pancakes

MicrostockGroup Sponsors


Author Topic: Histogram and white balance  (Read 3931 times)

0 Members and 1 Guest are viewing this topic.

aly

« on: December 12, 2013, 20:40 »
0
I am very confused as to what is correct white balance. I look at the histogram and it is in a curve as recommended but the  image gets rejected for incorrect white balance.Any ideas please?


« Reply #1 on: December 12, 2013, 21:55 »
0

« Reply #2 on: December 12, 2013, 22:36 »
+1
Not sure where you are on the learning curve here, but basically correct white balance means that things in the photo that were actually white, do in fact look white in the photo - not bluish or reddish.   It really has nothing to do with the histogram (unless you're looking at individual histograms for R, B, and G).

ONe way to get correct WB is to use a "gray card". 

 

Goofy

« Reply #3 on: December 12, 2013, 22:55 »
0
We need to know how you are currently processing your photos for white balance (i.e. Camera Raw) to give you a better answers. And how are you lighting - speedlights, strobes, natural light.

Uncle Pete

« Reply #4 on: December 13, 2013, 00:53 »
0
Yes. White balance and Histogram are not the same thing.

I'd add that white balance includes more than just white. It would help to have grey and blacks correct as well.

Histogram is a graphical representation of distribution of image data.

(And now can someone please explain Ohm's Law?  ;)  )


Not sure where you are on the learning curve here, but basically correct white balance means that things in the photo that were actually white, do in fact look white in the photo - not bluish or reddish.   It really has nothing to do with the histogram (unless you're looking at individual histograms for R, B, and G).

ONe way to get correct WB is to use a "gray card".

Ron

« Reply #5 on: December 13, 2013, 01:39 »
0
.

(And now can someone please explain Ohm's Law?  ;)  )




I studied that law

u=i*r

Uncle Pete

« Reply #6 on: December 13, 2013, 02:19 »
0
Yes, that part we know, the formula and it works... now explain it.  :)

I thought it was i*r=e ? Must be an old timer version. Or is it the American Spelling?

Anyway, explaining histograms can somethings be just as complicated. Yes it's right there and it's obvious, but all the graphical representations, in colors or black and white?

A bar graph of a frequency distribution in which the widths of the bars are proportional to the classes into which the variable has been divided and the heights of the bars are proportional to the class frequencies.

Next class, Heisenberg's Uncertainty Principle and how it affects photographic exposure. (but I'd only be guessing?)


.

(And now can someone please explain Ohm's Law?  ;)  )




I studied that law

u=i*r

« Reply #7 on: December 13, 2013, 08:02 »
+3
Heisenberg's Uncertainty Principle can be expressed as w=(r/b)*k where

w = "My certainty that these photons are absolutely spot-on"
r = "What the reviewer thinks of the photons"
b = "What the reviewer had for breakfast"

k = nominally a constant but when it comes to reviewing, it turns out that it's a variable and an irrational number too

w then, as a measure of certainty, tends toward infinite uncertainty
« Last Edit: December 13, 2013, 08:29 by Imagenomad »

« Reply #8 on: December 13, 2013, 10:26 »
+1
There are 2 white balances in a photo:

One that makes true white, so there is no colour cast.

another that makes exposure correct, THAT is the histogram.
So if the histogram is wrong, the whitebalance is also wrong, it could be lighter and darker or anything.
But the histogram does not show if the image has a colour cast.

So first adjust the histogram, to get exposure right.
Then set the white spot in the picture.

All that, of course, conflicts with " being true to light" and "seeing the light" and all that. Do you wish to capture light as it is, or do you want to capture the object as it should be?
That is a decision you will have to make in every photo. We stock photograpers are often loyal to the product, but the most interesting photos appear when lean to the side of the light, and especially when there is a whiteballance conflict, such as a redly lit pumpkin in a blue halloween landscape.

If you take a close (very close) look at Monets water lillies, you will find that every leaf has a white ballance conflict, between cold purple and warm yellow.
« Last Edit: December 13, 2013, 10:30 by JPSDK »

« Reply #9 on: December 13, 2013, 11:25 »
0
Correct the WB for your "golden hour" picture and you will end up with a "white hour".  Colour films had a fixed WB, so a daylight-balanced film would correctly record the distortion in colour around sunrise and sunset but now that WB is variable you need to understand when to allow colour shifts and when to correct them.
A custom WB of about 5700k should replicate the look you would get from a daylight balanced film. If you want to mimic the use of warming or cooling filters on film you can just reduce the temperature to make it bluer or increase it to make it redder. You can even make broad daylight look like deepest night by dropping the WB temperature way down low. But you all knew that, of course.

« Reply #10 on: December 13, 2013, 13:40 »
0
In my opinion there's no such thing as correct white balance, only differences in opinion as to what it should be. I often use it as a creative control - maybe making a snow picture a bit cooler to emphasize 'cold' or a people picture warmer to emphasize 'happiness'. It's a psychological effect and I only bump it a bit.

If the inspector commented on your white balance, his opinion is probably significantly different than yours. Personally, I would take the coaching and give him what he wants. You might also consider that your monitor may be calibrated differently than his.

White balance is reflected in the RGB histogram and the histogram will change when you alter the white balance. I don't know of any histogram controls you could use to specifically control white balance.

« Reply #11 on: December 13, 2013, 14:28 »
0
If you have anything that is any shade of pure grey in your image you can check the WB by looking at the RGB colour balance for that thing. Each of the channels should be the same intensity, then you know that the WB of that object is neutral.

Uncle Pete

« Reply #12 on: December 13, 2013, 19:38 »
0
Winner! Although all the answers are correct, this explains Histogram RGB balance. Is that white balance or grey balance?

Some things if you expose with the correct colors, will get rejected because the reviewer will expect White - Whites and the object is actually off white.

And continuing the color temperature ideas a real easy way to make things warmer is set the color balance for cloudy on a Sunny day. In the end, it's one more tool and adjustment that I'd say is the property of the artist to set as he/she wants it.



If you have anything that is any shade of pure grey in your image you can check the WB by looking at the RGB colour balance for that thing. Each of the channels should be the same intensity, then you know that the WB of that object is neutral.

« Reply #13 on: December 14, 2013, 05:32 »
0
I think WB and grey balance are different names for the same thing. If you have a pure white card and you underexpose it then you get a card that looks grey, but a spot reading of the colours on that card should always have the three of them the same - if you aim to get a perfect match on the actual colour of the object.

If, on the other hand, you want to record the colour of the light rather than the colour of the object the three channels may not be the same.  Sometimes your eyes do not fully adjust to the colour of the light in a scene so you may want to record what you see rather than the corrected colour. For example, the orange glow of a tungsten bulb in a room may give atmosphere to the image, just as the warm sunset colours do. So there are two different answers that are both right - one is to neutralise the colour of the light, the other is to record at least some of the colour of the light.

Beppe Grillo

« Reply #14 on: December 14, 2013, 14:59 »
0
White balance is a kind of nonsense invented by digital photography and not understood correctly by microstocks.

Take a picture of a room, with yellow walls, where there is, in the middle, a white object.
The reality is that the white object will normally appear yellow (and it is right like this).
The distorted reality of microstocks is that the white balance is incorrect

Do you really think that clouds are white??

aly

« Reply #15 on: December 14, 2013, 23:00 »
0
I shoot in raw without artificial lighting then using the warm and cool  adjustments try to get the histogram as even as I can then pick out a white area and use the eye dropper for correction but they still get all rejected-only 1 accepted last lot out of  many. I am learning fast but am very confused still. Thanks for all your helpful suggestions. Perhaps a few images may help.

aly

« Reply #16 on: December 14, 2013, 23:03 »
0
The crow on park bench was ok but the one on ground not.


aly

« Reply #17 on: December 14, 2013, 23:06 »
0
This one rejected for incorrect white balance.

« Reply #18 on: December 15, 2013, 01:46 »
0
I think you are making a mistake in "trying to get the histogram even". It all depends on what sort of light was around at the time and what colour the subject is, histograms are not "one size fits all". 
Also, a crow on a big patch of ground with nothing else to it is likely to have a reviewer looking for a reason to reject it. What use would that picture be to someone wanting to send a message in an advert?
« Last Edit: December 15, 2013, 01:49 by BaldricksTrousers »

« Reply #19 on: December 15, 2013, 09:12 »
0
I think you are making a mistake in "trying to get the histogram even". It all depends on what sort of light was around at the time and what colour the subject is, histograms are not "one size fits all". 
Also, a crow on a big patch of ground with nothing else to it is likely to have a reviewer looking for a reason to reject it. What use would that picture be to someone wanting to send a message in an advert?
That just about sums it up IMO. There's no real practical way to see exactly what the colour balance of an image is from the histogram. You can get a bit of an idea maybe, but that's all. Far better to do it by eye with a decent calibrated monitor set up.
FWIW I normally shoot auto WB outdoors, and find that 90% of the time that does it. Might occasionally need to warm up or cool down an image a bit.  Also sorry to say that I agree with the comment about subject. Small in frame common subject, dead centre on a lot of distracting background hardly makes compelling stock. Whether it gets accepted or not it's unlikely to make many sales.

Goofy

« Reply #20 on: December 15, 2013, 10:44 »
0
white balance is like putting salt & pepper on your dinner food- everyone has a different opinion! Yeah, the recipe calls for a certain about but the best cooks seldom follow the written recipe. 

ruxpriencdiam

    This user is banned.
  • Location. Third stone from the sun
« Reply #21 on: December 15, 2013, 12:09 »
0
Saw your post on SS and you need to remember that this type of rejection is usually a 3 part rejection.

http://submit.shutterstock.com/forum/viewtopic.php?t=132939&start=15

Poor Lighting--Poor or uneven lighting, shadows or exposure issue. White balance may be incorrect.

Your crow also had one or more of these.

Composition--Poor framing, cropping, and/or overall image composition. Please see Shutterbuzz for more info

Composition--Poor framing, cropping, and/or composition. Please see Shutterbuzz for more info

Focus--Your image is not in focus or focus is not located where we feel it works best. Please see Shutterbuzz for more info




 

Related Topics

  Subject / Started by Replies Last post
0 Replies
1881 Views
Last post May 21, 2009, 09:59
by melastmohican
11 Replies
5397 Views
Last post March 17, 2011, 00:11
by RacePhoto
White Balance

Started by tab62 « 1 2  All » Newbie Discussion

33 Replies
9534 Views
Last post April 10, 2011, 20:20
by luissantos84
8 Replies
2018 Views
Last post January 20, 2014, 10:10
by old crow
0 Replies
1211 Views
Last post December 18, 2017, 17:10
by Brightontl

Sponsors

Microstock Poll Results