Daylight low power china ball - Colour testing it

Peter Emery · June 29, 2007

This posting is about a bulb I just bought and a domestic way to colour check it

I've bought a photo balanced 105w 240V 5500K fluorescent bulb from Ebay with an ES screw. The idea is it could work in place of a photoflood. But run cooler, less power and at 5500K. You could even run a cluster of them as a large low power soft source. It's made by Miscansu.

It is supposed to kick out 525W tungsten equivalent. It actually gives about f2 500ASA at 3m through a china ball.

My gaffer said I should check the colour output

I took a shot of a colour/grey card with my pro digital camera. One lit from bulb and one outdoors.

Then I took them into Nikon Capture (my photography software) and using the mid grey off the card in the images I matched the grey point of each.

Then converted them to JPEG and imported them into AVID.

I put the images on top of each other in the time line. I then shunted between them and watched the vector scope for differences.

This is the result of the fluorescent compared to daylight

Green - bang on

Blue - bang on

Cyan - bang on

Yellow - very very slightly low

Magenta - very very slightly low

Red - very very slightly low

It's a really good result I think.

I'm not an expert of colour checking. I'd like to know if anyone can spot a flaw in my method.

Also has anyone tried these bulbs?

Peter

Richard Andrewski · June 30, 2007

Bulbs of that wattage and higher are available from a number of sources and have been in use by some for a while. I first started using them in 2005. Here's a 200w bulb for instance:

http://www.teksupply.com/webapp/wcs/stores...5054&isDoc=

Works with a mogul base so you need a mogul fixture. I've used them in a softbox without the silk on the front as its not really necessary unless you need to lose some light. I've also been using them in Chinese lanterns too for those times I need light at 360 degrees and need to diffuse it even further. Lots of ways to use them, but one thing though: the self-ballasted CFL models like this are never quite as efficient as those with separate ballasts. The lumens per watt of these models can be as low as 40 whereas a tube like a 55w PL lamp can get around 50 to 60 LPW with a great ballast driving them. Also, some CFLs are prone to flicker although thats less and less common. CRI is improving all the time too. Home Depot N:Visions 5500K are in the 80 to 85 range of CRI. That's becoming more and more common for household use bulbs.

I was going to sell the CFL spirals and 200w 8U units at one time on my site but haven't been able to find a manufacturer as reliable as the one making my 55w tubes. I still am actively searching that though and hope to offer these in 3200K as well as the more common daylight too when I'm sure the quality level will be what I like to offer.

In the meantime, I've really been impressed with the efficiency and versatility of metal halide / HMI and I think it offers even greater promise when some of the safety issues are addressed. Take a 150w metal halide bulb, relatively small, available in tungsten or daylight, at 85LPW and you've got a lot of hard light which can be left hard or softened up in all the normal diffusion ways. They really fulfill a lot of the promise of what I was originally looking for in the spiral and 8U CFLs. The light they produce is even relatively cool (temperature wise) although the bulb itself has to be in an enclosure as its intensely hot and emits a lot of UV.

JD Hartman · June 30, 2007

If these are the same as the spiral bulbs being sold on Ebay, marked Falcon Eye, the color rendition seems a little off. The skin tones were on the orange side in some video I saw, which was shot using them as sole light source. I don't think could have been the camera, makeup or any post processing. I saw the same person on the set, under the same lights and her flesh tones appeared orange.

Rodrigo Llano · July 1, 2007

Peter..

I haved a couple of questions in term of your procedure .

What was the original format of the picture?

Why did you choose JEPG to import in avid ? Why not a TIFF or other less compress format?

I feel with those details you could have a totally different image quality and readings.

Respectfully

Rodrigo

Chris Burke · July 1, 2007

This posting is about a bulb I just bought and a domestic way to colour check it

I've bought a photo balanced 105w 240V 5500K fluorescent bulb from Ebay with an ES screw. The idea is it could work in place of a photoflood. But run cooler, less power and at 5500K. You could even run a cluster of them as a large low power soft source. It's made by Miscansu.

It is supposed to kick out 525W tungsten equivalent. It actually gives about f2 500ASA at 3m through a china ball.

My gaffer said I should check the colour output

I took a shot of a colour/grey card with my pro digital camera. One lit from bulb and one outdoors.

Then I took them into Nikon Capture (my photography software) and using the mid grey off the card in the images I matched the grey point of each.

Then converted them to JPEG and imported them into AVID.

I put the images on top of each other in the time line. I then shunted between them and watched the vector scope for differences.

This is the result of the fluorescent compared to daylight

Green - bang on

Blue - bang on

Cyan - bang on

Yellow - very very slightly low

Magenta - very very slightly low

Red - very very slightly low

It's a really good result I think.

I'm not an expert of colour checking. I'd like to know if anyone can spot a flaw in my method.

Also has anyone tried these bulbs?

Peter

Not these particular bulbs, but ones just like it. What I have found out is that the CRI (color rendering index) number is the most import factor in getting "true" colors. the closer to 100 the CRI is, the better off you will be. I wouldn't trust a bulb that is lower than 90 CRI, because then you start to get a green cast. how did you rig this bulb into a china ball? I bought four similar bulbs and want to make a few china balls for it.

chris

Peter Emery · July 2, 2007

Just to answer various points that have come up.

My stills camera shoots NEF RAW (A Nikon Raw file) I used JPEG as I have to convert it to something. I guess a TIFF may have been a wiser option. I might try that. Although I'm not sure how much difference it would have made in practice.

The bulb was bought from a company called PH Miracle on Ebay. It's made by Miscansu

Checking the CRI perhaps is the way to go forwards. But how can I find this information out?

In terms of mounting it in a china ball. I just hung it from a good flex with a brass/ceramic ES lamp holder. If it all goes well I was thinking of building a triple mount and using a larger china ball. My Gaffer Bernie is talking about making a lighting tube using them; although I'm not sure how.

Peter

Richard Andrewski · July 3, 2007

Just to answer various points that have come up.

My stills camera shoots NEF RAW (A Nikon Raw file) I used JPEG as I have to convert it to something. I guess a TIFF may have been a wiser option. I might try that. Although I'm not sure how much difference it would have made in practice.

The bulb was bought from a company called PH Miracle on Ebay. It's made by Miscansu

Checking the CRI perhaps is the way to go forwards. But how can I find this information out?

In terms of mounting it in a china ball. I just hung it from a good flex with a brass/ceramic ES lamp holder. If it all goes well I was thinking of building a triple mount and using a larger china ball. My Gaffer Bernie is talking about making a lighting tube using them; although I'm not sure how.

Peter

Hi Peter,

I wrote an article about this subject a while back. A lot of people talk about CRI but don't understand how it is tested. It turns out, for now anyway, its mostly tested in factories with an "integrating sphere" and a spectroradiometric computer. I posted a picture of the testing apparatus. Unless your neighbor happens to have one or you have a light bulb factory somewhere nearby, it's not likely you can find one to use very easily as they run in the $20,000 range and higher ;-).

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

It turns out that the present methods used by lamp manufacturers to test CRI are really felt to be obsoleted by many modern discharge lamp technologies such as metal halide and fluorescent. Of course other sources like LEDs come into the question too now. Here's a link to some research on the subject:

http://www.knt.vein.hu/staff/schandaj/SJCV-Publ-2005/502.pdf

As mentioned in that article, the poor man's version of all this is a Macbeth Color Chart. Some examples of the Macbeth in use:

http://www.akdart.com/macbeth.html

http://www.akdart.com/mcc.html

http://www.mambo.net/as/view/113

I posted a picture below of the integrating sphere I have access to at my bulb factory here in China. We use this to test batches of all our lamps before they go out as do most all reputable lamp makers. It gives us a spectral energy distribution chart which we post for information purposes on our site along with a lamp test report which includes CRI value as tested.