the effect of white balance on the perception of color and tint

jon_slider · 2019-07-02T22:19:56.000Z

these three images are unretouched, however the pair of lights in the 3rd image have changed color due to auto white balance from the iPhone

in this next image, with annotation, is also not modified for tint nor color temperature in post processing:

All above photos by iphone, auto white balance, at 2pm on a sunny blue sky day near San Francisco in July
imo the light sensor in the camera is choosing the coolest CCT in the image to reference its white balance setting

in this next image the bottom pair of lights have been edited to change the color temperature towards the 3000k source in post processing, making the 4500k source appear blue

imo my brain does the same thing, with a 30 minute lag time for the brain to change its ambient white reference. Iow, our brain has (slow) auto white balance.

because of the effect of white balance on color perception, I recommend using daylight white balance when comparing beam colors and tints, for consistency

since my iPhone has auto white balance, I always include a cool white light to make the auto white balance adjust to cool white, which is near daylight white balance

BurningPlayd0h · 2019-07-03T00:38:55.000Z

“which is near daylight white balance”

5000-5500K has always been what I’ve seen for that and has always looked pretty true to life for me.

Does iPhone not have manual WB? I’m sure there’s a camera app out there that would include it.

To me that shot of the 3000 and 4500K only looks much closer to how the lights appear unless compared to BRIGHT sunlight (and then they’d look only slightly warmer).

scintillator · 2019-07-03T14:57:08.000Z

I am still struggling to grasp this,I found this film school blog on color temp easy reading.

https://www.lightsfilmschool.com/blog/filmmaking-white-balance-and-color-temperature

jon_slider · 2019-07-03T16:16:30.000Z

BurningPlayd0h:

“daylight white balance”

5000-5500K

I agree, D50 and D55, are Daylight White Balance references, so is D65…

.

scintillator:

I found this film school blog on color temp easy reading.

good link, definitely a confusing subject, thanks for taking the time to try and gain a better understanding

quoting from your link (in italics ):

Setting your white balance will allow your camera to identify the color of pure white.

Our eyes have a decent ability to correct white balance

Without setting your white balance, your true white colors may appear unnaturally blue’ish, orange’ish or even green’ish

white balancing identifies what is white in your footage

The two color temperatures you’ll hear most often discussed are outdoor lighting which is often ballparked at 5600K and indoor (tungsten) lighting which is generally ballparked at 3200K. These are the two numbers you’ll hear over and over again. Higher color temperatures (over 5000K) are considered “cool” (i.e. Blue’ish). Lower color temperatures (under 5000K) are considered “warm” (i.e. orange’ish).

Therefore if you are shooting indoors under tungsten lighting at 3200K you will set your white balance for indoor shooting at this color temperature. In this case, your camera will correct your camera’s settings to ensure that white appears white. Your camera will either have an indoor 3200K auto option (even the most basic camera’s have this option) or you can choose to set it manually. That is all you need to do to set your white balance. Once you’ve set your white balance you don’t need to change it after each shot unless the color of light changes.

.

Same info as quoted], in my words:

in this image the white balance that shows the optisolis as white, is 6500k (D65)

at 6500k (D65) white balance, the 3000k 219c (sm303) looks very orange.

at 6500k white balance, a piece of white paper illuminated by the optisolis, will look white in a photo, and the same piece of white printer paper illuminated by the 3000k 219c will look orange

.

So, what if we are taking photos of a piece of white printer paper, using the 3000k 219c as the only light source, and we want the photo to show the paper looking white? Then we would have to set the white balance reference for the camera, to 3000k.

The 3000k LED, can make a piece of white printer paper look white in a photo, IF the white balance reference point of the camera is set to 3000k also

And the paper illuminated by a 3000k source, can look white to our brain also,
IF our brain spends at least 30 minutes exposed to Only the light of the 3000k source.

Our brain can adjust its own white balance so that we see the piece of white paper look white, when illuminated by a 3000k source.

clemence · 2019-07-03T17:28:31.000Z

Jon, you might want to check the link below:

MUO – 15 Dec 19

The Best Camera Apps for Android and iPhone

Here are the best camera apps out there for Android and iPhone that'll bring out the best in your smartphone photographs.

[Clemence]

coherent-rambling · 2019-07-03T17:29:54.000Z

Our brain can easily adjust to different lighting conditions, yes. If we’re exposed to predominantly one color temperature and tint, it’ll generally look white. White balance is the camera’s way of dealing with the same situation; you can white balance to make just about anything look white.

Unfortunately, that’s not all there is to it. You have to remember that those photos are being viewed by people with different monitor calibrations, sitting in rooms lit differently. Nothing you can possibly do is going to make a beamshot accurately represent the true color of the beam to everyone who views it.

The easiest way to deal with this is to treat beamshots as a comparative tool, not an absolute reference. Use manual white balance, and indicate what setting you used (ideally always the same setting), and people will be able to understand more clearly how various light sources compare. You could make an argument for 6500K, since most monitors are backlit with something in that ballpark and you’ll probably end up with the most accurate absolute representation, but I personally use 5700K because I think it’s a more sunlight-like reference and gives me a better relative representation for how I’ll feel about each light.

Ultimately, the most important thing is just picking a consistent reference point and sticking with it. You can misuse white balance to make any source look good or awful, so it’s best to not let it. Use presets or manual and the problem disappears.

clemence · 2019-07-03T17:43:28.000Z

coherent-rambling:

Our brain can easily adjust to different lighting conditions, yes. If we’re exposed to predominantly one color temperature and tint, it’ll generally look white. White balance is the camera’s way of dealing with the same situation; you can white balance to make just about anything look white.

Unfortunately, that’s not all there is to it. You have to remember that those photos are being viewed by people with different monitor calibrations, sitting in rooms lit differently. Nothing you can possibly do is going to make a beamshot accurately represent the true color of the beam to everyone who views it.

The easiest way to deal with this is to treat beamshots as a comparative tool, not an absolute reference. Use manual white balance, and indicate what setting you used (ideally always the same setting), and people will be able to understand more clearly how various light sources compare. You could make an argument for 6500K, since most monitors are backlit with something in that ballpark and you’ll probably end up with the most accurate absolute representation, but I personally use 5700K because I think it’s a more sunlight-like reference and gives me a better relative representation for how I’ll feel about each light.

Ultimately, the most important thing is just picking a consistent reference point and sticking with it. You can misuse white balance to make any source look good or awful, so it’s best to not let it. Use presets or manual and the problem disappears.

Have we met before? I totally agree with all mentioned above.
Given all the variables in our cameras, monitors, eyes, etc…. Any reduction in variables for community based presentation is always worth the effort. A consistent fixed WB for all the beamshots we post is a good start….
You’ll present more information with less random variables and consistent order.

[Clemence]

jon_slider · 2019-07-03T21:19:39.000Z

coherent-rambling:

I personally use 5700K because I think it’s a more sunlight-like reference and gives me a better relative representation

very helpful post, agree in all regards including a preference for “daylight” white balance @5700k

also agree that it is a “relative” reference that can be helpful, not absolute reference, so My beamshots are never just One beam, and they always include a “known” LED for reference, such as in this case 219b 4500k, and 3000k 219c, Plus a cool white reference, in this case the very cool 6500k Optisolis.

clemence:

A consistent fixed WB for all the beamshots we post is a good start…

also complete agreement as far as that goes.

In addition Im tackling the effect of altering white balance to 3000k on a 3000k source, to illustrate the concept that “white” exists at more than just one CCT. This concept is critical to understanding why the same LED, viewed during the day, can look warmer than daylight, and viewed under incandescent at night, can look cooler than incandescent.

.

My motivation is to normalize the conversation regarding color and tint, where atm people use words like Pure White, True White, Creamy White, while failing to take into account the meanings of Warm White, Neutral White, and Cool White.

So we end up with people reporting that they Hate Warm White because it looks too orange (it does, but only if viewed from a vantage point of daylight white balance.)

then we have people that talk about how 3500k is the perfect CCT, for them, without filling in the details regarding the ambient color temperature, and hence the white balance of their brain, when they prefer 3500k.

Im not pretending to be an expert in any of this. My goal is to facilitate sharing information in a way that can be standardized, and shared, with visuals.

I do totally agree that it makes most sense to offer images that are white balanced to “daylight” white @ 5700k (± 300k)

thanks to everyone contributing to the conversation.

Happy 4th of July, if youre in USA

scintillator · 2019-07-03T21:16:29.000Z

Home monitors are a real wildcard here.I keep forgetting that I use F.lux program on my computer which is on most of the time and has one prime directive,
to protect my eyes and promote better sleep.(see link a little off subject)

f.luxometer

f.luxometer shows how screens make light and predicts how the light affects your body.

Most of the time I can tell when the f.lux is on and switch it off easily.

BurningPlayd0h · 2019-07-04T20:24:03.000Z

jon_slider:

then we have people that talk about how 3500k is the perfect CCT, for them, without filling in the details regarding the ambient color temperature, and hence the white balance of their brain, when they prefer 3500k

For some people that is not at all the reason they prefer very warm white light though. I don’t think my brain ever really adjusts to either very warm or very cool white light. I can be around only incandescent-temp lights in my house for hours and they are still noticeably warm.

jon_slider · 2019-07-05T01:24:37.000Z

BurningPlayd0h:

For some people that is not at all the reason they prefer very warm white light though. I don’t think my brain ever really adjusts to either very warm or very cool white light. I can be around only incandescent-temp lights in my house for hours and they are still noticeably warm.

I agree
LED choice is not only influenced by a single factor, such as how cool or warm it looks, in a given ambient environment.

There are other variables that people may prioritize differently, for example

The CCT of the LED under which the operator chooses or prefers to operate, can be due to personal preference for a mellow relaxing lighting mood, or it could be that their priority is not to fall asleep, and they want to maximize their ability to work under cooler light that potentially is more energizing, and has more daylight color fidelity.

For example, it is possible that someone who uses their light as a support for their work, may tend to prefer cooler CCT. I dislike working on a car in the sun with a 4000k led, and prefer a CW LED for that application.

people who prioritize the use of their light for nightstand use, may tend to prefer warmer. I find CW rather blue for nightstand use, and I prefer sub 4000k for that application

even after driving in the bright sun, some prefer to use a 4000k LED. even during the day, with their brain adapted to cool white. they accept and are not disturbed by the relative warmth of the beam at that time.

others will differ, and claim 4000k looks too “yellow and dim”, to them

no one variable, I like Im talking about one factor at a time, and youre also right, that there are many factors that influence a users LED preference

thanks for keeping me honest