Building your own Fluorescents

[Glenn Woiler]

For those interested in building balanced flourescent lighting... this is what I have found out. For starters, you can look at pro lights and see what bulbs they are using and then google them. I bought mine - 55 watts each for about $10 each. I bought 4 of them. I also got electonic ballasts and sockets made for these bulbs that have a high frequency so there will be no flicker problem. I found an outlet that sells polished reflectors here:
http://ahsupply.com/36-55w.htm
I think I will end up getting their kit minus the bulbs. Mine have a CRI rating of... they say... 98 but I think it is really 92. Anything over 90 should yield very good results. I will put all 4 bulbs in one fixture. Then, with the experience gained I will build a second and be done. This should save me big bucks over buying them from a name brand company.

glenn

[Emre Safak]

How did you determine the actual CRI?

[Jaron Berman]

I found that site a month or two ago and have since made a number of flo's. THEY ROCK! AH supply is my new hero. I went to chinatown and had some stainless pans made to fit 2-tube units, as well as yokes to go around them. A few pop rivets and some time later, they came out perfectly, and the light is BRIGHT. I haven't actually gotten a chance to measure yet, but I believe from looks that these ballasts w/ kino bulbs are equal to the output and look from Kino's own parabeam series...minus the price. In the end I paid about 180 for the whole setup including bulbs....not too bad! Glad someone else happened upon that site too.

[Richard Andrewski]

Check out this thread,

http://www.dvxuser.com/V6/showthread.php?t=61114

and this one too:

http://www.dvxuser.com/V6/showthread.php?t=58947

Richard

[Glenn Chan]

I believe you'd use a spectrophotometer to measure CRI. That device will measure the light intensity at each wavelength. From the data you can figure out the color rendering index. Those devices typically cost tens of thousands of dollars.

The practical thing to do:
A- Take your camera and shoot the light. The light on camera depends on:
1- The kind of fluorescent light.
2- The camera you're using. Different cameras will render color differently.

B- A prism or diffraction grating might be a way to get a poor man's spectrophotometer... it'll break up the light into its wavelengths. Then you can eyeball the spectrum for evenness.

[Bob Grant]

Osram publish CRI data for all their fluro tubes, did deep enough you'll even find spectral emmision graphs. Better CRI is the result of adding more phosphors into the mix. Down side is slightly less apparent light output and more cost which is why not all fluros are created equal. Realistically for video not that much of an issue so long as you're using all the same type of tubes to light and even then, well, no one seems to care that much these days.

You can see the differences though, some 'daylight' tubes are 5,400K and others well over 6,000K.

[Marcus Marchesseault]

I really don't like the look of the 6000K+ fixtures. They look okay as a rim light to simulate moonlight, but they make people's faces look colorless. I prefer to err in the direction of 5000K if I can't find 5600K.

[David Tamés]

The color does matter, as Marcus Marchesseault points out, skin tones, especially, are affected. Often I choose to warm the light coming from a fluorescent (regardless of CRI) by using a 1/8 or 1/4 CTO or Bastard Amber gel or even a Surprize Pink to get a little more warmth from the source relative to other light sources. Doing experiments with a Macbeth chart and a variety of gels with your fluorescent sources can be fun and instructive.

[Rob Gregory-Browne]

For those interested in building your own fluorescents, there's a great new DVD available to show you how. It's called Cool Lights. I've seen the DVD and it's very good. I'm already planning several projects myself.

http://www.coollights.biz/cool-light...ting-p-28.html