CFL 5000k vs 6500K

[Jim Froom]

I have some 6500K CFL light bulbs (800 + watt output) that were $$$$ in cost.

Today, at home depot, I saw some cfl bulbs with 1000 watt output and they were 6500K and I'd say maybe $ or $$ in cost.


Questions shooting with Canon XH AI HDV and HV20.......
1. Issues mixing the 6500 and 5000k lights as fill and/or key light?
2. Which set would work best when mixing with light coming in via the windows?
3. Does cloudy day vs brilliant sunshine make a difference in the middle of the day?


From wikepeida
"# 5000 K: Typical warm daylight
# 5500–6000 K: Typical cool daylight, electronic flash (can vary between manufacturers)
# 6420 K: Xenon arc lamp
# 6500 K: Daylight°"

thanks in advance for any advice....

[Emre Safak]

1. Obviously there is a noticable color temperature difference. However, if you were going to gel it slightly warm anyway (e.g., for the key for an interview), then you just use a 5000K bulb.
2. Definitely 6500K, since the color temperature is higher in the shade (more bluish).
3. Overcast weather is Mother Nature's softbox. It decreases the intensity and saturation, and increases the diffusion and color temperature.

[Shiv Kumar]

Jim,

With CFL lights you have to take at least one other factor into consideration.

CRI or Color Rendering Index.

a CRI of 100 is considered ideal for Photo/Video work and corresponds to the "ideal" incandescent bulb.

CFL you buy at Home Depot are typically 82 CRI. In other words the quality of those bulbs are not too great. I've seen "Full Spectrum" CFL bulbs at Home Depot as well. They don't specify their CRI, but are more expensive than the other bulbs even by the same manufacturer.

Take a look at this thread for more info, cost and where to buy.
http://www.dvinfo.net/conf/showthread.php?t=103871

[Richard Andrewski]

We've had many post around here verifying as I've been saying for a long time that CRI 82 is just fine for most digital video work. Home Depot N:Visions are less expensive and more accessible than many and available in 6500K, 5500K as well as 3500K and sometimes even 2700K.

http://dvinfo.net/conf/showthread.php?t=90647

You'll find that many have different opinions on the subject and you eventually will have to make up your own mind whether the super expensive bulbs which are usually only about 8 to 12 points better in CRI than the ones like Home Depot sells are worth it to you--not to mention the added costs of shipping. The N:Vision 5500K bulbs are really superior and are well thought of in conventional use too with many surveys showing their bulb coming out on top.

My feeling on this subject is you have two philosophies in the fluorescent lighting for media production world.

The really pro solutions with separate T5HO ballasts, higher power factor (effective watt output) and efficiency that use tubes like the 55w biax, which is slowly becoming the standard where fluorescent tubes are used in TV and some film production. The T12 75w is somewhat well thought of too especially in film production. Lots of color temperature and different CRI options available in those biax types. Every major studio lighting manufacturer has a 55w biax fixture solution and many small companies too.

Then you have the CFL solutions with internal ballast which can't compete with the T5 ballasts in efficiency or power factor (a 55w CFL does not put out the same light as a 55w biax) but there is a perceived convenience or ease of use factor for photography and portable videography use. Power factor and real output is always lower on CFL bulbs though, because you simply can't make as good a ballast that can fit in small area of the base of a bulb. You never see power factor quoted in the self-ballasted models because they're not proud of that spec but it's usually between .55 and .75 whereas a good T5 ballast for a 55w tube is .95 to 1 which means that it uses most of it's theoretical watt output and therefore attains higher lumen output too.

IMHO, these are pretty much DIY and convenience solutions and to pay large sums of money for them seems ridiculous--particularly since you're paying for the ballast over and over again and throwing it away when the bulb is worn out or is broken--leaving a ballast that usually still works permanently attached to a tube which is worn out. Thus, a Home Depot spiral is priced right for this kind of "disposable" or convenience use when you're not a stickler for the extra 10% or so on the CRI.

Bottom line, if you need higher CRI, go with the 55w biax tube and fixtures with the ballast built-in, otherwise don't overpay for spirals in the lower end / low cost / DIY solutions. Hope this helps.

[Shiv Kumar]

Richard,

I'm curious. How have you concluded that 82 CRI is just fine for video work? In tests I've conducted (and they may not be scientific tests) there is a marked difference between 82 CRI and 92 CRI.

Yes, I have read the thread you pointed to. The discussion was not really about CRI.

Honestly, would love to learn about your tests and conclusions.

[Emre Safak]

Everyone has a different idea of what is acceptable. IF your scene does not contain a lot of saturated colors you may get by with a lower CRI.

[Richard Andrewski]

The post may not have been about CRI but it was about inexpensive CFLs. I didn't see anyone saying that the color rendering wasn't acceptable and haven't heard one person complain about the N:Vision brand of 5500K's. No scientific tests, just good video of interviews and such as many around here have used it for that as well as general indy filmmaking use.

I didn't say it was for the person that's shooting a 7-Up commercial for a huge client and he has to make sure the colors on the 7-Up can are exactly the same each time. In any case, why a person like that would be buying spiral CFLs and building a DIY or putting together a 9 spiral CFL solution to shoot his work with would be beyond me. The pro solution I mentioned using 55w or 75w tubes would be a more normal choice for that guy who's clients are paying him huge sums anyway. And by the way, that is what that person is using if they use fluorescent and are not still using hot lights.

Spiral CFLs for interviews and small/indy videography I don't see as a problem though.

[Bob Grant]

Just to add a bit to what Richard is saying.
I've built fixtures using both Osram Lumilux 950 36W biax tubes which have a CRI >95 and their Studioline 55W tubes which have a CRI >85. This is pretty much an apples to apples comparison.

We had the odd complaint about the Studioline tubes from people shooting film. For those shooting video no one has ever noticed a difference and this is in line with Osrams own advice. If you're shooting film or lighting an art gallery then yes, the higher CRI is worthwhile. For video the Studioline tubes are optimised for video, they're slightly cheaper and give more lumens per watt of power.

One thing probably worth a mention too, if rolling your own fixtures using CFLs try to give them plenty of ventilation. The built in ballasts typically need all the cooling they can get else they don't last very long.

[Richard Andrewski]

It is so hard to see the difference in 8 points of CRI (from 82 to 90 or especially 85 to 90 like Bob said).

When we were engineering the Cool Lights tubes we manufacture here in China, I started with a CRI of 90 as my goal. I got a real education in fluorescent tubes beyond what I ever thought I would. What the engineers told me was that the lumen output drop between 85 and 90 is pretty drastic because to get beyond 82 to 85 CRI you are using a higher quality phosphor which cancels out most of the dreaded green spike everyone is always talking about. The problem with this is, it's the green component which contributes the most to the light output in a flo bulb. We experimented and found 87 to be about the optimum place where you still have good lumen output and color rendering. So we have what we call a CRI >87 bulb. So because of all this, I'm a bit dubious of claims about CRI 90 bulbs which have normal lumen ratings comparable to others of the same wattage. Particularly the CRI 95 ones I see occasionally. One trick that many do is overdrive the bulbs a bit with a higher wattage output from the ballast. Of course, this shortens life of the bulb but makes up a bit for the lack of green in the bulb which cripples it a bit.

[Richard Andrewski]

I just noticed Bob mentioned something about film. Film is a whole other animal consistent with what he said. Film will be far more sensitive to a green spike and more care has to be taken to use minus green or choose the best bulbs you can. Those shooting digital (video or stills) though, which I would say is the vast majority of members here, will find their cameras are very forgiving of green spikes.

Also, on the comment about giving them some room--that's best for sure. But one more reason to give them some room is that those spirals function best from a light giving standpoint that way too. If they're all crowded together, they simply don't radiate light as efficiently as if they're too close together as many of the spiral "panels" do in the name of keeping the unit compact. Especially if they have any reflectors beneath. The reflector does practically nothing if there isn't some room around the bulbs for the reflector to take some light and reflect it back out.

[Greg Boston]

Originally Posted by Richard Andrewski

Those shooting digital (video or stills) though, which I would say is the vast majority of members here, will find their cameras are very forgiving of green spikes.

What you're saying is very true. My XD HD camera, and some others I'm sure, have various preset color matrices, one of my choices being FL for fluorescent lighting. It's a custom table that counteracts the nasty green spike. Here's a frame grab showing the FL preset in a meeting room using only standard overhead fluorescent lighting. Gamma setting =2.2.

-gb-

[Richard Andrewski]

Aren't digital cameras great? I've only shot 35mm still film in my life, never movie film, nor do I ever want to. It's just so much more convenient from so many different angles to be shooting stills or moving images via digital. The white balance aspect alone is one of the biggest features. It should be interesting to see how the RED camera handles these type issues but my guess is that it will also be forgiving.

[Jim Froom]

Thanks for all the info. Nice to get a little education from people actually using this stuff.

Two more points/questions.

1. In regards to the CRI, I pulled this off of

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

in the middle of their description of CRI.

"Although an objective measure, the CRI has come under a fair bit of criticism in recent years as it does not always correlate well with the subjective color-rendering quality for real scenes, particularly for modern (e.g. fluorescent) lightsources with spikey emission spectra, or white LEDs. It is understood that the CIE is looking at developing newer color-rendering performance metrics." Any comments?

2. Color saturation aside, if I'm doing an interview inside a building or room, and am using sun light (direct or indirect) coming in from a windows or skylight plus light from either 5000k or 6500k fluorescents, are my issues going to be in color shifts that are going to cause problems or are they going to be more "artistic" or "subtle" in nature that few will find offensive or ugly. (mainly referring to skin tones and/or neutral colors)

[Richard Andrewski]

Originally Posted by Jim Froom

"Although an objective measure, the CRI has come under a fair bit of criticism in recent years as it does not always correlate well with the subjective color-rendering quality for real scenes, particularly for modern (e.g. fluorescent) lightsources with spikey emission spectra, or white LEDs. It is understood that the CIE is looking at developing newer color-rendering performance metrics." Any comments?

This is sticky and requires a lot of back information to get it. Take a look at this article I wrote on the subject and see if you still have questions. A lot of people talk about CRI but don't even have a real idea of how it's measured. I covered that in the article if your interested.

http://www.coollights.biz/wordpress/archives/13

Here is a short article on why most discharge lamps use mercury too:

http://www.coollights.biz/wordpress/archives/14

There is a lot of talk surrounding what exactly the new color rendering measurement system would look like that would be all-encompassing for everything from real sources like daylight or tungsten to the discharge types and LEDs. But there is also a lot of inertia around the subject because there is so much marketing material and infrastructure built around the old way of describing things. Also, factories that produce light bulbs of various kinds are the primary users of the very expensive test equipment that measures color temperature, correlated color temperature and CRI. They are not keen to have to reinvest in new test equipment so I can imagine it's going to be a long while before anything happens there.

Originally Posted by Jim Froom

2. Color saturation aside, if I'm doing an interview inside a building or room, and am using sun light (direct or indirect) coming in from a windows or skylight plus light from either 5000k or 6500k fluorescents, are my issues going to be in color shifts that are going to cause problems or are they going to be more "artistic" or "subtle" in nature that few will find offensive or ugly. (mainly referring to skin tones and/or neutral colors)

Depends on lots of things including your camera, white balance, etc. On some of the worst discharge lighting green spike offenders, you'll see more green in the picture when mixing sources of different color temperatures. It's better if it's all uniform as possible. If the flo's are in the range of CRI 82 or above though, you probably wouldn't notice so much difference.