Photography: Auto vs White Balance 5000K ["VERY" PIC HEAVY]

[MascaratumB]

Last weekend I took profit of the amazing light on an Autumn afternoon and went outside to take some photos and capture some coloured things / places around my parents’ house.

In the process I decided to experiment one thing with my BQ Aquaris X Pro smartphone: take some photos on Auto and with White Balance on 5000K (no more adjustments) with daylight.

I also tried this under the light of 2 flashlights, the Wuben TO46R (3 x XP-G3 5000K 90CRI) and the Wuben TO50R (4 Samsung LH351D 90CRI), during the night.
Below you will find the results!

NOTE 1: that this exercise was not just for recreation but also to check how “Auto” and “White Balance” may affect our perception of tints and colour rendering when taking photos for our flashlight reviews.

NOTE 2: I won’t draw any conclusion because your phones and cameras would certainly produce different colours from mine. Still, I found it a good exercise.

NOTE 3: In real life, colours look different from both photos with “Auto” and “White Balance”. So none of them reflects what my eyes were seeing.

NOTE 4: In real life (against a white wall), the Wuben T046R, even with a sheet of diffuser that I use, shows some greenish spill that is not completely portrayed in the photos.

NOTE 5: The photos were taken with my phone (12MP camera, 4:3 format, High Quality, JPEG). They were originally in my G00Gl3 account, then were downloaded and the uploaded into IMGUR with the setting of “1024 x 768 (17inch monitor)”, with no treatment.

NOTE 6: My first intention was to mix daylight and night beamshots in rows, but the images were very small, so I opted for this scheme for the comparison:

	Auto	WB 5000K
Daylight
Wuben TO46R
Wuben T050R

---

---

---

---

---

---

---

---

---

---

————— //// —————

Other examples:
Auto vs WB 5000K

---

Thanks for reading!
Leave your comments or questions! And I hope this may help in some way! When taking photos for our flashlight reviews

[mattlward]

Very nice phone camera! That is an interesting study, I seem to move my WB around as I am usually looking for tint shift and going for a setting that produces pure white on a white target. Thus finding a very close measure of the true tint of the emitter. Only when comparing multiple lights do I set at 5000k, most of my lights have been modded to 3500k to 4750k as that is what I like.

[MascaratumB]

Thanks mattlward!
The phone camera is not very good, but it does the job! My older phone (Nexus 5) had a better camera and too better photos, but it is dead now, so this is the one I have

About the different WB, I normally try that too to show the beams specificities, mainly when it has shifts or nuances (more rosy, more yellow). But, most of the times I try use to between 4900K and 5200K that tend to show them better.

However, I don’t know if that is from my camera or not, when comparing the Auto and the 5000K, it tends to produce a more rosy aspect. And that is what I wanted to show too !

To show your 3500K lights’s beams, what settings do you normally use?

[jon_slider]

Thanks for all the photos.

If you have control of White Balance, I recommend 5600k as the consistent Daylight White value. Any other value will change the colors.

Since the LEDs have relatively low R9, Im not suprised they dont do well with reds. LEDs are most challenged by Red objects, other colors are easier, so I do not recommend using multi colored targets.

I would focus on a real Red target, and that will allow comparing how good the LED is at showing that color.

Auto white balance works OK, IF there is a 5600k reference in the image and it looks white. Otherwise the auto white will change the Kelvin temperature of the photo, and of all the colors.

[mattlward]

@MascaratumB

Thanks mattlward!
The phone camera is not very good, but it does the job! My older phone (Nexus 5) had a better camera and too better photos, but it is dead now, so this is the one I have

About the different WB, I normally try that too to show the beams specificities, mainly when it has shifts or nuances (more rosy, more yellow). But, most of the times I try use to between 4900K and 5200K that tend to show them better.

However, I don’t know if that is from my camera or not, when comparing the Auto and the 5000K, it tends to produce a more rosy aspect. And that is what I wanted to show too !

To show your 3500K lights’s beams, what settings do you normally use?

For 3500k it depends on the situation. If I am comparing a 3500k to a 4000k then I will work at 3500k. If I am comparing a 4000k to a 3500k then I will work at 4000k. I intend to make a photo that comes as close as possible to what the eye sees. There is something to be said about always shooting at the same WB for comparison purposes. But to me, it can make both leds look like crap if they are not both around 5000k. I do note my WB settings when I post photos of beamshots.

Here is one that I took years ago, compares A01-219b-sw40 M3-xm-l2 5C tool-xp-g3 5 cu maratac-xp-g3 4 ss maratac-xp-g3 3 ti tool-219b sw45. White balance set at 5000k, from memory EXIF did not contain the setpoint.

This one was white balanced on the 4C, compared to a 5A1 and 7A

[MascaratumB]

@jon_slider

Thanks for all the photos.

If you have control of White Balance, I recommend 5600k as the consistent Daylight White value. Any other value will change the colors.

Since the LEDs have relatively low R9, Im not suprised they dont do well with reds. LEDs are most challenged by Red objects, other colors are easier, so I do not recommend using multi colored targets.

I would focus on a real Red target, and that will allow comparing how good the LED is at showing that color.

Auto white balance works OK, IF there is a 5600k reference in the image and it looks white. Otherwise the auto white will change the Kelvin temperature of the photo, and of all the colors.

Thanks for the tips jon_slider I guess you gave me more reasons to experiment some photos and new settings
Thing is, most of the times, the beamshots don’t have red spots to show :zipper_mouth_face: So, the photos may miss some things then…
I am learning with this stuff and with you guys! It pisses me off when I want to show something and the photo is no accurate in colours. So that’s the reason I’m experimenting this.

@mattlward

For 3500k it depends on the situation. If I am comparing a 3500k to a 4000k then I will work at 3500k. If I am comparing a 4000k to a 3500k then I will work at 4000k. I intend to make a photo that comes as close as possible to what the eye sees. There is something to be said about always shooting at the same WB for comparison purposes. But to me, it can make both leds look like crap if they are not both around 5000k. I do note my WB settings when I post photos of beamshots.

Thanks for the examples mattlward! I guess I will have to tune things better an according to the contexts I have when taking the photos!
Sometimes I try to follow some mentions (in photos) from “pro” people, but in the end it turns even worst than my auto and doesn’t show anything well enough :zipper_mouth_face:

[mattlward]

I do find that a picture taken in a context is more valuable than the same image without the thought of the context it is intended for. Does that make sense? For example, I could have gone with 5600k on the second image, but all would have had a yellow tint. I wanted to reference the difference in context with the 5A1 and 7A. I do tend to find 7A to warm even for me. I really like the 5A1, 2 and 3. I have never been able to get ahold of a 5A4, but suspect that I would really like it. I wish that Cree had better control over the tint shift across the beam. I currently find that I still really like the 219B sw45k and the LH351D UPP6 3500k.

[Pavlo]

I think the problem with any photo is that if it is taken in JPG, an algorithm is used to automatically add color, saturation, contrast, etc to where the image is lacking, making it near impossible to compare 2 lights. Even if taken in RAW, it is subjective and up to the editor to export the images as they best feel it suited.

For tint and colour temperature comparison, I think the only way is to ensure the beamshots are taken within the same photo. In this scenario, you can atleast discern the difference in tint between the two lights, regardless of how they have been altered by the camera into JPG form.

[P33]

If you could only tell me what point were you trying to make/prove. I just don’t understand this exercise.
Thanks.

[MascaratumB]

@mattlward

I do find that a picture taken in a context is more valuable than the same image without the thought of the context it is intended for. Does that make sense? For example, I could have gone with 5600k on the second image, but all would have had a yellow tint. I wanted to reference the difference in context with the 5A1 and 7A. I do tend to find 7A to warm even for me. I really like the 5A1, 2 and 3. I have never been able to get ahold of a 5A4, but suspect that I would really like it. I wish that Cree had better control over the tint shift across the beam. I currently find that I still really like the 219B sw45k and the LH351D UPP6 3500k.

Yes, it makes sense and I guess that is one if the reasons why P33 made the question (below), because I was comparing without a context…
So I guess that this “method” of showing auto vs WB without a broader comparison will not make much sense to show other differences… Got it!

Still, it serves to show how the cameras will modify the colours just for adding a different setting!

@Pavlo

I think the problem with any photo is that if it is taken in JPG, an algorithm is used to automatically add color, saturation, contrast, etc to where the image is lacking, making it near impossible to compare 2 lights. Even if taken in RAW, it is subjective and up to the editor to export the images as they best feel it suited.

For tint and colour temperature comparison, I think the only way is to ensure the beamshots are taken within the same photo. In this scenario, you can atleast discern the difference in tint between the two lights, regardless of how they have been altered by the camera into JPG form.

I won’t repeat my words above but I do realize what was badly done in this exercise, the lack of a contextual comparison (at least when using both flashlights)! We’re still able to see the differences during daylight, despite some are minimal, but I realize the exercise was not as methodic as I thought it was… :person_facepalming:

@P33

If you could only tell me what point were you trying to make/prove. I just don’t understand this exercise.
Thanks.

Well, P33, I guess my answers are above :zipper_mouth_face: I was trying to show the differences when using WB or not when taking beamshots, but it failed to prove that, as it wasn’t contextualized as it should.

So probably this was an worthless effort, that doesn’t show what I thought it would show!

Folks, sorry, and let’s move on!

EDIT: At least you have some nature photos!! :innocent:

[Pavlo]

I hope my previous msg was not misunderstood. I was in no way discrediting anything in your post. I think your images bring value and help others understand how White Balance in a camera can affect beamshots and interpretation of colour temperature. My comment was only to add additional insight on beamshot comparisons

Thanks for sharing your images and comments.

[MascaratumB]

@Pavlo

I hope my previous msg was not misunderstood. I was in no way discrediting anything in your post. I think your images bring value and help others understand how White Balance in a camera can affect beamshots and interpretation of colour temperature. My comment was only to add additional insight on beamshot comparisons

Thanks for sharing your images and comments.

Hey Pavlo, they were not misunderstood
I just realized that I was missing some context and comparison with the beamshots. They may still be useful, but they lack something more, specially a bigger input when comparing different tint flashlights! These are isolated photos, so they only show what a simple setting may do when showing something more accurate or not!

Despite I’ve been making many reviews and trying to portray things as faithfully as I can, I find myself struggling with good settings to show beamshots sometimes. In resumer, I’m also learning and this thread is proof of it

Thanks for the words here, yours and other members, and share your experiences and thoughts :+1:

[Pavlo]

Obrigado

[P33]

Myself I’m an hobby photographer so I kind of know that getting WB right is essential for getting correct colors. Add LEDs with the incomplete spectrum and it becomes much harder.
That’s why I can’t be bothered to do beamshots - it’s just never right

[jon_slider]

Bruno

I suggest you try a beamshot lineup against a piece of white paper,
XP-G3 5000K 90CRI and LH351D 90CRI and a cool white reference light
in the same image,
and set white balance to Daylight (5600k)

this format:

I do not expect the photos to match sunlight, and I do not expect the photos to match what my eyes see.

I only use the photos to compare the beam colors to each other, within a single photo.

this standing lineup can also give useful beam color comparisons:

here is a more challenging comparison format: