Flourescent studio lights?

[oleleclos]

Has anyone on CiC used flourescent daylight studio lights with digital cameras, and if so what are your experiences/recommendations?

I have Bowens studio flash but it's heavy and cumbersome and overpowered for what I do now, so I wonder if these newfangled daylight bulbs are any good. I have found some heads with Bowens bayonets which would let me use my existing reflectors, softboxes etc. which sounds tempting.

I'm well aware of the pros and cons of flash and continuous lighting and have used both extensively; what I am looking for here is advice specifically on flourescent "bulbs" which I have never used.

Thanks in advance.

[Inkanyezi]

I am unaware whether there really are any screw-in bulbs that are good enough. I have built fluorescent studio lamps with straightt tubes and with compact fluorescent tubes. I have also used "low-energy" screw-in bulbs for product shots, and they can be used if colour is not critical and at long exposure times, 1/30 and longer. All screw-in lamps modulate their output by twice the mains frequency, a cycle of 10 ms in Europe.

The standard FL tubes that work best are daylight type with Ra 90+ colour rendition, for example colour 954. There are also special tubes as Osram Studioline, which have excellent colour rendition and also maintain colour properties when dimmed, but they are extremely expensive.

The critical property when using fluorescents is temperature. The FL lamp must attain a temperature ov between 40 and 45 degrees centigrade in its coolest point for proper colour and light output. A tube that is cooler or hotter will have less output and will also have a colour drift, mostly toward green if hot and other colours if cool. Therefore ventilation of fixtures is important.

For screw-in bulbs, the modulation has a depth of about 30%, and they cause uneven lighting with too short shutter times. I recommend at least three full cycles of the modulation, and shutter times under one cycle (1/100 in EU, 1/120 in USA) will inevitably show banding. Other lamps than screw-in should have sufficiently smoothed HF ballasts for proper use, and then any shutter time can be used.

[rpcrowe]

I have a four CFL daylight bulb soft box that I use for video shooting. One day I had it set up near where I was shoting a rescue Maltese and decided to try in in conjunction with my bounced 430EX and Joe Demb Flash Diffuser Pro.

I set the softbox almost directly over the dog at camera left. The lighting turned out pretty nice despite the mixing of light sources. I determined the white balance by shooting in RAW and including a WhiBal card.



However, using just the CFL bulb softbox in portrait work is somewhat problematic because it is difficult to get enough light for a decent shutter speed and f/stop...

[oleleclos]

Thanks guys for the helpful input (and the very nice muttshot). I'll give it a try; it's for small scale still lives, not portraits (not even of dogs), so exposure times are not a concern.

The Rocwing head* with Bowens bayonet seems a simple way to try this as it'll fit with the kit I have, but I don't think it has a cooling fan to keep the bulb and thus the colour temperature in check. I'll report back here when I have tried it.

*) http://rocwing.co.uk/epages/eshop138.../Products/1470

[Inkanyezi]

Usually, free air flow is sufficient to keep temperature at bay. The "cool spot" on a screw-in type of lamp is the one that is farthest from the electronics, hence the lamp "top", and the hottest part always is where electrodes are. The cool spot temperature regulates mercury gas pressure, which in turn governs light output and colour.

Screw-in types should preferably not be pointed upwards, which may cause them to be overly hot. The best burning position of a screw-in lamp is pointing downward. Compact fluorescent lamps, CFL, indoors should be pointed down when feasible.

A fan is usually not needed, and if shooting video, fans should always be avoided as they emit sound. Fan cooling might render the tube too cool, making it less efficient.

[rpcrowe]

This is basically the type of four CFL bulb softbox that I use.

http://www.ebay.com/itm/JS-Continuou...item2a1ddff1fc

I bought the softbox alone (from another seller) and I don't remember how much I paid for it but it was really inexpensive. I purchased the four daylight CFL bulbs at Home Depot and they cost me more than the softbox unit did. It provides nice soft light and a couple would be quite good for general still life setups. The only thing that it would not be capable of is any type of spot lighting. You would need another type unit for this.

These free videos have a lot of information on fabricating inexpensive set ups for product and still life photography...

http://www.prophotolife.com/video-library/

[oleleclos]

Thanks Richard and Urban. What about colours? One issue with flourescents is that it's not a continous spectrum, so I am interested in how that affects digital sensors.

I'm one of those people who is affected by the line spectrum (and my wife even more so) it and we therefore avoid energy saving bulbs as much as possible, but that doesn't necessarily translate to photography. Richard's dog portrait suggests it doesn't.

[Cparoyan]

I'm just starting out and don't have a lot of experience but I have a Canon Rebel and shoot with fluorescent day light balanced bulbs. I think they work great! I work at a production company so I often use the Kino Flo kits with the 56K lights that we use here for video. I have also purchased a very inexpensive kit from Cowboy Studios which are the large screw in type bulbs that come in softboxes. Again, I'm knew to this so I don't recall what causes this issue, but with the cheaper lights you can get some banding effects on the background. I can't remember if it was when my shutter speed was too slow or too fast. There is little to no issue with the Kino Flo kits, but those are also about $1k each. My monitor is calibrated and I find from looking at the raw images the color accuracy is pretty dead on. Sometimes you have to adjust the magenta slightly but that's all. I really like them since they don't get hot and have a nice even color. You also need a faster lens than with the standard hot lights.

[Colin Southern]

You also need a faster lens than with the standard hot lights.

With my studio strobes I'm usually shooting at F11 - and for tight head and shoulders shots, it's needed for sufficient DoF. Personally, I'd be concerned about lack of DoF if I had to use fast glass to get sufficient shutterspeed when using continuous lighting. Pupil constriction would also be a problem.

[Cparoyan]

https://www.facebook.com/photo.php?f...type=1&theater

I'm not sure if I am adding this picture correctly but here is an example. I think the DOF is good and minimal pupil constriction.

[ajohnw]

I looked around on this subject but aiming at getting the correct lighting for viewing a monitor semi serious but I need to do something with the room lighting I work in anyway. The following may be of interest

The cri of the fluorescent gasworks type bulbs on ebay tends to be around 80 which is rather low really. Lower CRI usually means dips in the reflection of certain colours of light as far as taking photo's go. It can be very marked.

Many companies are describing tri phosphor tubes as multi phosphor - the tri phosphor can't reach cri's of 90+. I was interested in d5000 lighting. These are what ISO recommend for graphics works. This is because D6500 is extremely difficult to achieve with a decent cri. Impossible I suspect

Led's are worse than tubes. There is some misleading comments about on hi cri led's too. Good skin tones maybe but not anything else. The low colour temperature power leds have a more even spectrum but still nothing like sunlight. No surprise to me as I know that r g g b leds have been produced for illumination purposes and they improve the rendition of a few colours.

There are some filter halogen lights about that do have a "sunlight" type spectrum but I haven't checked if this is actually correct. They are mentioned for photography and SAD.

Last but not least and filament bulb has a cri of 100%. Then it's just a case of setting the colour temperature correctly.

The last one really makes me wonder. I am sitting in a room with a cool white strip light. Should be about 5000K or so. My colourimeter reads it as 3,500K as it doesn't have an even spectrum. I also wonder about certain types of lights used in many shop to give good colour rendition but have no data on them,

-

[Colin Southern]

https://www.facebook.com/photo.php?f...type=1&theater

I'm not sure if I am adding this picture correctly but here is an example. I think the DOF is good and minimal pupil constriction.

It's a bit hard to tell because it's not sharpened optimally -- but it definitely "makes it easier" when the eyes are on the same focal plane. Would be a different story with shots like this cover shot of the singer Yulia though (actually shot at F22) (although not for DoF reasons)



As a "case in point", if I take a typical studio setup of 70-200mm lens @ 140mm - shoot at 1.8m - at F3.5 I get a DoF of only 3.2cm. At F11 I get 10.4cm. The other issue with fast glass due to low lighting is that ambient light (in this case meaning light other than the continuous lights) (eg from windows etc) can start to creep in and thus giving a mixed colour temperature situation.

[Cparoyan]

It's a bit hard to tell because it's not sharpened optimally

How would I go about sharpening the image? Again, I'm new to this so any input is greatly appreciated. I didn't even know until recently that photos are supposed to be sharpened on output. I'm teaching myself so there's a lot of info I'm missing.

[Colin Southern]

How would I go about sharpening the image? Again, I'm new to this so any input is greatly appreciated. I didn't even know until recently that photos are supposed to be sharpened on output. I'm teaching myself so there's a lot of info I'm missing.

Hi Cheryl,

It's the stuff "books are written about" (so hard to articulate in just a few sentences), but for an image like yours, standard settings would work pretty well ... so ... assuming a RAW capture then:

Capture sharpening of 0.3 @ 300%

Content / creative sharpening of 4px @ 40%

(both of the above on the full resolution image)

and then (after down-sampling) ...

Output sharpening in the region of 0.3px @ 50 t0 100%

I've written a bit about it in the past - you might find this helpful ...

Sharpening and Noise Reduction Sequence

Having just said all that, portraiture requires a lot of additional post-processing around the eye area - including additional localised sharpening.

Does that help?

[Inkanyezi]

The cri of the fluorescent gasworks type bulbs on ebay tends to be around 80 which is rather low really. Lower CRI usually means dips in the reflection of certain colours of light as far as taking photo's go. It can be very marked.

Colour rendering index, expressed as Ra[nn+] is often misunderstood. Those rendering indexes follow two different standards, and both of them are perceptive, not measured.

Light sources with colour temperatures of 5000 K and above are estimated according to a "daylight" standard, which does not have an even spectrum. The standard daylight accentuates a few colours that contain blue, as well as green, and are poorer in yellow and red than the other standard.

Light sources below 5000 K are estimated against a "black body radiation" standard, which is the explanation why incandescent bulbs generally score as Ra 99 or better, 100 being the limit. So any unfiltered incandescent lamp, whether halogen or regular household lamp, will have a CRI close to 100. The incandescent lamp is substantially stronger in the red and yellow parts of the spectrum, and its spectral curve is smooth, much more even than the daylight spectrum.

The reasons why daylight does not have a spectrum similar to the black body radiation is that the sun is a nuclear reactor, and its light emission is not purely from heat, but also from the nuclear reaction that causes ionisation of its materia, as well as filtration through the atmosphere, that also affects the spectrum received at the surface of our planet. Sunlight hence, although it has a complete spectrum with no evident modulation, does not have an even spectrum, not even smooth. It has many dips, and it has a top at either end, both in the red and in the blue parts of the spectrum. Daylight has considerably more blue than a black body radiator.

Many companies are describing tri phosphor tubes as multi phosphor - the tri phosphor can't reach cri's of 90+. I was interested in d5000 lighting. These are what ISO recommend for graphics works. This is because D6500 is extremely difficult to achieve with a decent cri. Impossible I suspect

Not impossible. There are such fluorescent lamps, although the standard lamp that is closest to daylight is colour 954. The colour code first denotes CRI, and 9 means 90+, which typically is Ra 97. The two last figures are colour temperature, and it is in hundreds of K, hence 954 is a 5400 K lamp with Ra 90+ CRI, typically 97. Those lamps may be mixed with daylight with no apparent difference.

Led's are worse than tubes. There is some misleading comments about on hi cri led's too. Good skin tones maybe but not anything else. The low colour temperature power leds have a more even spectrum but still nothing like sunlight. No surprise to me as I know that r g g b leds have been produced for illumination purposes and they improve the rendition of a few colours.

The above doesn't hold completely true nowadays, and there are continuous improvements of the LED:s. However unlike FL lamps, LED usually is not declared against any standard, so you cannot know beforehand when you buy a bulb whether it is a "full spectrum" bulb or just the old "twilight" type. My label "twilight" is hackish jargon for a two-colour lamp. The first white LED were a blue LED with yellow fluorescence, which lacked red more or less completely.

There are some filter halogen lights about that do have a "sunlight" type spectrum but I haven't checked if this is actually correct. They are mentioned for photography and SAD.

This has a technical explanation, and there are also other incandescent bulbs with the same type of filtering. The filter is a parametric filter coating of neodymium oxide, which looks a bit violet to the eye. The result of filtering is that the spectrum is more like the daylight spectrum, hence less even. Those lamps render more lustre to colours, because when you adjust for the colour temperature, the image will look like exposed with daylight. Because the lamp therefore does not comply with the black body radiance standard, its CRI decreases to about Ra 70, but its effect for colour photography is superior to regular incandescent bulbs with CRI of Ra 99, as long as you want a neutral colour rendition that does not resemble incandescent. In fact they are very good for colour photography, particularly product shots, flowers and the like, and they are also very good for showing the colours of fabrics. Unfiltered incandescent lamps however are better for presenting the "mood" of incandescent light.

Last but not least and filament bulb has a cri of 100%. Then it's just a case of setting the colour temperature correctly.

And the correct setting for incandescent bulbs mostly is about 200 higher than the K of the bulb, if you want the perceived colour with about correct metamerism. They will however never render violet tones well, no matter how much you correct in PP.

The clue is that the incandescent lamp does not have a spectrum that closely resembles that of daylight, so it will never give a colour rendition close to what you get with daylight, and it is really not as simple as setting a K value in your camera or PP.

The last one really makes me wonder. I am sitting in a room with a cool white strip light. Should be about 5000K or so. My colourimeter reads it as 3,500K as it doesn't have an even spectrum. I also wonder about certain types of lights used in many shop to give good colour rendition but have no data on them,

And the explanation is simple. An even spectrum is not actually needed for good colour rendition, because we have a medium that registers three, and only three colours. Our eyes too do so, at least when we are over ten years of age. So we are sensitive mainly to three spectrum colours, and our cameras are tuned to the same colours with the filter array in front of the sensor, with one coloured filter for each sensel. Then our presentation media as well, so your CRT or LED or LCD screen will display just those three colours and nothing else. If we have three lamps with those colours, even monochromatic, we will perceive the image as if illuminated with white light. Not truly though, but not far from. So FL lamps have most of their emission in those three bands, which makes them more efficient, and they will render good colour if the mix is reasonable. Their colour temperature is only a matter of the balance between red, green and blue.

As a sidetrack, children, and also cataract-operated people, are also sensitive to ultraviolet, which is perceived as vibrance or clarity. We lose that ability when growing older, as the lens in the eye becomes yellow and filters the UV radiation, but after cataract operation, when the lens is removed and replaced with a PMMA plastic lens, UV perception is restored.

I think there is reason to explain light modulation as well.
When a lamp is supplied with alternating current, it emits more light at the top of the current curve and less when the current goes through zero. For FL lamps this modulation usually has a depth of about 30% when driven through a magnetic ballast, and when they have an electronic ballast, modulation may become even deeper, due to poor smoothing. That is why you may perceive a flicker from screw-in bulbs, and LED screw-in bulbs are even worse in this respect than FL lamps. At fast shutter times, when the curtains travel simultaneously over the sensor with only a slot open, modulated light will cause banding.

When shooting video, modulation is very evident, and if you use an electronic viewfinder, you will see the same effect in the viewfinder or on the screen of your camera, that light varies with time. This is the reason why you should choose a ballast that has very good output smoothing from the rectifier at the supply side, because poor smoothing will cause modulation.


Incandescent bulbs also are modulated, but to less depth, due to thermal inertia. You may perceive the modulation from low-wattage bulbs, which enables fine tuning of the turntable speed of a gramophone with a stroboscope disk. The cycle of modulation always is twice the mains frequency.

Post edit: Thermal inertia in an incandescent bulb is greater with a thick filament than a thin one, and therefore low wattage high voltage bulbs have considerably less thermal inertia than low voltage high wattage bulbs. Lower voltage and higher wattage lamps show less modulation because their filaments are thicker.

[ajohnw]

All ok Urban but I was relating my comments to the spectral sensitivity of the eye and the spectral quality of sunlight to a 1st approximation. It's rather hard to tie down the later as it depends on atmospheric conditions, place on the earth etc etc,

Taking a few of the points. CRI is sometimes shown as a polar plot who's radius depends on the colour of light that is reflected from a given colour. Taking leds and tri phospor tubes they both fool the eye into seeing white light just as a pc monitor does. In the case of a power led the blue that is emitted is actually just short of UV. As the eyes sensitivity in this region is very very low rapidly increasing amounts of it have to be emitted to raise the apparent colour temperature. Taking typical daylight type this blue peak can be as higher or even higher than twice the green peak. A warm white led has a blue peak of about 1/2 the green. The red and green outputs are broad and caused by the blue exciting phosphors. Another name for this blue light is dental blue. It's used to cure fillings. Colour rendition of these leds is poor even visually so along comes r g g b leds. The extra green colour improves shades of yellow. The only current application for them as far as I am aware is for microscope lamps and architectural lighting. Tri phosphor tubes have a similar spikey spectrum and again the cri is low in real terms. Multi phosphor types are much more even. It is the spikey nature of some sources that causes the low cri. The problem with spikey spectrums is that they can not illuminate colours well. This is why colour calibration spectometers such as those used to read colours to check printer outputs etc use a very specific form of lighting. The source must emit a full spectrum without any gaps and at a level that allows measurement. Out eyes are broadly similar in some ways - carefully tuned to noting colour lit by sunlight over a surprisingly wide range of colour temperatures. As an example I can post a shot taken through a microscope where the tungsten bulb was glowing orange. My eyes are perfectly capable of reducing this to a neutral grey with poor signs of colour in the right places. More blue than green. The camera was set for tungsten lighting.



The reason for mentioning bulbs was that I have noticed that even under driven halogen provides much better and far more accurate colour rendition. I suspect that is a low cost option even without filters. On microscopes tungsten lights have to be over driven to take realistic photographs. A 12v halogen bulb is usable from 10v which is probably there actual normal operating voltage.

As to the spectrum of all of these light sources they are about on the web apart from halogen which can be difficult to find. Cheap tri phosphor daylight tubes may even emit uv. 6500K tubs emit a lot of blue just like led's, Any simple tube does.

-

[Cparoyan]

Hi Cheryl,

It's the stuff "books are written about" (so hard to articulate in just a few sentences), but for an image like yours, standard settings would work pretty well ... so ... assuming a RAW capture then:

Capture sharpening of 0.3 @ 300%

Content / creative sharpening of 4px @ 40%

(both of the above on the full resolution image)

and then (after down-sampling) ...

Output sharpening in the region of 0.3px @ 50 t0 100%

I've written a bit about it in the past - you might find this helpful ...

Sharpening and Noise Reduction Sequence

Having just said all that, portraiture requires a lot of additional post-processing around the eye area - including additional localised sharpening.

Does that help?

Thanks Colin! I will take a look at the article. I shoot raw so I'll give it a try.

[Colin Southern]

Thanks Colin! I will take a look at the article. I shoot raw so I'll give it a try.

No worries

Just sing out if you need a hand with anything. The actual values vary a bit -- sometimes slightly more aggressive sharpening is needed to compensate for slight focussing errors or lens performance - and sometime less is required depending on the frequency of the target (eg fine hair will get frosty).

Happy to process a RAW capture for you as an example if it helps.

[oleleclos]

Thanks for all input; I'm very impressed with the depth of knowledge on this forum

I have now made the following comparisons between a Bowens Traveller flash head and a 5400K 125W 92CRI flourescent lamp mounted in the same Bowens softbox. All exposures were at f/14 and the exposure time with the tube was 1 sec.

1: Flash, AWB
2: Flash, daylight WB
3: Flourescent, AWB
4: Flourescent, daylight WB



After WB correction in Lightroom, the same pictures look like this:



Not much between them, except slightly darker yellow and brighter cyan in the flash exposures. And arguably slightly better separation between magenta and red in the flourescent exposures.

I'm aware that a grey card and colour patches do not tell the full story of how real world objects will appear, but it's encouraging enough for me to continue playing with this light source.

I like it because it shows me more clearly than flash exactly what I get BEFORE I press the shutter and being less bright it gives me a wider choice of apertures. I'm not worried about exposure times as this is for inanimate objects.

[oleleclos]

After those boring grey cards I thought I'd round this thread off with something on my favourite "light, colour, form and texture" topic, lit by my new flourescent lightbox.

C&C welcome.