40D: Weird colours

[yakim_peled1]

As the difference between the two shots is 1 stop I took the normal one and underexposed it by 1 stop at DPP. As you can see, it just looks darker but the colours don't look like the weird one.

 

In both cases the center AF point was used and it was placed on the grasshopper's face.

 

 

Happy shooting,

Yakim.<div></div>

[phule]

Unless you mean the blue and green colors at the bottom of the frame, which, to me, appear to be objects in the foreground that are simply out of focus.

[phule]

[[As the difference between the two shots is 1 stop I took the normal one and underexposed it by 1 stop at DPP. As you can see, it just looks darker but the colours don't look like the weird one.]]

 

Of course it doesn't. One shot is being lit directly by the sun and one is not. You can't just take a subject fully in sunlight and drop the exposure by 1 stop in post production and expect it to come out looking like the sun wasn't there in the first place.

[yakim_peled1]

Rob, in the first shot it was sunny. In the second it was slightly cloudy but I can assure you that the colours were nothing like you see in the weird pic. In fact, they looked very much like the underexposed pic.

 

 

Happy shooting,

Yakim.

[phule]

[[Rob, in the first shot it was sunny. In the second it was slightly cloudy but I can assure you that the colours were nothing like you see in the weird pic. In fact, they looked very much like the underexposed pic.]]

 

Yakim,

 

You do understand that the camera does not see the world the same way your eyes do, right?

[yakim_peled1]

Rob and Peter, you obviously try to tell me something but I just don't get it. I think that the key sentence is the last one in Peter's post but I simply don't understand it.

 

Now please treat me like a complete moron and explain in the simplest terms possible. I really don't mind to be considered as such as long as I am able to understand this issue to the core.

 

It's midnight now in Israel and I have a busy day tomorrow. I'll look at it tomorrow and hopefully be able to figure it out. Remember: Use the simplest terms possible.

 

And of course, my deepest thanks.

 

Happy shooting,

Yakim.

[phule]

Yakim,

 

Might I suggest you go to your nearest library and/or book store and pick up a few books on the fundamentals of photography? You seem to have dug yourself into a deep whole of misconceptions and misunderstandings and I'm not sure a web forum is the best place to try and correct these. I'm not saying this to be insulting.

[yakim_peled1]

Hi Rob,

 

I am shooting for 18 years and I honestly think I got the basics well covered. What I know is that aperture and shutter determine the amount of light getting into the film/sensor. Now, what does that has to do with the colours of the image?

 

BTW, the libraries and book stores in my vicinity has no books about photography.... :-(

 

Happy shooting,

Yakim.

[petemillis]

Yakim, re your grasshopper shots....Chris mentioned that AWB measurment is on the circle around the centre AF point. In your picture where the grasshopper is small in the frame, there is nothing white or grey around where the centre AF point would be - all the colour around the centre AF point is warmer than white/grey, and so AWB will try and pull it down to neutral giving a colder image. If you use the tool in DPP to select an area of white/grey (I would choose the area between the grasshopper's front feet) then the WB comes up with a much warmer picture (see attached). However, with the picture where the grasshopper is bigger in the frame, the AWB region includes a large area of off-white/grey on the grasshopper's right hand cheek (left cheek in frame) giving a much warmer picture as the computer in the camera sees this already as neutral.<div></div>

[petemillis]

You need to remember, a camera and its software are dumb - the camera has no way of knowing how your eyes and

brain see things. You can't rely on the algorithms in the camera's software to give the results you expect all the time.

[philipp_wetzler]

Yakim,

 

direct sunlight, especially in the morning and evening but also during the day, is warm (yellow / orange). When the sun is

covered by something (like a cloud, or when you're in the shade) most of the light that illuminates your scene comes from

the sky, which is neutral (when it's cloudy) or cold (blue-ish, when the sky is blue).

 

The brain and the eyes are very good at compensating for this effect, which is why, unless we train ourselves to see it,

we don't much notice that the light is more yellow or more blue. But the sensor in your camera sees the difference very

clearly, and on the computer screen so do you.

 

AWB tries to compensate for this effect, but it is very conservative about it, so a picture taken in cold light will generally

still look somewhat cold, while a picture taken in warm light will generally still look warm. I don't shoot Canon, but I've

noticed that with my camera, setting the white balance manually will give much more consistent results. I'd suggest you

play with the various manual WB settings on your camera and get a feeling for what they do; alternatively you could get

in the habit of checking & adjusting WB after the fact on the computer.

[yakim_peled1]

I must admit I am thoroughly confused.

 

Rob and Peter are convinced that this is an exposure issue, a notion I find difficult to believe because I can't

understand how exposure can be related to colour.

 

Pete is convinced that this is a WB issue, which I find much more reasonable, but then comes the question: If it

is indeed so, why can't I bring the original colours back. After all, RAW files are supposed to enable you just that.

 

Whatever the answer, it must be the same for both pictures and also account for the fact that this has never

happened to me in the 18 years that I shoot. What was so special in those particular sets? I have shot the same

subjects before and in that location before and in similar conditions before.

 

Would you please make up your mind? :-)

 

>> You need to remember, a camera and its software are dumb - the camera has no way of knowing how your eyes and

brain see things.

 

I agree.

 

>> You can't rely on the algorithms in the camera's software to give the results you expect all the time.

 

Well, it has been doing just this to date. I wonder what was so special in these particular conditions that

sparked the change?

 

P.S. All of a sudden all these questions regarding lens/camera A vs. lens/camera B seems so easy to answer, don't

they?

:-)

 

 

Happy shooting,

Yakim.

[philipp_wetzler]

Yakim, you said

 

"Pete is convinced that this is a WB issue, which I find much more reasonable, but then comes the question: If it is

indeed so, why can't I bring the original colours back. After all, RAW files are supposed to enable you just that."

 

You have to understand that in your sun-lit picture, parts of the picture are in direct sunlight (i.e. warm light) while parts

are in the shade (i.e. cold light). In the other picture, everything is in the shade (cold light). By changing the overall white

balance, there's no way you can make one of them look like the other -- if you make the shade picture warmer, it will be

warm all over, including the parts that are cold in the sunlit picture.

 

Scene lighting is a very complex issue, you almost always have a mix of different color temperatures and tints. AWB will

give you a rough starting point, and manual calibration can get you closer to getting consistent results, but most of the

time there's just no way to fix differences in lighting in the camera or on the computer, which is why it's very important to

get the lighting right in the first place, by waiting for the right moment or by adding your own light / shade. And in order to

do that you have to understand the light first.

 

It's hard to believe that you've never noticed the effects of changing light in 18 years of shooting, unless most of that

was with negative film (photo labs will try to adjust white balance for you when they make prints).

[philipp_wetzler]

Proof of concept: Adjusted the exposure by 1.25 stops and turned the color temperature WAAAY up. This is as close as it

gets without doing localized adjustments in Photoshop. But it doesn't look natural, because you wouldn't get such warm

light without any shadows; nature's diffuse light isn't generally warm.<div></div>

[yakim_peled1]

Philipp, I am willing to believe you if Rob and Peter will..... :-)

 

>> nature's diffuse light isn't generally warm.

 

It was in this case. It looked very much like my DPP -1 underexposed example.

 

And yes, 15 of my 18 years were shooting film and it was mostly developed and printed in photo labs.

 

Can you please comment on the original picture?

 

Happy shooting,

Yakim.

[philipp_wetzler]

"Can you please comment on the original picture?"

 

You mean the sunset picture? Without having seen the actual sunset you photographed I can only speculate. I have in

the past frequently seen sunsets where the whole part of the sky around the sun takes on various shades of the same

golden color, just like in this picture.

 

Now you can say "no, I know what I saw, and the sky didn't look like that". But these are the two alternatives:

 

1) You made the simple, very common and very human mistake of thinking you saw one thing when in fact it was

another

 

or

 

2) the sensor in your camera, suddenly and arbitrarily on only this picture, recorded light in a fundamentally different way,

resulting in a different but still plausible picture; also, this happens only with _your_ camera.

 

Given what I've seen, alternative 1) is vastly more likely, and so I'll go with that. Think about it: Your eyes see the whole

rest of the sky, and it is blue; so your brain perceives it as blue, even in the small area where it's not. Your eyes see

white clouds everywhere, and your brain knows that clouds are supposed to be white, so it sees white clouds. In the end

you see a variety of colors where physically there's only shades of the same color. Our vision is easily tricked into

seeing what we expect, rather than what's actually there; only with practice we can learn to sometimes compensate for that,

at least to some degree.

[yakim_peled1]

No, I know what I saw, and the sky didn't look like that.... :-)

 

But seriously, the whole point of my incapability to understand is that in each and every other shot (tens of thousands just with digi: 1D and 40D), what I saw is what exactly what I got in the picture. What was so different in these particular shots?

 

Happy shooting,

Yakim.

[mark u]

EXIF of Grasshopper pics:

 

I AWB, FLASH, 1/250th f/6.7 Tv mode ISO 200 (-0.5EV compensation, no FEC info)

 

II AWB, no flash, 1/750th f/2.8 Av mode ISO 200 (-0.5EV compensation)

 

First image should be around 5600K, second image much higher (7500-10000K depending on whether light is diffuse via cloud or open blue sky in shade because sun blocked by smallish cloud).

[philipp_wetzler]

"But seriously, the whole point of my incapability to understand is that in each and every other shot (tens of thousands just

with digi: 1D and 40D), what I saw is what exactly what I got in the picture. What was so different in these particular shots?"

 

The difference is that you're beginning to pay more attention to the results you're getting from your photography. Light hasn't

changed, your camera hasn't changed -- you have: You're getting more discerning, and your ability to see is improving. Accept

it, don't tell yourself it isn't real or that your camera is lying to you, and instead get out there and take pictures! Find different

kinds of light and try different WB settings and just play with them and see what you get.

[yakim_peled1]

I tried to reproduce the sunset pic. Very similar conditions. ISO 100. Av. F/9.5. 1/8000. AWB. The only difference is the metering (center weighted, by accident) and the lens: 17-55/2.8 instead of 100/2.8. Colours are true to reality.

 

Mark, I did not understand what you mean. Could you please elaborate?

 

Happy shooting,

Yakim.<div></div>

[petemillis]

Yakim, all this pointing your camera at such a bright sun, and looking through the viewfinder...you're probably burning up the sensor and your retina. Could lead to all sorts of difficulties!

[philipp_wetzler]

Yakim, I don't get your point with your most recent picture of the sun -- it looks just like the other one, only it's not a

sunset so it's more blue-ish, less orange-ish. It's still all shades of the same color. If you could show us two raw photos

of the SAME sunset at the SAME time taken with the SAME white balance, and one of them has different colors than

the other, that would be meaningful. This isn't.

 

Seriously, stop trying to find faults with your camera and the pictures it takes. You're pushing yourself into seeing

problems that don't exist. All the photos you've shown here look exactly the way I'd expect them to look. You keep

acting like somehow the camera didn't take the picture right when, in fact, all it did was record what was there.

 

Once you stop trying to find fault with the results your camera gives you you can start trying to actually understand the

relationship between what you THINK you see, what is actually there, and what your camera records (at its various

settings). And it involves taking lots of pictures out there, playing with WB settings on the camera and on the computer

and comparing the results, and believing that your camera works the way it was designed to work.

[yakim_peled1]

You know Philipp, your last paragraph reminds me of early replies to the first reports of the AF problems in the 1D Mk III......

 

Thing is, I know what I saw and it was nowhere near those weird pictures. It has not happened before and not since. I just want to get to the bottom of this and unfortunately, despite of all the best efforts of all the good men here in PN, I failed. Some think it's exposure issue, others think it's WB issue and others think it's a combination of the two. Can you blame me for being confused?

 

Happy shooting,

Yakim.

[ad trace]

I'm definitely no expert (and I'm not getting into the camera v eye business) but..

 

It's my understanding (and interpreting, hopefully correctly, the above posts) that the exposure could cause the loss of colours - in order to expose the sun correctly (the sun is close to both the centre of frame and the selected focus point) the camera has grossly underexposed the surrounding sky.

 

When you *really* under or overexpose then you end up losing colour info (much like being able to bring back blown highlights from RAW files. You can get details back but the colours are gone). In the case of your sunset picture, changing the white balance in pp doesn't regain the colours as they were because the colour information was never recorded in the first place.

 

If you underexpose a shot by several stops you would not expect to be able to pull a perfect picture from the file in pp.

 

Beyond that, I'm stumped... :-)

 

Good luck

AD

[dogbert]

I am thinking it is more likely a camera problem, perhaps slightly corrupted firmware or possibly but less likely a card with slightly corrupted segements. My reasoning is simple: Yakim is an experienced photographer and I doubt he is seeing/imagining somethign different to what the camera is seeing and the photos are clear anomalies compared to tother shots he has just taken.

 

The sunset photo looks quite desaturated - easy to produce in photoshop but impossible to produce if trying to take a natural looking shot that is full of golden warms colours. I have taken thousands of shots in warm light on either AWB or daylight and never seen anything like this.