Willis and his gaffer had very sharp eyes and never used any lamp that was not performing at its best. HMI’s – especially back in the 80’s – were notoriously inconsistent in the color they delivered. Though they’ve improved, I remain a little suspect of them every time I use them, even today.

Leave a Reply

Your email address will not be published. Required fields are marked *