Lamps For Fluorescent Heads

Hi List

I am working with one of the main lamp manufacturers to develop a range of fluorescent tubes. We are thinking that we will develop two lines - 1 for TV/Video and another for Film.

Can I ask the list for their requirements in order of preference.

IE for TV I think light output is No 1 followed by stable 3200K colour temperature and a high CRI of around 95 or better. I am less worried about the green spike for TV.

For Film I think the green must be as low as possible even if I have to loose some light output (I can always put more heads in to up the light level). I also want a stable colour temperature even when dimmed.

When we talk about "daylight" tubes what colour temp are we looking for? I have seen figures that range from 4200K right up to 7600K quoted as daylight! Whilst this may be correct, as it depends where you are in the world, time of year etc etc, am I right to assume that for film use we want a stable 5600K?

SO I guess in summary what I am asking for is your specs for the perfect fluorescent tube!!! Your chance to custom design a tube. Your answers can be as technical as you want or as basic as you like - we just want input at this stage.

Then we can make a housing for it and we should have the perfect light - Yeah Right!

Thanks in advance for your comments.

Regards

Peter Daffarn
UK Based Lighting consultancy, supply and installation
Projects Department Ltd
http://www.projectsdepartment.com

>When we talk about "daylight" tubes what colour temp are we looking >for? I have seen figures that range from 4200K right up to 7600K quoted >as daylight!

I think of 6000K as a good average for "daylight' tubes.

A little bit of dimmability is nice but I don't think it's necessary to
try and have it dimmable over the entire range unless this would help with marketing purposes. it reminds me of a joke, "What do you call a 2K baby junior with 3 double scrims?" Answer, "The wrong light!"

And, this is probably more in terms of developing a ballast, but I've always thought it would be cool if there were a little switch to easily warm or cool the tubes in a lamp head in terms of color temp- such as a "+1/8 CTO" or a "+1/4 CTO", and varying degrees of warmth and cool. But I have no idea what it would take to actually achieve that!

Toby Birney
Director of Photography, LAC
Vilnius, Lithuania

>And, this is probably more in terms of developing a ballast, but I've >always thought it would be cool if there were a little switch to easily >warm or cool the tubes in a lamp head in terms of color temp

It is a cool idea, but the way fluorescent tubes work (independent of ballast design) is this :

The inside surface of the fluorescent tube is covered with a mixture of phosphors - minerals which fluoresce when excited.

The tube is then filled with mercury vapour. An arc is struck down the length of the tube in the mercury vapour which excites the phosphors and causes them to spit out light in the frequencies that they are wont to do.

The color output of a fluorescent tube is determined by the mixture of phosphors coating the inside of the tube plus the specific wavelengths of light represented in the emission lines of mercury gas.

As we know, Mercury gas emits some light at a number of different wavelengths but has a strong spike in green.

Just about the only variable you can adjust once the tube has been manufactured, is the amount of current that you pass through the mercury cloud. To some degree, the more current you pass through the tube, the more output you get from the phosphors, BUT as you raise the current (and the internal temperature of the tube) you raise the pressure of mercury gas and the output from the mercury arc disproportionately more than you raise the output of the phosphors...so, as you raise the current, you get a little more green and blue in the spectral distribution of the tube, compared to its overall increase in output.

Thus, as you raise and lower your ballast current, you change the proportion of green and the overall output but not really the color temp (slightly bluer with more current)

There have been some ballasts that do manage to dim the tubes to some degree, but it is REALLY hard to change light output without changing color rendition a little for the above reason.

I suspect that even PWM (pulse width modulation) is likely to change color a little since the tube will heat up or cool down and hotter means higher pressure of mercury gas means slightly greener.

By the way, manufacturers like Kino-Flo that make high output ballasts design the phosphor mix in their tubes to compensate for the higher green output of the mercury arc - boosting regular tubes in high output fixtures results in more green the tubes are specked for.

Conversely, specialty tubes made for high output fixtures, if burned in standard fixtures, may balance out slightly more magenta than their design spec.

Mark Weingartner
LA based

If you're going to design a tube for video with a CRI of at least 95 why would you design a separate tube for film? It seems to me it would make more sense to develop one set of tubes for use with both mediums. Why double your production and research costs if you don't have to?

Just be very careful that you do everything right the first time. Get the formulations right, design sturdy ballasts, and make sure your lamp housings are light and easy to use. There are many Kinoflos imitators who have played and lost at this game. Videssence and Icelight come immediately to mind but I'm sure there are others.

(I never worked with Videssence but a client of mine has Icelights in their studios. The housings are metal and very heavy, and the tubes have a very noticeable magenta shift to them. No wonder the company isn't around anymore.)

Art Adams, DP [film|hdtv|sdtv]
San Francisco Bay Area - "Silicon Valley"
http://www.artadams.net/

>...but a client of mine has Icelights in their studios. The housings are >metal and very heavy, and the tubes have a very noticeable magenta >shift to them...

I have to interject something here: Back in the 80's I worked a great deal with the Softube fluorescents - using them on docs, features, and commercials. These fixtures were VHO units - the 4 ft tubes used 1500milliamps per tube, significantly higher than the Kino's and some other early "film" fluorescents.

As a result of the high current, there was a strong green spike in the output from the green emission lines of the mercury gas. The tube phosphors that were used to make the tungsten balanced tubes made a noticeable magenta cast to everything they lit - but on both film AND video they balanced out to tungsten without either a green or magenta shift.

The only reason I point this out is that where the eye saw an imbalance, the film didn't.

We had to spend a little time reassuring new users that the pink didn't show up on film - in video, of course, we could point to the monitor.

Mark Weingartner
LA based erstwhile NY gaffer

>The only reason I point this out is that where the eye saw an >imbalance, the film didn't.

I've seen that when using Flourofil (sp?) gel on fluorescents. The color is like an orange/chocolate color (if I remember correctly) but it photographs tungsten.

These Icelights, however, appear magenta even on video.

Art Adams, DP [film|hdtv|sdtv]
San Francisco Bay Area - "Silicon Valley"

>If you're going to design a tube for video with a CRI of at least 95 why >would you design a separate tube for film? It seems to me it would >make more sense to develop one set of tubes for use with both >mediums.

There is a trade off between light output and CRI. From the TV Lighting Directors I have spoken to most are not bothered too much about the green spike as it doesn't show up on video and are more concerned about getting more light out off the tubes.

I have not had much contact with the film side (hence posting the question on the list) but I think I am right assuming that the film DoPâ€™s are very concerned at the green spike and would ideally like no spikes at any frequency.

Given that the above is correct it is not possible to make a fluorescent tube that is correct for both TV/Video and Film. However the TV LD's could use the Film tubes with no problem (other than a small reduction in light output).

Thanks for the replies posted so far - they are very useful.

Regards

Peter Daffarn
Projects Department Ltd

Peter Daffarn wrote:

>Given that the above is correct it is not possible to make a fluorescent >tube that is correct for both TV/Video and Film.

Why would a tube be correct for video and not for film?

Blain Brown
DP
LA

>Why would a tube be correct for video and not for film?

Film is more sensitive to the green spike that you get from fluorescent tubes. Video is either less sensitive or less receptive to colour balance. I am not sure why but it seems that TV can get away with using standard 55W fluorescent tubes such as the Osram Lumilux de luxe 3000K which has a CRI of 93 or even the Osram 55w Lumilux (increased brightness) 3000K but with a CRI of only 85. Both these tubes are used over here in Europe for TV applications without any problem.

If you were to use these for film I suspect your would get a green tint on the developed image.

The reason I think we need 2 sets - film and TV is because it is cheaper to make the TV fluorescent tubes and so where you are not worried about the green you can use the lower priced and higher light output tubes. For film you would require the more expensive, better colour rendition and lower output film range.

So I guess the answer is that you could just have one made for film but the TV people would be paying a higher price for their tubes then they need to.

Please correct me if I am wrong - that is the reason for the posting!

Regards

Peter Daffarn
Projects Department Ltd

> Why would a tube be correct for video and not for film?

It would and it wouldn't. By the nature of design of a florescent tube, you are never going to get an even distribution of color. Video ahs always been more forgiving because of it's ability and sensitivity to color. Green in itself is the reference channel in video so video likes green, but films physical characteristic doesn't mask that green tinge as well. I'd say that there is a lamp for both. It's the best lamp made for film.

Video is more forgiving so if film sees it well, video will see it better. Problem is I haven't found a fluorescent that really works well with video. It's a different kind of illumination.

Walter Graff
BlueSky Media, Inc.
www.bluesky-web.com
Offices in NYC and Amherst Mass

Walter Graff wrote:

>Problem is I haven't found a fluorescent that really works well with video.

Huh...Care to expand on that Walter?

John Roche, gaffer
NYV

>Problem is I haven't found a fluorescent that really works well >with >video.

>Huh... Care to expand on that Walter?

Whoops should have said film.

Yes fluoro's give a unique look to film, but a limited look. I don't like fluoro's for general purpose film unless I am going for a cold look. There is something unnatural about the texture of light produced by a fluoro's.

Un-natural doesn't mean bad because of course if you are going for an unnatural look then fluoro's work well, such as at a desk on a computer or phone (high key), or in a phone booth, or on a subway station.

Walter Graff
BlueSky Media, Inc.

>IE for TV I think light output is No 1 followed by stable 3200K colour >temperature and a high CRI of around 95 or better. I am less worried >about the green spike for TV.

Nothing in a TV studio is 3200K. That's a lab test bench number. Real world is more on the order of 2850. Round it to 2900 if you like. Shoot the light through a lens or bounce it off of a white surface...send it through a piece of opal and you get a substantial drop in color temperature. Since fluorescents are often bare, they don't get the same drop as you would with any tungsten/halogen luminare.

You've got to be able to mix the fluorescents with the incandescent lights or they would never make it to my studio. Also, we DO care about the green spike. A 95 CRI is nice, but 92 is acceptable. Good luck dimming while maintaining your CRI. Even with dimmable fluorescent fixtures you rarely have the same CRI balance throughout the range.

Bruce Aleksander
ABC / Disney
Houston

>So I guess the answer is that you could just have one made for film but >the TV people would be paying a higher price for their tubes then they >need to.

Still...KinoFlo doesn't make a separate tube for video, and we still use them for video, and they're still the best. And all the companies that I know of who have tried to corner the market on video-only fluorescents (Videssence, Ice Light) are long gone. So even if you make a cheaper tube for video production they'd still better be really good tubes, and easily discernable from the film tubes so that no mistakes are made.

>There is something unnatural about the texture of light produced by a >fluoro's.

Actually, that can be used to one's advantage. The quality of shadows produced by raw Kinos can be very interesting. I've seen some stuff that was obviously lit by raw Kinos but in the context they looked marvellous.

Having said that, though, I always use them with diffusion and they look great, even if I do have to warm up the 3200s to taste. There's something about a glowing diffused source totally without specularity that can be magical.

Art Adams, DP [film|hdtv|sdtv]
San Francisco Bay Area - "Silicon Valley"

Walter Graff wrote:

>Video has always been more forgiving because of it's ability and >sensitivity to color. Green in itself is the reference channel in video so >video likes green.

One thing you gotta say about CML. Ya learn stuff. A very clear and simple explanation.

Blain Brown
DP
LA

Blain,

>Video has always been more forgiving because of it's ability and >sensitivity to color. Green in itself is the reference channel in video so >video likes green.

and

>One thing you gotta say about CML. Ya learn stuff. A very clear and >simple explanation.

Yes, you do, but be careful.

Yes, green is the predominant colour in video systems. From my recent post on CML-video...

"Green is split first because its the predominant component of white light in encoded colour television systems [NTSC, PAL etc]. Remember Y = 59% green + 30% red + 11% blue [approximate figures]. So green carries most of the detail and hence its best to give it the cleanest, most direct path to the tube/chip."

...which was part of a post discussing green/magenta chromatic aberrations in out-of-focus high contrast transitions in video pictures.

However, just because green is the "reference channel" does NOT mean video cameras "like" green spikes in fluoro tubes.

Be careful with what you read on the internet!

Cheers,

Clive Woodward
Video Technician
Perth, Western Australia.

>Green in itself is the reference channel in video so video likes green.

I disagree that. I think green is one of the hardest colors to work with. Either with film, video or HD. That's why it's so common to see so many movies, tv shows, etc with brownish forests or woods. Or take a look at those fluoro-green Kawasaki supercross dirt bikes. Video never really get that color.

Regards

Rodrigo Lizana Lamarca
Pixine, Cine Digital
www.pixine.cl