Small room for color grading and sound mixing

[Marcel Birgelen]

Was it a big issue when smooth spectrum carbon arc lamps were replaced by spiky spectrum xenon lamps? Did colour timing rooms stick with arc light until forced to adopt xenon?
I remember trying to match colour on two timing room xenon projectors (for changeover operation).
Our matching involved getting about a dozen Super Lume-X mirrors and finding the two with the closest colour temperatures. Once that was done, measurements were taken and expensive custom filters were put in the light path to be "perfect".
I respect that colour timing is important.

There is a difference here. Xenon lamps do have a spiky spectrum, but:

- Several filters inside the machine help to get rid of light we don't want, like too much of the rather harmful UV spectrum.
- In the visible range, it's not all that spiky and it's continuous. It's also easier to filter light away in a certain spectrum than to fill up the gaps created by the light source.

Also, with digital projectors we have the opportunity to tune the image generation part, something that wasn't possible with film.

Color timing in the digital age has become a whole lot more flexible and less of a dark art than in the analog ages. Still, without a good reference source you're nowhere.

[Steve Guttag]

Color timing on Xenon was always an issue. Normally, one refers to its color temp with the target being 5400K. The problem with the single number is that it presumes one is moving on the "Daylight" scale. When you filter, you tend to move off of the daylight scale and are not approximating its color temperature. Filtering also tends to affect all colors with a very broad brush when you may only be wanting to fix parts of the spectrum. Think about audio EQ. If the circumstances of the room/speaker cause you to have a non-flat response, you'd like multiple ways of "correcting" it. Conversely, using a filter to "fix" wrong color is like using just a treble control to "fix" EQ. You can do a bit with it but you are moving a whole range of the spectrum rather than fixing a particular problem.

Strong did use a 3-letter code on is reflectors as they measured them. I found them to be reasonably well matched if those letters matched. For our more critical screening rooms, we always specified the reflectors required (and got them) and they were acceptable. I had one review room where they needed to be able to judge for both xenon and tungsten. We had in-front-of-the-lens filters to shift th spectrum from xenon to tungsten and while you could get a meter to show it in the tungsten range of 2700-3200K...it never looked right and if you used a projector with an actual tungsten lamp and shot it side by side...they were not the same.

Now, with spectroradiometers, you can not only see the color temp but where your white point is on the CIE chart and why the single number colortemp is insufficient for specifying color on a projector. However, back when your tools were reflectors and filters to change color...you worked with what you had. Such is not the case for digital where the three primarly colors can be worked with, to some degree.

There was significant desire to keep with the 5400K spec for film after the switch to xenon. I was one of the minority that wanted us to abandon it and look for something within xenon's natural color as that would be more achievable rather than a filtered version of light that reads right one one number but is off of the daylight scale so it is an approximation. As I recall, xenon wants to be closer to 6000-6500K. Whatever number it is, if the DPs and color timers would have used that as their reference, they could have then timed for its more natural light and had a more repeatable result, right down to the commercial cinema. Then again, show me two reels of film that are tightly color matched.