In practice, a 7000 watt lamp is about as big as can be used for 35mm projection. Larger lamps have larger electrodes, so the plasma of the arc is difficult to efficiently focus on the "postage stamp" sized 35mm frame. And even if you could, the concentration of energy would likely burn the film. For really huge screens, 70mm is needed, since you can then spread out the energy over a much larger area of film.

As far as drive-ins, SMPTE Recommended Practice RP 12 says that you should try to get 16 footlamberts, but you are allowed to be as low as 4.5 footlamberts. So about 1.5 watts per square foot of scope screen area would be the minimum "rule of thumb" for a drive-in. At 5000 watts, I suspect your "half bad picture" is only half of the SMPTE 4.5 footlambert minimum. It may be heresy, but you would actually improve quality if you didn't try to fill the screen, and used a smaller image, say about 40 x 96 feet, which would at least have a chance of reaching 4.5 footlamberts.

 |  IP: Logged

Standard SMPTE 196M specifies a screen luminance aim of 16 footlamberts for indoor theatres, with an allowed range of 12 to 22 footlamberts. SMPTE RP 12 sets the same aim for drive-ins, but allows as low as 4.5 footlamberts. It is certainly possible to achieve 16 footlamberts on a 25 x 60 foot screen, but any bigger with 35mm prints risks heat damage to the film and other problems, so the really large screens are probably well below 16 footlamberts.

I agree that getting 14+ footlamberts on a moderate-sized outdoor screen should look "fabulous". But actually MEASURING the light with a screen luminance meter is the only way to be sure what the light level actually is. Some pictures can look okay at 10 footlamberts, but they will look much better at 16 footlamberts.

------------------
John P. Pytlak, Senior Technical Specialist
Worldwide Technical Services, Entertainment Imaging
Eastman Kodak Company
Research Labs, Building 69, Room 7419
Rochester, New York, 14650-1922 USA
Tel: 716-477-5325 Fax: 716-722-7243
E-Mail: john.pytlak@kodak.com

I know you may not have time to do all of this. But the reason I suggest it is because there are so many different factors that effect the light. For example, if you measure two different screens, but one has a gain screen and the other doesn't, it will throw off the accuracy of the formula.

THX doesn't get real involved in the "picture" part of theaters. I mean they do make sure the image is not serverly cropped, and it is decently lit, but just because the theaters are THX approved does not mean you are compairing "apples to apples."

Generally speaking, I found that the height x height x 12 formula works to 60mm. Then I would add 5% for each 5mm step less. Even then, we would buy the next bigger lamp, even if the wattage value fell on a standard lamp size.

For top/bottom masking situations, we found that the light could be improved by adjusting the tilt of the projector to the correct angle for scope first. Then, move the framing knob and file a plate for flat. The light is a little better than doing it the other way around. I thought I was so clever when I though of this, but then found a SMPTE article that, although didn't say this directly, implyed it strongly.

I think that without doing a real careful survey, the existing "rule of thumb" is pretty good. But, I would think it kinda fun to work out a better one.

Start with screen gain, then factor in its curve, and then audience distribution, and masking type, and lenses and film formats and the amount of cropping of each format, and xenon system efficiencies, a bulb degradation factor, and then make an allowance for focus drift...

Might be better to take a basic formula, look at one or two other major variables and then add a little "Kentucky Windage" based on experience.