Jump to content

Flourescent questions (CRI and brightness)


Recommended Posts

So, I'm doing a 2 month shoot in a school while it's in session (yup, crazyness) . As a result I can't hang and leave lights for the whole time, but we have got permission to upgrade the lights in the school with certain standardized fluorescent fixtures.

 

Problem is I just got the word today that the order need to go in tomorrow, so I don't have time to test, so any help would be greatly appreciated.

 

My main question is does anyone have recommendations for a 4' daylight-ish globe that's cheaper than kinos.

 

We're shooting on the Red, and I would prefer no green spike, which I've noticed the Red seems more sensitive to than CCD cameras used to be...

 

I found this one that looks promising:

 

philips natural sunshine

 

Thanks!

 

(next I have to guess if three bank fixtures will give enough light, but I should be able to take some measurements tomorrow morning. Interestingly it is rated at 1915 lumens, while a 40w kino (2ft) is rated at 1600 lumens, so it should be brighter, which I guess makes sense since it's 4 foot so greater surface area even if the same wattage)

Link to comment
Share on other sites

  • Premium Member
*sigh* no-one ever responds to my posts, am I being to specific or did I just forget to put deodorant on this morning?

 

> My main question is does anyone have recommendations for a 4' daylight-ish globe that's cheaper than kinos.

 

Sorry, but I don't know anything useful on this. We just use KinoFlo's, and haven't tried anything else.

 

 

 

 

-- J.S.

Link to comment
Share on other sites

  • Premium Member

Give us a chance - we're not all in the US.

 

I don't have much time on Red, but you're right, it does seem unusually sensitive to CRI problems for an electronic camera.

 

There are various high colour-temperature, high CRI fluorescent lamps around, most of which will, one would reasonably assume, be better than whatever was there by default. Without testing, though, or finding someone who's done it, that's about the best information I can give you.

 

On the upside, daylight flos tend to be better in terms of CRI, since the blue phosphors have a wider bandwidth than the red. CRI figures into the mid-nineties are available, although it's not the only measure.

 

P

Link to comment
Share on other sites

  • Premium Member

I just used some Philips tubes (colortone, and the labeling is F40T12/C50 Colortone 40Watt) bulbs on a shoot with the EX1. They were pretty good with a CRI of 92 I believe which was enough for me (as I wanted some green spike) and the best part is I just found 'em in Home-Depot.

Link to comment
Share on other sites

Thanks guys,

 

I just checked out the location and the fixtures, and it turns out they are T8 so those philips TL-950's Kar recommended may be perfect - A CRI of 98 is pretty crazy considering kinos are rated at 95...

 

Of course now I'm also looking at the jealously at the osram 22026 FO32/850XP/ECO which put's out roughly 40% more lumens than the philips: 2850 vs 1860 at the same wattage but only has a CRI of 85. It's amazing to me that they could put out that much more light at the same size and wattage.

 

However, I think it may be best to go with the safe philips if I can't test...

Edited by Mathew Rudenberg
Link to comment
Share on other sites

Bear in mind that CRI has to do with human vision, not film or video. And it's kinda kludgey, too....

 

 

 

 

 

-- J.S.

 

 

John, would you mind explaining/expanding upon this concept?

I always thought CRI described the variety of wavelengths produced by a light source, as in it's relative purity and accuracy in illuminating colors... Wouldn't this matter more for analog sensors than the human visual system, since it has the ability to adapt to color casts ect?

Sorry i'm still learning. This whole concept of CRI gets very confusing to me

Link to comment
Share on other sites

  • Premium Member
John, would you mind explaining/expanding upon this concept?

 

Well, it's not so easy without being able to draw some pictures.... or even with pictures. The Wikipedia article is OK at the moment:

 

http://en.wikipedia.org/wiki/Color_rendering_index

 

In reading it, bear in mind that the "two degree standard observer" the whole thing is based on is a mathematical representation of human color vision that was created by having people look at pairs of test colors and adjust one to match the other. Note well the "Criticism and resolution" section at the end.

 

 

 

-- J.S.

Link to comment
Share on other sites

Bear in mind that CRI has to do with human vision, not film or video. And it's kinda kludgey, too....

-- J.S.

 

What about the Spectral Power Distribution? does that indicate how much green will read on camera?

 

post-13335-1255235247.jpg

 

This is the Osram CRI 85 that puts out much more light

 

post-13335-1255235254.jpg

 

This is the philips 950 with a CRI of 98

unfortunately the onlly graph I found for the philips is barely legible and the scale is different, it looks like "per 5nm per lm" to me.

 

So to convert the Osram to the same scale (if I can still do math at all) -

1000 x 0.09 (the tip of the green spike) = 90

90 x 5 = 450

 

So I think I get a 450 green spike on the osram, and a 310 on the Philips, which makes sense since the Philips has a higher CRI.

 

But what does it mean to the Red camera?

Link to comment
Share on other sites

  • Premium Member
What about the Spectral Power Distribution? does that indicate how much green will read on camera?

...

But what does it mean to the Red camera?

 

Spectral power distribution shows you how much light you get at each wavelength. The absolute height of the spikes doesn't matter so much as the ratio between them and the lower level phosphor output. The curves have pretty much the same shape. Scaling the whole curve up or down would be equivalent to just having a brighter or dimmer light.

 

Kodak, IIRC, had spectral sensitivity curves that you could probably use to predict how the film would respond to a given spectral power distribution. But the result would be more curves on graph paper -- basically the three dye density functions, each multiplied by the spectral power distribution function. That might show you the 546.1 nanometer mercury spike landing where the green layer has a lot of sensitivity, but getting from that to knowing what the resulting picture would look like isn't particularly practical. The way to deal with that problem is to shoot tests and look at them.

 

I'm not sure if Red publishes any spectral sensitivity curves. Even if they do, the important thing is how the pictures look to your eye. So, looking at tests is even more so the way to know, because with electronic cameras, you can see immediately what you get.

 

 

 

 

-- J.S.

Link to comment
Share on other sites

  • Premium Member

Hello Matthew,

 

It's good you got out the graphs. Matching the spikes to the sensor or stock is a constant challenge for DPs. You can get into a usable ball bark with any of these lamps and a little lens filtration and even better with white balancing. The beef I have with flos are the spikes. Every so often, someone here will agree. But, for the most part, folks just work around the spikes.

 

So, here's the beef with flos- They're spiked. Even if you can fudge the light and sensor to get a useful spread, you still end up with holes in the color. You may find that you like the look you get. People seem to like the slightly sterile look of HMIs, for example. I have a preference for tungsten light because it delivers a nice, smooth and full light.

 

An easy trick to fill out flos is to throw some gelled tungsten into the scene. Having a few tungstens and an assortment of gels also keeps you from absolutely having to relamp every location. You might be acceptably surprised how little viewers care if the flos that are visable in the shot are off color when the subject matter looks right under set-up lights.

 

This really doesn't address your question. But, you may find it useful somewhere down the road.

Link to comment
Share on other sites

Well, I ordered the fixtures, but postponed the globes until Monday.

 

I agree that Tungsten is the prettiest of lights, but my main beef against HMIs is that they always manage to go down at the worst possible moment. I love the simplicity of a Tungsten light - if it doesn't work usually the globes blown or it's not plugged in...

 

In this case the purpose of the flos is not to light primary talent but to provide ambiance in a long school hallway where I'm not going to be able to rig space lights, and I may not even be able to hide ceiling bounce or two.

 

So, I've flipped 180 degrees, and I'm just going to assume the worst; the light quality will be imperfect, in which case I'll have to add some minus green. In this case I can always take away light (add more diff or ND gel) but I can't increase it - so I'm thinking the higher output globes will be a better choice in the end.

 

I'll try to post some frame grabs in a couple weeks to show how green or not they turn out...

 

Thanks for all the help!

Link to comment
Share on other sites

Of course now I'm also looking at the jealously at the osram 22026 FO32/850XP/ECO which put's out roughly 40% more lumens than the philips: 2850 vs 1860 at the same wattage but only has a CRI of 85. It's amazing to me that they could put out that much more light at the same size and wattage.

 

It puts out more light precisely because the CRI is lower. Given the same type tube, wattage and all other factors being equal (including the ballast being the same type and power factor thats driving both high and low cri tubes), the lower CRI tube will always be higher in lumen output. Its simply because the green spike which is not appreciated in filmmaking is well appreciated in the lighting world for the extra "umph" it provides to electronic lighting from fluorescent to metal halide and even now to LEDs. Kinoflo got around this issue by simply overdriving the tubes to try to win back the lost lumens when the CRI was made higher.

Link to comment
Share on other sites

It puts out more light precisely because the CRI is lower. Given the same type tube, wattage and all other factors being equal (including the ballast being the same type and power factor thats driving both high and low cri tubes), the lower CRI tube will always be higher in lumen output. Its simply because the green spike which is not appreciated in filmmaking is well appreciated in the lighting world for the extra "umph" it provides to electronic lighting from fluorescent to metal halide and even now to LEDs. Kinoflo got around this issue by simply overdriving the tubes to try to win back the lost lumens when the CRI was made higher.

 

Interesting, I did notice the kino globes are much higher wattage for their size than the standards, but don't put out much more lumens.

 

If I then gel with minus green, will I simply be filtering out the extra light?

Link to comment
Share on other sites

  • Premium Member

The important thing here is that lumens aren't physics, they're perception. They're weighted by the sensitivty curve of the human eye (the photopic = cone = color vision one -- it's different for scotopic = rod = B&W vision). The peak is 683 lumens per watt, at 555 nanometers, which is green. It's not far from the 546.1 nanometer mercury spike.

 

The manufacturers use lumens because the primary purpose of their product is for people to see stuff. Lumens are for eyes, not cameras.

 

 

 

 

 

 

-- J.S.

Link to comment
Share on other sites

If I then gel with minus green, will I simply be filtering out the extra light?

 

Yes you will. A lot of people don't understand or believe it but there is an explanation.

 

Most filters of any kind reduce the output of what they're filtering. Light waves as well as sound waves. Listen to your stereo with the bass turned up. Those eq knobs are filters for adding or subtracting frequencies in a simplified way of explaining it. Turn just the bass down thus filtering out low frequencies. You may find yourself reaching for the volume knob to make up for the frequencies being filtered that gave the sound more "volume" or at least perception of volume. Same with electronic light such as flos or hmi. They lose a lot of volume or perceived volume when the peak area is filtered in the name of higher CRI.

 

When a tube manufacturer manufactures a higher CRI product, they in essence add more magenta to the phosphors used in the tube. This is no less of a filter than if you added minus green gel to the outside of the tube. Thus since the green spike is what really gives the light its efficiency, you lose light.

 

Kino simply "turned up the volume" so-to-speak by driving the bulbs harder to replace the lost light of a higher CRI tube and 95 is realistically about as high as you're going to get anyway. Hope that helps.

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...