The Importance of Choosing High-Quality CFL Bulbs for Continuous Light Shooting


Anyone interested in shooting photos or videos using continuous light without breaking the bank has likely looked at purchasing or building a lighting setup based around CFL bulbs. Earlier in the year, I decided to go this route myself because I film interviews and generally prefer the way that soft continuous light reacts with a subject’s pupils and comfort levels for photography.

To start with, I purchased a two-softbox light kit from “Limo Studios” on Amazon. This included eight 45W 6500K CFL bulbs listed as “daylight balanced.” I was aware that 6500K was a little cold, but I’ve spent enough time messing with white balance that I didn’t think this would be a problem.

What I didn’t realize at the time was that these bulbs do not have a high enough CRI (Color Rendering Index) for people. Amazon didn’t list the CRI, but there were plenty of positive reviews, so I pulled the trigger and tried it out.


In every shoot I had problems with skin tone and color. In some conditions without enough ambient light I could bring back some tone by dropping the exposure, increasing the shadows, and messing with the individual color saturation points but it was always sketch. The low CRI completely killed lip tones, turned skin pale, and created shadows around the eyes. Every time I found an excuse for it, mostly blaming white balance (even shooting gray cards).

On my last shoot I finally just gave up and shot with strobes, then went home determined to figure out what I was doing wrong. I learned about CRI for the first time and the importance of it, so I turned around and immediately bought a new set of CFL bulbs with a CRI of 91.

The next morning I set up a single softbox and took some images while switching between the crappy bulbs and the good bulbs. The improvement in color and quality was immediately apparent. The old bulbs are going in a box for use around the house — they’re just no good for photography/videography.

Here are some examples showing the difference, straight out of the camera with auto white balance (and slight variance in exposure since I was doing it manually and the Alzo bulbs put out more light):

Example 1: Note the lips and cheeks

Example 1: Note the lips and cheeks

Example 2:  When the poor CFL bulbs are closer and brighter the difference isn't as stark, but it's still very clear.

Example 2: When the poor CFL bulbs are closer and brighter the difference isn’t as stark, but it’s still very clear.

Example 3: Shot with gray cards. Neither of these bulbs resolve a perfect neutral gray using the gray cards, so white balance is better set by hand.

Example 3: Shot with gray cards. Neither of these bulbs resolve a perfect neutral gray using the gray cards, so white balance is better set by hand.

It’s easy to blame white balance, but you can see that the gray cards fall in very close to the same color, while the skin looks completely different. That’s my point (look around the eyes and lips especially).

Next stop is exposure, but when you compare exposure with the histograms, the average adjustment to match exposure was either increasing one set by +.1 or decreasing the other by -.1. So yes, there was some exposure variance but it’s not what’s causing the disgusting effect.




The problem, in the end, is the quality of your CFLs. So what’s the lesson here? Buy better quality CFL bulbs, and pay special attention to the CRI. It matters a lot!

About the author: Pete Waterman is a photographer, adventurer, technologist, and writer who lives in the Washington D.C. Metro Area. You can find him on his website, LinkedIn, Facebook and Flickr. This article originally appeared here.

  • Sincerely Want To Know

    Which example is the good one; left or right? I am honestly not trying to troll but want to make sure I am aligned with the author.

  • harumph

    If you can’t tell, then you need to calibrate your monitor. Left is the good one, except in the 5th set down.

  • Sincerely Want To Know

    Thanks, Harumph, that is what I was thinking but the fifth one down threw me. I still maintain the author should label good v. bad, though.

  • RafaƂ Drumla Nitychoruk

    achieving “perfect neutral gray using the gray cards” is impossible… gray card is for lightmeter not a color balance…

    if you want to correct colour perfectly, you need to use color cards like Colorchecher Pasport which solves metamerism problem…

    still you are absolutely right about light quality – it is worth to know how tight is (as example) LEDs light spectrum…

  • Zos Xavius

    This is true actually. I have a grey card I use and it helps, but does not give perfect WB.

  • Dr Hob

    You lost me at “Auto White Balance”

  • LostandConfused

    You say “the alzo bulbs put out more light” but you never distinguish if the Alzo lamps are the good ones or the bad ones…

  • markz

    Been there done that though as I shoot for B&W conversion (or B& film in medium and large format) I put off purchasing new CFL’s until just now.
    one thing that became obvious, even with B&W shooting is the huge gaps in the spectrum of cheap CFLs using colour filters to change the rendering tone values showed there was almost zero in some red and orange areas of the spectrum.

  • Jon Woodbury

    I thought the same thing. The shots should have been shot on a consistent white balance for the comparison to be valid.

  • matt jones

    he has the disappointed look on his face on the shots with the good bulb, good way to confuse people, if that’s what he is trying to do.

  • SpaceMan

    You lost me at “film”

  • ffs gah!

    thanks for not telling us what the ‘good’ lights were. would hate to actually learn something from your article.

  • markz

    that’s OK not every one who looks can see… or reads can understand. – medium format digital is expensive, very expensive, large format almost non existent (and the few instances of digital LF are usually both cumbersome and priced around the “small house” budget) so at those resolutions film is the an incredibly cheap option, by several orders of magnitude.(it’s also classy, and needs skill somethings in short supply these days)

  • Peter Waterman

    This seems to be what provokes the biggest confusion. The entire point is that the different quality lights produce images which look different even when properly “white balanced” (as shown by the comparison image with the gray cards pasted together to show they are almost exactly the same). The bluer lower CRI bulbs need significant tint/hue adjustments beyond white balance to look normal.

    Also, many people seem to think cameras are stuck in the 1990’s when it comes to AWB, but with a clean white background my 5d3 sets a visually better white balance on auto then Lightroom does from a gray card. The whole point of shooting RAW is I can choose the balance I want though.

  • Peter Waterman

    The cheaper ones are the Limo Studios bulbs, as mentioned earlier in the article. Cheers!

  • Peter Waterman

    The fifth one down was added in later in an edit and I swapped the two sides to make sure the gray cards lined up while comparing exposure and balance.

    To me the good vs bad is so completely obvious I didn’t think it needed to be labeled!

  • Peter Waterman

    I just woke up and took a bunch of random shots of myself in the living room while my gf looked on and made fun of me. There was no intent behind any of the expressions because I didn’t draw conclusions until after the test.

    I’ll keep that in mind for next time though. ;)

  • Rabi Abonour

    Quick Google search will find you high-CRI CFLs. While I agree it wouldn’t have hurt to mention the brand, the more salient point of the article is that there is more to light than temperature, which is something people often forget.

  • Lance

    Lolz, so now I know how to get photos that look like mugshots–get 6500k bulbs with a low CRI. :D

  • Eugene Chok

    it was clear, as was what you wrote, good job man

  • markvturner

    Where do you get CFLs with a CRI above 90? Every single one I found has a CRI below 90.

  • an engineer

    I’m an engineer who actually works on writing imaging standards. A word of warning: the CRI metric is highly susceptible to gaming. Knowing how CRI is calculated, I could easily design a light that looks terrible but gets a CRI > 90.

    At this point, I don’t suspect the lighting manufacturers have done so (though I have no proof) because the high CRI market is so small, but it may someday go the way of megapixels (cameras) or dynamic contrast ratio (TVs) – gamed by marketing until it doesn’t tell you anything anymore. Basically, it’s a good indicator, but doesn’t tell the whole story.

  • Pontiaku

    I’m curious if a home made spectrometer would clearly show a difference between these CFL’s.