DirtFarmer Loc: Escaped from the NYC area, back to MA

Flourescent lighting has been a problem for many photographers over the years. I wanted to start a discussion on the subject with some observations. Feel free to talk about your problems with fluorescent lighting (not including spelling) and your solutions.

Black body light sources

The sun, which produces our daylight, is roughly a black body source. That just means it looks like a body at a certain temperature. The apparent temperature of the sun is around 5500 degrees Kelvin. You all know that if you heat something up enough it will glow red. If you heat it up more it will get orange, then yellow. Red heat is around 900 degrees Kelvin. Yellow is around 3000 degrees Kelvin. The sun is very hot. At low temperatures, a body will emit light, but mostly in the infrared. When it starts to glow red, that’s the first indication of visible light. Light from the sun peaks in the green, which is very hot (near 9000 degrees Fahrenheit for you guys that don’t think in Celcius). The solar spectrum is not purely a black body curve since there are elements in the sun’s atmosphere that absorb light at certain wavelengths. That modifies the spectrum but in a broad sense it looks pretty much like a black body and it takes a spectrometer with good resolution to see the absorption lines, known as Fraunhofer lines, after the guy that discovered them. In the chart below the colors represent the visible part of the spectrum (horizontal axis) and everything to the right is infrared.


Source: Wikipedia article, image By Darth Kule - Own work, Public Domain, commons.wikimedia.org/w/index.php?curid=10555337

An incandescent light source (regular light bulb) has a spectrum similar to the 3000 degree Kelvin curve above. If you look at the 3000K line and the 5000K line in the above chart you can see that although the 5000K curve has about 8 units at the blue side of the curve and 11 units at the red side, the 3000K line has about 1 unit at the red side and very close to zero at the blue side. That’s why there’s a difference between the white balance for daylight and incandescent lighting. The light bulb will look redder because there’s more energy at the red side of the spectrum (relative to the total energy).

Fluorescent lights

Fluorescent lamps generate light by a two-step process. The lamps have changed in recent years but the basic process is similar. A fluorescent bulb is a glass tube with a gas in it at low pressure, and a phosphor is coated onto the inside of the tube. A high voltage is applied to the ends of the tube, generating an electrical discharge in the gas. Energy is transferred to the gas, which then emits the energy in the form of light. The gas is chosen to have a lot of energy emitted in the UV region. Mercury was used in early bulbs but I suspect they have been substituting other gases as they try to phase out uses of mercury. The UV light hits the phosphor on the inside of the tube, which absorbs the energy and re-emits the energy at a lower (visible) wavelength. Many types of glass absorb UV light so little of it is emitted by the fluorescent bulb (but you do get some).

The discharge in the gas is electrical, and is run on AC at the normal line frequency (50-60 Hz). Since AC goes in both directions, the discharge is generated at twice that frequency (100-120 Hz). Newer fluorescent lamps use a higher frequency which alleviates some of the problems described below.
The light from the gas discharge occurs at a group of wavelengths specific to the type of gas used. The phosphor converts that to a more continuous spectrum, but the shape of the spectrum will depend on the phosphor used. Generally a mixture of phosphors is used to refine the shape of the output spectrum. Fluorescent lights are available in a number of different spectral distributions, for example “daylight”, “cool white”, and “warm white”, as well as some specialty distribution such as growing lamps, which match the spectrum needed for plant photosynthesis, generally heavy in the red and blue and light in the green (which is reflected by the plant leaves).

Since the gas discharge turns on and off driven by the line voltage, the UV light is intermittent. The phosphors emit light after the UV light turns off but the intensity of the emitted light will diminish exponentially with time until the voltage rises and the UV light turns on again. The decay rate of the phosphor depends on the particular phosphor used, and since phosphors are generally a mixture of materials, the spectrum will change in a complex way with time. Every different material has a different decay rate for the emitted light

I found a couple examples of fluorescent light spectra online, presented below. As noted above, these are not static spectra, but most likely averages over an integral number of power line cycles. The broad curves are the spectra due to the phosphors and the spikes are the spectra due to the gas discharge. There is clearly a wide variation in the spectra you will get from these light sources. Just to remind you, UV light has wavelengths shorter than around 400 nanometers and visible light is roughly between 400 and 700 nanometers. Above 700 nanometers is infrared.


Sources: www.comsol.com/blogs/calculating-the-emission-spectra-from-common-light-sources/
en.wikipedia.org/wiki/File:Spectrum_of_halophosphate_type_fluorescent_bulb_(f30t12_ww_rs).png

To see the time variation of the spectrum I took my camera and took several shots of a fluorescent lamp with the shutter set at 1/1000 second. My camera has a focal plane shutter so the actual exposure takes about 1/250 second, which is the time taken for the shutter to fully transit the frame. The image below is five such shots cropped to show just the fluorescent bulb and placed side by side. The travel of the shutter is along the length of the fluorescent bulb so what you see is the variation of the light coming from the bulb over that 1/250 second. The timing of the shots is random compared to the phase of the AC line so the shots will show the variation at different starting points. At the left of the illustration, the bulb is illuminated for the whole 1/250 second of the exposure. Just to the right of that, the gas discharge is on at the start of the exposure (bottom of the frame) and turns off just after the exposure starts so you can see the light from the phosphors start to decay. In the center the decay is further along and the light starts to look yellowish. Then there’s a wide yellow band in the center and the gas discharge starts up at the top of the frame. Last image (right side) shows the gas discharge off (bottom of the frame), then it turns on so the top of the frame is fully illuminated. The diffuse band between off and on is partly the time it takes for the discharge to get to full intensity and partly due to the width of the shutter opening. The shots are not sequential since they are at a random phase relative to the power line. I just arranged them to look sequential.



Since there is a variability of the effective color temperature of the illumination with fluorescent lighting, different shots will need different adjustments for white balance, particularly if the shutter speed is shorter than half the time between power line cycles. The effect will be lower for longer shutter speeds but unless the shutter speed is not an integral number times the time between power line cycles there will be some effect.
As mentioned above, some of the newer fluorescent systems use a higher frequency for the discharge so the phosphor decay will not be as important and the variability of the white balance between shots will be reduced.

LED light sources

While LED lights have less of the spiky nature of fluorescent sources, they are not really like a black body source. They do have some structure, and the structure will depend on the blend of different LED sources in a lamp.


Source: electronics.stackexchange.com/questions/149683/do-standard-white-leds-produce-a-full-spectrum-of-light

The color of a LED depends on the materials used to construct the LED. Many LEDs are combinations of different color LED components, so the spectrum shown above is not representative of all LED light sources.

timcc Loc: Virginia

Thank you for sharing this detailed explanation -- very informative.

therwol Loc: USA

I copy some of my negatives and slides with my Nikon D810 over a 6000K fluorescent light box. In Live View, there is a visible shimmer or flicker on the LCD screen, also seen when I tether the camera to a computer. I pretty much know what this means. I crank the ISO all the way down to 64 and shoot at around f/10 so that I get a shutter speed of usually 1/4 to 1/8 of a second. I don't see any variation in exposure or color at these long exposure times.

I'm not sure that fluorescent was the best choice. I haven't tried anything else for this. I though that a cheap LED light box might cause color issues. I know that this has been worked out for more expensive LED lighting, but that didn't seem cost effective.

Gene51 Loc: Yonkers, NY, now in LSD (LowerSlowerDelaware)

I've had good results with the anti-flicker feature on my Nikon DSLRs.

Haydon

I don't pretend to know the inner workings of lighting specifically as you've detailed DirtFarmer but I mostly shoot in a controlled studio lighting environment using monolights. You are welcome to correct me and I will take no offense but from what I've read both LED & florescent lighting are considered discontinuous lighting missing key parts of the light spectrum which to many professionals are a shortcoming. Despite improvements in these light sources, they do not perform apparently as well as studio strobes.

That shortcoming is of little consequence to me. I'll openly admit I'm colorblind and would not recognize this alleged inadequacy. Peter Hurley uses florescent lighting and is the top headshot guy in New York bringing in over $1000 per sitting. Joel Grimes confesses he's colorblind but is internationally known for his work bringing in top commercial contracts for the last 40 years. It certainly hasn't hurt either one of these photographers with their use and or visual inadequacies.

So my statement is nutshelled to, work with whatever works for you. The biggest shortcoming one has is their own vision and talent and can't be rooted to equipment or their own perceived handicaps.

dpullum Loc: Tampa Florida

Excellent educational post, UHH needs a searchable file specific to Education and Tutorials.

Walmart has a 4' florescent LED replacement for $21 ... 5000 Lumens.... I could not find temp specification but one review said: "looks to be about 4500k to 6000k". Useful for Photography??? perhaps.

Halloween, my favorite holiday, is the "Black Light" prime time. I bought a 4' fluorescent tube also on eBay I bought a 100 LED UV flashlight for $15. A walk at night using the flashlight changes your view of the glowing world.

Bipod

Very informative post. Raising awareness of color rendering problems caused
by lighting is a beneficial for photography. Color temperature is easy to fix in
post processing, but color rendering problems are usually impossible to fix.

A great way to undrestand the importance of lights in color rendering is to try
taking photographs by a sodium vapor streetlight; roses are black, violets
are gray, and sugar is yellow.

When buying lights, photographers have to rely on color rendering data provided
by the manufacturer. Unfortunately, it is impossible to convey spectral data as
a single number. And there is little independent testing of photographic lights.

The most common measure, the Color Rendering Index (CRI), is less than ideal.
It is based on 24 colors, which is rather small sample. Lights can have a high CRI,
but still not work well for photography.

The Extended CRI rating (ECRI) adds more colors. But few light manufacturers
give a ECRI rating.

One problem with fluorescent lights is that the formulation of the phosphor can
change without notice.

"White LED" is a generic part in the electronics industry. It's often hard to tell
who made them, or whether this batch is different from the last batch you bought.
Most LEDs are not marked -- you have to rely on the markings on the box the
shipment came in.

LED lights intended for photographic use often contain two more more different
LEDs to increase the number of phosphors. A certain combination of LEDs
may work well -- but the next shipment of the "same" part may not.

Given how difficult it is to assess lights, I prefer to use only continuous spectrum
lighting: xenon flash and the types of incandescent continuous lighting that are
closest to daylight (i.e., "quartz halogen"). But even household incandescent
are far better than fluorescent lights.

We all hate the complexity of flash and the annoyance of hot continuous lights.
But all worthwhile pursuits have inconveniences. Technology is not a
panacea and is limited by physics. The universe wasn't designed to make
phtoography easy.

larryepage Loc: North Texas area

The big problem with fluorescent lamps is that the phosphorescent process is destructive to the coating inside the tube. For most designs, output will have dropped to about half of the initial intensity by the time the lamp finally fails, and color temperature is constantly changing at least a little during this decay process. The exact characteristics of the decay vary and are not generally predictable, at least not completely.

fourlocks Loc: Londonderry, NH

Interesting post; lots of good information. A minor point on fluorescent lamps: It is my understanding the two step process is that when power it applied, the mercury vaporizes and emits ultraviolet (UV) radiation. That UV radiation then hits the phosphors on the inside of the tube causing them to fluoresce and give off visible light. A fluorescent lamp will not work without mercury. If the fluorescent coating is left off the inside of the tube, only ultraviolet light is emitted and you end up with the black light so familiar to us baby boomers back in the late 60's and '70's.

DirtFarmer Loc: Escaped from the NYC area, back to MA

Basically right. It's not the vaporization of the mercury that produces the UV, but the electrical discharge through the mercury vapor. Of course the vaporization is necessary to produce the mercury vapor to produce the discharge. As I recall, black lights were like fluorescent lights, but without the phosphor coating and with the addition of a filter to remove visible light, thereby producing the term "black light".

Any gas that produces significant amounts of UV will work for the fluorescent tube. Mercury is a really good UV producer so that is what has been used. But since there's a lot of effort to remove mercury from the environment, substitute gases are being considered for the discharge. I am not really familiar enough with the research to know just which gases are used but if fluorescent lighting survives technology advances, the mercury will most likely be phased out.

DirtFarmer Loc: Escaped from the NYC area, back to MA

I suspect the photographers who use fluorescent lighting are using the newer types with faster cycling times (maybe 1 KHz or higher). That would mitigate the phosphor decay variations and give more consistent lighting. Since the light from the fluorescent bulb is not like black body radiation (sunlight or incandescent light) some adjustment in postprocessing will be needed to balance out the difference. Once that balance is found, and if the light is consistent, the fluorescent lighting is capable of producing good white balance.

After all, the monitor you are looking at right now uses three colors to produce the entire range of colors your eye perceives, so having a few bumps in the spectrum is something that can be corrected. Of course the response of your eye is a whole 'nother subject.

Fotoartist Loc: Detroit, Michigan

Very interesting. I'm setting up a table top Black light project now.
I have various neon, day-glo, and fluorescent paint colors (not sure if there is a difference but they all glow under black light) but where do you get fluorescent white paint as a base coat?

I've also noticed Epson matte white printing paper fluoresces.

Bill P

I've also noticed Epson matte white printing paper fluoresces.[/quote]


That means it contains something called optical brighteners. Many papers have them. The goal is to produce brighter whites. And yes, it fades with time.

larryepage Loc: North Texas area

Cotton fibers also fluoresce blue. Most archival paper and board contains a significant amount of cotton fiber.

Architect1776 Loc: In my mind

Hopefully by now most all intelligent businesses and commercial locations have dumped Florescents by now. All new construction if properly designed does not have them at all anywhere.
So this should not be an issue going forward except for backward places wanting to just throw money away on old technology and wasteful lighting.