Reading mode: Light Dark

Two of the same cameras produce different image when set to the same settings?

Go to last post
A

alexdrake

Member
Messages
24
Reaction score
3
Hi!

I have a Sony A7III and recently rented another one for a project. However, when testing the cameras side by side I noticed that one had a more red tint than the other. I put the two cameras in front of a white paper, same distance, lens and white balance and still one seemed to have a more red tint. I tested this several times and still the same result.

How can this be?
 
Last edited:
alexdrake said:
https://drive.google.com/open?id=1FKOjFXMhs_G6W2SvklrXmUdnYeGbTOI6

Uploaded Raw-files here if anyone wants to view them.
Click to expand...
These files are badly underexposed which will exaggerate differences when adjusted. As well on of them reads as a tint of 16 magenta and the other says plus 17 when opened in acr.

I think that the problem is in the lighting.... possibly some stray light of low intensity, a light on or off in another room, a shift in the ambient light from a cloud passing in front of a window? who knows... you need a better exposed test to sort this out.

How significant a problem this is - and whether camera variation is the problem - needs a better quality test to sort this out. It would be better to see a properly exposed image that was not all squeezed to the left of the histogram as a basis to evaluate this issue.
 
justinwonnacott said:
alexdrake said:
https://drive.google.com/open?id=1FKOjFXMhs_G6W2SvklrXmUdnYeGbTOI6

Uploaded Raw-files here if anyone wants to view them.
Click to expand...
These files are badly underexposed which will exaggerate differences when adjusted.
Click to expand...
But there are differences. I see different % pixels per color underexposed in FastRawViewer. Might be useful to dig deeper in exif
justinwonnacott said:
As well on of them reads as a tint of 16 magenta and the other says plus 17 when opened in acr.
Click to expand...
One point shift in Tint is not insignificant and this is with everything reset
justinwonnacott said:
I think that the problem is in the lighting.... possibly some stray light of low intensity, a light on or off in another room, a shift in the ambient light from a cloud passing in front of a window? who knows... you need a better exposed test to sort this out.
Click to expand...
More paired shots would help, but I think the OP has already been there and done that
justinwonnacott said:
How significant a problem this is - and whether camera variation is the problem - needs a better quality test to sort this out. It would be better to see a properly exposed image that was not all squeezed to the left of the histogram as a basis to evaluate this issue.
Click to expand...
 
robert1955 said:
justinwonnacott said:
alexdrake said:
https://drive.google.com/open?id=1FKOjFXMhs_G6W2SvklrXmUdnYeGbTOI6

Uploaded Raw-files here if anyone wants to view them.
Click to expand...
These files are badly underexposed which will exaggerate differences when adjusted.
Click to expand...
But there are differences. I see different % pixels per color underexposed in FastRawViewer. Might be useful to dig deeper in exif
robert1955 said:
As well on of them reads as a tint of 16 magenta and the other says plus 17 when opened in acr.
Click to expand...
One point shift in Tint is not insignificant and this is with everything reset
robert1955 said:
I think that the problem is in the lighting.... possibly some stray light of low intensity, a light on or off in another room, a shift in the ambient light from a cloud passing in front of a window? who knows... you need a better exposed test to sort this out.
Click to expand...
More paired shots would help, but I think the OP has already been there and done that
robert1955 said:
How significant a problem this is - and whether camera variation is the problem - needs a better quality test to sort this out. It would be better to see a properly exposed image that was not all squeezed to the left of the histogram as a basis to evaluate this issue.
Click to expand...
Click to expand...
The arw files are like... four stops underexposed!
 
I developed both in Capture One Pro using everything turned off, or the same on both images, no NR, same white balance settings etc.

There is a difference, however for having been pushed 4 stops it's surprisingly small.

I do believe that there can be slight differences in cameras of the same model, perhaps it's too difficult to create every sensor exactly the same, just as there are differences in lenses.

It's also difficult to completely rule out the surroundings in this situation, the test needs to be more controlled.



Green Hue C1Pro
Green Hue C1Pro



Red Hue C1Pro
Red Hue C1Pro

What differences do you see now? Is it still the same?
 
Yeah, maybe sensors still differ slightly. I still see the difference though and I tried taking pictures in different surroundings and one still had a more red hue and the other one green.
 
Yes sorry, I know the files are underexposed but the difference was most noticeable in these pics. I took some others as well and the difference in hue was still there.

Even if they are underexposed, shouldn't they have the same hue and not one leaning toward red and the other one toward green in all the pictures I took?
 
alexdrake said:
Yes sorry, I know the files are underexposed but the difference was most noticeable in these pics. I took some others as well and the difference in hue was still there.

Even if they are underexposed, shouldn't they have the same hue and not one leaning toward red and the other one toward green in all the pictures I took?
Click to expand...
The white balance on the images are different. As noticed previously, one has a tint of 16, and the other 17.

How did you set the white balances on these cameras?

--
http://therefractedlight.blogspot.com
 
Last edited:
I did a custom white balance on a piece of paper and tweaked so both cameras were set to 2800K, A-B: 0 G-M: M3. Tested different white balance values as well but still the same result.
 
Last edited:
alexdrake said:
I did a custom white balance on a piece of paper and tweaked so both cameras were set to 2800K, A-B: 0 G-M: M3. Tested different white balance values as well but still the same result.
Click to expand...
So how did they end up with different tints, which would be the G-M value?

I compared the metadata in both images, and here is the relevant difference:
alexdrake said:
alexdrake said:
< <crs:Tint>+17</crs:Tint>
---
> <crs:Tint>+16</crs:Tint>
Click to expand...
Click to expand...
 
I also downloaded the files and I too see the same difference in the tint values in ACR.

When I move the exposure slider about 4 stops to the right in ACR the histograms for each image are more different from each other than what I would expect but the images look virtually identical.
 
Last edited:
Hm, that is weird. Perhaps I set a difference white balance in the photos I uploaded by mistake ... but I did set the white balance several times and I'm 100% sure that they were the same settings and yet the images still had different tints when I looked at them later.

What tool are you using to see the tint level metadata? I used ExifTool and couldn't seem to find a value for tint?
 
Last edited:
ANYTHING / everything that is made...either by human hands OR machine...has +/- tolerances.

No matter how carefully the detail parts are made, no matter how carefully those parts are assembled, said assemblies most likely will be different. Perform different in one matter or another.

You've heard the comment...don't buy a car (anything) made on Friday, or before a holiday...there's a good reason for that..!

Mike
 
alexdrake said:
Hm, that is weird. Perhaps I set a difference white balance in the photos I uploaded by mistake ... but I did set the white balance several times and I'm 100% sure that they were the same settings and yet the images still had different tints when I looked at them later.

What tool are you using to see the tint level metadata? I used ExifTool and couldn't seem to find a value for tint?
Click to expand...
I used Photoshop.

Did you edit these images at all on the computer? I'm wondering if the software you are using actually writes metadata back to the raw file instead of saving it to a sidecar file. This is usually frowned upon, but it does happen.
 
Just for brainstorming, could there be something going on with the camera sensor?

If the sensor on the rental is cleaned each time it is returned then it could have been cleaned a lot of times which might have either left some residue or removed a bit of coating.
 
alexdrake said:
Yeah, maybe sensors still differ slightly. I still see the difference though and I tried taking pictures in different surroundings and one still had a more red hue and the other one green.
Click to expand...
I think there is something else going on here. Maybe the rental has had a much harder life in terms of being knocked about and something internally is functioning as it should anymore.

On my screen onlyfreeman's edited images look 99% identical to me.

What difference do you see in onlyfreeman's 2 images?

When I also push both images by 4 stops in ACR the differences in the images are just about indistinguishable.

You can get the images from the 2 cameras to look just about indistinguishable with a bit of basic post processing, so I wouldn't get too hung up on the slight difference between the 2 cameras especially when one is a rental.

You could end up spending a lot more time trying to work out why there is a very slight difference in the 2 cameras than it would take to make them look virtually identical in PP.
 
Last edited:
David1961 said:
alexdrake said:
Yeah, maybe sensors still differ slightly. I still see the difference though and I tried taking pictures in different surroundings and one still had a more red hue and the other one green.
Click to expand...
I think there is something else going on here. Maybe the rental has had a much harder life in terms of being knocked about and something internally is functioning as it should anymore.

On my screen onlyfreeman's edited images look 99% identical to me.

What difference do you see in onlyfreeman's 2 images?

When I also push both images by 4 stops in ACR the differences in the images are just about indistinguishable.

You can get the images from the 2 cameras to look just about indistinguishable with a bit of basic post processing, so I wouldn't get too hung up on the slight difference between the 2 cameras especially when one is a rental.
Click to expand...
These images are noisy. Sensor temperature and noise are strongly correlated. Perhaps a temp difference is involved too.
 
You must log in or register to reply here.

Similar threads

simonalexander
  • Question Question
Struggling with white balance on Q3 43
Replies
19
Views
6K
Rodenmg
Rodenmg
RobbieV
  • Question Question
Face Detect on OM-3 & OM-1.2 the same?
Replies
8
Views
5K
Skeeterbytes
S
DNature
Manual white balance issues sometimes when shooting video S5II
Replies
6
Views
1K
DNature
DNature
C
Sensor Performance -- Different generations and different systems
2 3 4
Replies
73
Views
3K
RealWorldPhoto
R
M
  • Question Question
Why does an Adobe RGB test image not show gamma compression on my monitor but does when printed
Replies
5
Views
275
JohnWheeler
J

Keyboard shortcuts

f
Forum
m
My threads
Mobile site
About
Editorial content
Cameras & Lenses
Community
All content, design, and layout are Copyright © 1998–2025 Digital Photography Review All Rights Reserved.
Reproduction in whole or part in any form or medium without specific written permission is prohibited.
When you use DPReview links to buy products, the site may earn a commission.
©GPS Media - Guides, Products, Services.
Back
Top