cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

Custom White Balance kinda sucks

Stonius
Contributor

It could be so much simpler, couldn't it?

 

It takes 10 operations to set up a custom white balance, which seems ridiculous for something that should be a quick and easy single button operation during a shoot. For a start it shouldn't require the image to be shot as it only takes the resultant setting anyway. Just stick the settings in the metadata. Secondly, it shouldn't require that the entire screen is filled with the grey card. If you're on a wide lens, it requires moving the camera to the card (light may be different where the camera is). It should just take into account say, 100 pixels in the center of the image.

 

As it is you have to;

1) switch to an auto exposure mode

2) zoom in to fill the frame with your calibration object (or move the camera). If you don't fill the frame the colour is skewed.

3) Take Photo

4) Press Menu

5) Find Menu 2 under camera settings

6) Enter the custom white balance option

7) Scroll to the photo you just took

😎 Select it.

9) Confirm selection

10) make sure your camera is in 'custom white balance mode'.

 

Needless to say, this is unnecesarrily slow and cumbersome during high pressure shooting environments.

 

I know people will say just to dial it in by eye using Kelvin, but that won't balance for the Magenta/Cyan axis the way a proper white balance does.

 

Markus

 

23 REPLIES 23

Hmm okay. I'm going to revisit this test and post my humble apologies when I find out that I was wrong all along. I mean, sorry, start a flame war! What was I thinking, where are my interenet manners?. 🙂


@Stonius wrote:
Hmm okay. I'm going to revisit this test and post my humble apologies when I find out that I was wrong all along. I mean, sorry, start a flame war! What was I thinking, where are my interenet manners?. 🙂


Don't worry about that too much. I just think there may have been some confusion about what you're final goal was. The responses in this thread are going to be from people that shoot only or mostly still photos. W/B is pretty easy to fix if you shoot in RAW.

 

If you're shooting video that might give you a whole different set of challenges. Are you shooting under really controlled lighting conditions?

Okey doke, yes it was experimental error. Testing it again reveals that it is a tiny area in the middle of the sensor that is the sample area for the white balance calculation.

 

I'll attempt to attach an image that shows the white balance skew with the accompanying images that were used for the white balance in that particular piece of video. The white piece of paper changes only a tiny increment across the field of view, but you can see the difference is dramatic. The middle shot has some colouration so obviously it straddles the boundary of the white balance sample. Not sure if it is affected by metering modes (spot or whatnot) or any other factors, but so far it seems either humble apologies or a good old flame war is in order. 🙂

 

White Balance Comparison

 

Cheers & thanks

 

Markus


@Stonius wrote:

Okey doke, yes it was experimental error. Testing it again reveals that it is a tiny area in the middle of the sensor that is the sample area for the white balance calculation.

 

I'll attempt to attach an image that shows the white balance skew with the accompanying images that were used for the white balance in that particular piece of video. The white piece of paper changes only a tiny increment across the field of view, but you can see the difference is dramatic. The middle shot has some colouration so obviously it straddles the boundary of the white balance sample. Not sure if it is affected by metering modes (spot or whatnot) or any other factors, but so far it seems either humble apologies or a good old flame war is in order. 🙂

 

White Balance Comparison

 

Cheers & thanks

 

Markus


What are we looking at?  I see three near identical looking photos, which appear to have different WB settings.  And?

--------------------------------------------------------
"The right mouse button is your friend."

There are three captures from video of a person that happens to be me on the bottom row.

 

Directly above each of them, are three subtly different photos that show the image that was selected for custom white balance and used in the image below. The subtle difference lies in the fact that the white piece of paper moves incrementally to the left. Once too much of the green backdrop hits the central part of the sensor, the white balance in the resulting image yields a magenta cast.

 

The point is to get a feel for how much of the image is used to determine white balance. So, for instance, you could have an 18% grey card in the middle that doesn't take up the entire frame and if you need to put a colour chart or slate in there to label them, you can without affecting white balance.

 

The reason I'm so concerned about this is I want to create a library that illustrates the effects of a whole bunch of different gels on skin tone, so I need to know that my white is bang on since the whole point of it is the colour.

 

Best,

 

Markus

Learn how to take an accurate sample of a grey card.  

 

Like I said, the coverage does not need to fill the image.  But, the sample shot does need to fill the center area of the image, where the exposure system samples the scene, which none of your sample shots do.

 

What you are doing reminds me of an anecdote.

 

[Athlete]: “Coach, when I hold my arm like this, my arm hurts.” (Actor holds arm above head,touching opposite ear.)

 

[Coach}:  “Well, do not hold your arm like that, because it hurts you.”

--------------------------------------------------------
"The right mouse button is your friend."


@Waddizzle wrote:

Learn how to take an accurate sample of a grey card.

 

 


Thanks, but I don't have an A2 size grey card in my kit. I imagine you don't either. I find it harder to detect tonal shifts in greys than white, so I went with white. It worked, hey.

 


@Waddizzle wrote:

 

Like I said, the coverage does not need to fill the image.  But, the sample shot does need to fill the center area of the image, where the exposure system samples the scene, which none of your sample shots do.


Yes, but the question for me was *how much of the centre frame needs to be filled. I don't think you understand what I'm trying to do here. The white piece of paper *does in fact cross the center of the sampled area - *that's why the resultant white balances are different!

 


@Waddizzle wrote:

 

 

What you are doing reminds me of an anecdote.

 

[Athlete]: “Coach, when I hold my arm like this, my arm hurts.” (Actor holds arm above head,touching opposite ear.)

 

[Coach}:  “Well, do not hold your arm like that, because it hurts you.”


Maybe you want to make sure you know what's actually being attempted before you post condescending comments? Even better, maybe just don't post condescending comments at all. It reflects poorly on you.

I am not trying to insult you.  I'm just being honest.  Sorry, if you do not like it.

 

Your testing is bad science.  You are trying to see if you can fool it, and you succeeded.  Congratulations.

 

You have been told to use the center area of the image, where the AF point and the exposure system analyze the image.  What do you do?  You seemed to have done the exact opposite, in attempt to fool the camera.  Again, congratulations.

--------------------------------------------------------
"The right mouse button is your friend."


@Waddizzle wrote:

I am not trying to insult you.  I'm just being honest.  Sorry, if you do not like it.

 

Your testing is bad science.  You are trying to see if you can fool it, and you succeeded.  Congratulations.

 

You have been told to use the center area of the image, where the AF point and the exposure system analyze the image.  What do you do?  You seemed to have done the exact opposite, in attempt to fool the camera.  Again, congratulations.


Might not look pretty, but it did what I set out to achieve.

 

I had a hypothesis, I tested it, I provided the experimental evidence to prove my claim. That's pretty much how science works.

 

Again, I don't think you understand the point of the exercise; it was to determine the approximate size of the sampled area so I can put other things in the shot without affecting the white balance. 'Fooling it' was kind of the point here. Three perfectly white balanced shots would tell me exactly nothing. I'm sorry you missed that, and that I had the temerity to actually experiment rather than just taking you at your word, but replication of results is also how science works, and you're just some dude on the internet who misunderstands the scientific method.

 

Look, there are a lot of places to get information on the internet. I'm not interested in a pissing contest with an 'Honoured Contributor' with comprehension issues. I'll leave you to your digital fifedom. With 5,377 posts in two years, you've earned it, though probably not much money in all that time you've spent on here (you don't mind me saying that, do you? I'm just 'being honest').

 

PS, I specified that you don't have an *A2* sized grey card (so you can fill the frame at a distance). You really expect me to believe you have a grey card that is more than half a meter on the long side? You should really get those reading and comprehension skills checked out, you know.

 

Bye, Troll


@Stonius wrote:

 

 


 

 

 

 


Maybe you want to make sure you know what's actually being attempted before you post condescending comments? Even better, maybe just don't post condescending comments at all. It reflects poorly on you.


"What are we looking at? I see three near identical looking photos, which appear to have different WB settings. And?"

 

 

Excuse, you.  But, I asked you what was being attempted, and you replied that you were testing the guidelines that you were given, regarding using the center AF point area of the image.  It worked!  You fooled it.  Congratulations.

BTW, I do have a grey card. You lose the bet.

--------------------------------------------------------
"The right mouse button is your friend."
Announcements