DPReview TV: The quality of light, and how different types of lights affect your photos

Yes if you can do something to standardize your vision in some situations and you can see differences. Like have a large sheet of light grey paper and aim a tungsten bulb and a daylight bulb at each side of it. But color is still being interpreted by your brain and it is pretty hard to calibrate it.

This is an extremely complex subject once you go beyond scratching the surface.

When I was a student we had this example in our color class. Everyone watched a slide show in a darkened room. What we didn't know was as the slides progressed they were accompanied by steadily warming filters.

In the end they presented the first image... very blue and the last image ... very warm and we understood how our brain compensated because we had no "calibration" reference.

When you look at an image on a monitor, or at a print, your brain calibrates your vision to your surroundings as a reference. If you work on your computer in a darkened room color adjustment will be tricky.

I copied a bunch of old slides using an LED for focus and framing, and a flash for exposure. Electronic flash has a continuous spectrum and colour temperature close to that of the Sun.
Kodak Ektachrome colours had survived far better than Agfa on these 60 year old slides, but none were unchanged by time depite being stored in darkness for most of their lives.

Right. CFL is not emitting some light frequencies, and there is no filter to balance for information that simply isn't there.

Halogen bulbs work well for "scanning" slides and film with a camera. However they get very hot and are easily broken if bumped while on. Electronic flash is another good source, You can mount it on the camera and bounce the light off of a white card. You may need to turn the output down when working at close range. AE with flash may be problematic.

I get good and consistent results using LED "replacement" bulbs in a desk lamp. These bulbs use a UV LED to excite a phosphor. The spectrum is very smooth, and the slow decay of the phosphor eliminates flicker and banding caused by direct LED operation.

I always wondered about the continues spectrum claim re LED lights. If it is just a phosphor like CFL, how to they do it? CFL is very not continues

Hats off to you, your brain must be an absolute library of color science.

Thanks. Yes I studied color science and human vision as part of my course at RIT. But applying it to shooting unforgiving transparency film under a range of lighting was extremely challenging. That kind of photography was more like a science project each time than photographic art. Sometimes there were many steps to make a single exposure...

Beyond the color science was the limitations of the film medium itself... color and density shifts as exposure time changes, every film type having its own unique interpretation of color. Every batch of film needing testing.

All of this is still very useful for shooting digitally but it is way easier now.

Typical shot in an office space at 3AM when my brain is well past its prime:

Shoot image for 15 seconds with specific room lights on and a filter on the lens. Turn off those lights and send assistant into room with ladder to remove the two "security bulbs" that can't be shut off. Use broom to sweep up marks he made in the carpet. Carefully remove filter, connect shutter to strobes, take a frame using those lights. (I had self cocking shutters of course.) Sometimes we would put filters over the bulbs in the ceiling or over windows.

Repeat this process for about 10 frames of film or so for bracketing and backup to hold from the lab. We also had to cover all exit lights with material that matched the wall or ceiling. (Clients hated seeing these.)

I owned every Wratten filter in 4 and 3 inch for front and back of the lens plus a lot of glass filters. (I still have them.) I had holders that let me add or remove a filter without moving the camera.

I did get tired of this.

I found and posted some of the typical film examples from a long time ago. Some of these old scans could be better adjusted.

https://alangoldstein.photoshelter.com/gallery/Old-architecural-film-images/G0000UDNd4kFNaow

Wow, you got access to some high up there places, like those board rooms...

It seemed normal for me to be hired to shoot large projects. Moving the gear over these spaces was a hassle.

There are good back stories to some of these photos. Mobil was on a large wooded campus with high security. They told me the boardroom was bombproof and the doors were vault like. Spouses of workers were not even allowed to visit. Everything coming in was x rayed and inspected. I was illustrating how the security was blended with the architecture as part of the assignment.

Amazingly we spent a few days there rolling ten + cases of gear onto site and they never inspected anything. Over the weekend a car raced through the site and threw a brick through the entrance window.

Another one was shooting a fancy hotel lobby late at night and watching the expressions on the "escorts" faces when they exited the elevator and had to cross over an area with 10,000 watts of lights on them.

Those are some really amazing images!

The one I love to say to people, especially outdoors, is what color is the grass?
Green? Nope, it's every color BUT green. All the colors of light are absorbed, BUT green is reflected, which is why we perceived it as green. ;)

So you are saying the M in CYMK is fake?
Excuse me while I go tell my color enlarger that it has been lying for the last 50+ years. (It has a dicroic head) similar but an older model of this.
https://www.bhphotovideo.com/c/product/105697-REG/Beseler_8011_03_23CIII_XL_Dichro_Enlarger.html

PDL, magenta is a "metamer", of which many exist but in this case you cannot find a corresponding response in the normal visible spectrum. Look up the term and get ready for some head-scratching reading... but your magenta light in your enlarger is made of two distinct wavelengths of light being emitted at the same time. All colours are dependent on human perception more than we realize in our everyday lives.

Don,
With due respect a dichroic head on a color enlarger does not have a "magenta" light. The lamp is halogen and the filters are thin layers of metal (I don't know what type) evaporated over optical glass.

I would imagine that Cyan is also a metamer but if you look at a common color wheel - there they are. You can even buy filters that are Magenta, so to say "it does not exist" is not valid. If you say it does not exist in the spectral response of light sources, then yes what you state is true. So lets not talk about other "colors" that do not exist when using the spectrograph - like Brown, Orange or Puse. 😁

To say any colour does not exist is valid, since every colour is a creation of our minds and isn't seen the same way by other creatures - but I digress. We're strange beings when it comes to perceiving light - and the senses only get stranger with smell and taste. We've developed tools in the visual space to define "beauty" to our eyes just as we've refined the culinary experience to be delicious to our tastes.

I'm going a little overboard there, I know. I agree with you. Maybe I should do another video on metamers in the future. I remember some old graphics cards actually had a dedicated "brown circuit" to handle some of those colours!

CRI index is an OK guide for lighting quality, but it is only defined by certain colour patches. This means there is still the potential for significant spikes or dropouts in the spectrum to be present even with a high CRI. This is the case for many compact fluorescent bulbs which are marketed as having a CRI well over 90.

The only way to be really sure of the lighting quality is to look at the spectral plots.

You can do a simple test with a sheet of white printer paper. That paper is designed with optical whitening agents that essentially fluoresce under UV light - which is also the reason why this paper appears bluish when placed outdoors on neutrally-white snow.

If the paper appears brighter under one light source, that means the light emits more UV light. This is also the reason why I favour printing my photos on high quality papers that do not use optical brightening agents - the "look" of the image may change more drastically than one would otherwise anticipate based on the lighting conditions. They also decay over time.

Beware of the yellow snow!

If you have a Leica, you don't even need light, you can shoot with lens caps on and the photo will still be highly acclaimed.

Tonnes of Leica shooters actually photograph with lens caps on.

http://leicaphilia.com/eric-clapton-forgets-to-take-the-lens-cap-off-his-leica/

http://1.bp.blogspot.com/_mmBw3uzPnJI/TA6YtGyUPaI/AAAAAAABUxU/T-UOl1xfmeo/s1600/camera_fail_01.jpg

https://www.google.com/amp/s/www.thephoblographer.com/2014/05/07/photographer-forgets-remove-leica-m-lens-cap-shoots-10-perfectly-black-frames/amp/

Because money can never buy skill.

You're right - different materials will transmit light to different degrees. A great example is the previous video I did with a lens made of all quartz glass that transmitted more UV light compared to "regular" glass. Coatings, optical design, and material choices all have an affect on the final "balance" of light before it hits the camera sensor.

Some objects may, for the human eye in sunlight, appear to have the exact same color. Those same objects might differ in color in other lightsources. Secondly, if you take a photo of these objects, even in sunlight, they may appear having a different color on the picture, due to differences in the exact responsiveness of the camera to specific colors in the spectrum, in comparison to the resposiveness of the human eye to those specific colors of the spectrum. (Let alone when you take a photo of these specific objects using a fluorescent light source.)
If you ever have done some photography on military stuff you may know that not everything that appears to have the same hue of military green necessarily has the same color in the pictures. Flowers are famous for this effect too. (also the colors of flowers are in many times so saturated they fall out of the camera's color gamut, but that's a different effect)

Rather than correcting the colors, it would be interesting to test how good they are at capturing the different spectrums, since the sensor's bayer filter and the sensor itself won't have the same response to every red or every green in the picture.

That would be interesting Txoni. You have to realise however, that every camera as far as I understand already as quite an amount of corrections even before saving the result als a rawfile. I've read things about those corrections and soon concluded to better stop reading, make photo's and correct them to a point where I like them* as long as I don't work for a museum or so.

*: on my calibrated screen that kind of resembles the screens of the people that look at my pictures.

Using the Color Checker Passport to try and give the most optimal response for corrected colour and it still couldn't match things properly, as Leonp suggests this is because the light source isn't emitting the frequency of light required to reflect off the subject, pass through the lens, and be detected by the sensor.

You can get closer to "reality", but it all starts with a good light source.