Color change from D7000 to D7100

Started 7 months ago | Discussions
Bernard Delley
Regular MemberPosts: 237Gear list
Like?
Color change from D7000 to D7100
7 months ago

These images were taken one after the other using the same lens, left D7000 right D7100. The raw images were extracted by ACR and assembled in PS without corrections. Conditions were under heavily overcast sky with Auto WB. The auto WB went a bit more blue-ish on the D7100, clearly visible in the skin colour. This may easily be adjusted in PP.

The most striking is the colour of the windbreaker: clearly blue according to both images. It appears that way in images consistently. To the eye the windbreaker is consistently and clearly "green"! Below is a version with the windbreaker corrected to the proper eyeball colour.

The hue shift needed is about -30 for the D7000 windbreaker colour to arrive at the eye observed blue green - left image below. For the D7100 the required hue shift is reduced to about -20 - not shown here. To conclude, I like the slight progress in colour differentiation coming with the D7100.

Nikon D7000 Nikon D7100
If you believe there are incorrect tags, please send us this post using our feedback form.
rkumar
Regular MemberPosts: 408
Like?
Re: Color change from D7000 to D7100
In reply to Bernard Delley, 7 months ago

Wow! So, the windbreaker actually looks like the left image on your second corrected photo. But it looks blue as in the 1st pic when the pic is taken with AWB and no color correction.

Do you know why the leaves in the back-ground did not change color when you did the hue correction?

Reply   Reply with quote   Complain
Bernard Delley
Regular MemberPosts: 237Gear list
Like?
Re: Color change from D7000 to D7100
In reply to rkumar, 7 months ago

rkumar wrote:

Do you know why the leaves in the back-ground did not change color when you did the hue correction?

I selected just the windbreaker in PS for the hue rotation.

Reply   Reply with quote   Complain
bakhtyar kurdi
Regular MemberPosts: 402
Like?
Re: Color change from D7000 to D7100
In reply to Bernard Delley, 7 months ago

I would like to know what was the real color of the jacket, in three pictures it looks blue, in the last one it is green.

Reply   Reply with quote   Complain
Pi lover
Regular MemberPosts: 377
Like?
Re: Reality: Pictures don't look "real".
In reply to Bernard Delley, 7 months ago

Bernard Delley wrote:

These images were taken one after the other using the same lens, left D7000 right D7100. The raw images were extracted by ACR and assembled in PS without corrections. Conditions were under heavily overcast sky with Auto WB. The auto WB went a bit more blue-ish on the D7100, clearly visible in the skin colour. This may easily be adjusted in PP.

The most striking is the colour of the windbreaker: clearly blue according to both images. It appears that way in images consistently. To the eye the windbreaker is consistently and clearly "green"! Below is a version with the windbreaker corrected to the proper eyeball colour.

I was talking to a friend the other day and pointed out pictures don't look like the actual scene you are photographing.  This confused him.  Colors are one aspect.  Blur/ bokeh is another.  Just choosing the part of the scene to show would be another.  Point in time/ context etc. etc. etc.

We aren't recreating "there" in many ways.

The blue jeans also show an overly saturated blue.  Did you have the saturation set to 0 on both examples and if you adjust settings can you get closer to true colors? (on both cameras)  That would save you a lot of PP time.

Grant.

Reply   Reply with quote   Complain
Mako2011
Mako2011 MOD
Forum ProPosts: 15,594
Like?
Adobe?
In reply to Bernard Delley, 7 months ago

Bernard Delley wrote:

These images were taken one after the other using the same lens, left D7000 right D7100. The raw images were extracted by ACR and assembled in PS without corrections. Conditions were under heavily overcast sky with Auto WB. The auto WB went a bit more blue-ish on the D7100, clearly visible in the skin colour. This may easily be adjusted in PP.

The most striking is the colour of the windbreaker: clearly blue according to both images. It appears that way in images consistently. To the eye the windbreaker is consistently and clearly "green"! Below is a version with the windbreaker corrected to the proper eyeball colour.

The hue shift needed is about -30 for the D7000 windbreaker colour to arrive at the eye observed blue green - left image below. For the D7100 the required hue shift is reduced to about -20 - not shown here. To conclude, I like the slight progress in colour differentiation coming with the D7100.

You might be seeing the difference in how Adobe Standard in ACR (which is actually different for the two cameras) is processing the image.

Here are two RAW nef's of the same cameras processed with Nikon software (no PP other than conversion).

If you make your NEF's available...might be cool to see what happens with the jacket with Nikon conversion and WB held constant. Very interesting. Thanks

-- hide signature --

My opinions are my own and not those of DPR or its administration. They carry no 'special' value (except to me and Lacie of course)

Reply   Reply with quote   Complain
Bernard Delley
Regular MemberPosts: 237Gear list
Like?
Re: Adobe? no, VX too.
In reply to Mako2011, 7 months ago

It is not a WB issue. The windbreaker colour is wrongly differentiated in both cameras. Contrary to the leaves, this bluish green records as a colour with almost no green value. Actually the 'bluish-green' field in the upper right of the ColorChecker has a similar blue shift issue. I find it slightly less spectacular, as I would tend to call this field blue-green.

Even the WB setting to shade in VX will not get it to look "green". Not even shade 9090K and tint 12. That certainly makes the face look very sick green.

Reply   Reply with quote   Complain
Mako2011
Mako2011 MOD
Forum ProPosts: 15,594
Like?
Cool
In reply to Bernard Delley, 7 months ago

Bernard Delley wrote:

It is not a WB issue. The windbreaker colour is wrongly differentiated in both cameras. Contrary to the leaves, this bluish green records as a colour with almost no green value.

Actually the 'bluish-green' field in the upper right of the ColorChecker has a similar blue shift issue. I find it slightly less spectacular, as I would tend to call this field blue-green.

Even the WB setting to shade in VX will not get it to look "green". Not even shade 9090K and tint 12. That certainly makes the face look very sick green.

Do you have ADL on in  either cam by chance? Any change you can make the Nef's available. Thanks.

-- hide signature --

My opinions are my own and not those of DPR or its administration. They carry no 'special' value (except to me and Lacie of course)

Reply   Reply with quote   Complain
Bernard Delley
Regular MemberPosts: 237Gear list
Like?
only images of reality
In reply to Pi lover, 7 months ago

Pi lover wrote:

Bernard Delley wrote:

These images were taken one after the other using the same lens, left D7000 right D7100. The raw images were extracted by ACR and assembled in PS without corrections. Conditions were under heavily overcast sky with Auto WB. The auto WB went a bit more blue-ish on the D7100, clearly visible in the skin colour. This may easily be adjusted in PP.

The most striking is the colour of the windbreaker: clearly blue according to both images. It appears that way in images consistently. To the eye the windbreaker is consistently and clearly "green"! Below is a version with the windbreaker corrected to the proper eyeball colour.

I was talking to a friend the other day and pointed out pictures don't look like the actual scene you are photographing. This confused him. Colors are one aspect. Blur/ bokeh is another. Just choosing the part of the scene to show would be another. Point in time/ context etc. etc. etc.

We aren't recreating "there" in many ways.

The blue jeans also show an overly saturated blue. Did you have the saturation set to 0 on both examples and if you adjust settings can you get closer to true colors? (on both cameras) That would save you a lot of PP time.

of course one would expect from a basic colour photograph that the colours look correct. That is the case to a surprising degree. And for demanding people, there is colour management to push this colour correctness to a higher level.

This fun object reminds me that things can go wrong, because the camera does not exactly reproduce the tri- stimulus spectra of the eye. I believe this windbreaker has a narrow spectrum between green and blue, which happens to fall clearly to the blue side by the CFA of the camera.

I do not feel a categorical imperative to PP to 'correct' colours.

I normally us SD or NL settings on the camera, albeit that should not matter with raw files.

Reply   Reply with quote   Complain
Bernard Delley
Regular MemberPosts: 237Gear list
Like?
Re: Cool, NEF to play
In reply to Mako2011, 7 months ago

http://homepage.swissonline.ch/delley/data/DSC_3911.NEF

remember, the colour seen by eye is shown with reasonable approximation in the original post. The manipulated left part of the lower composite image.

Reply   Reply with quote   Complain
j_photo
Senior MemberPosts: 1,166Gear list
Like?
Re: Color change from D7000 to D7100
In reply to Bernard Delley, 7 months ago

Bernard Delley wrote:

These images were taken one after the other using the same lens, left D7000 right D7100. The raw images were extracted by ACR and assembled in PS without corrections. Conditions were under heavily overcast sky with Auto WB. The auto WB went a bit more blue-ish on the D7100, clearly visible in the skin colour. This may easily be adjusted in PP.

The most striking is the colour of the windbreaker: clearly blue according to both images. It appears that way in images consistently. To the eye the windbreaker is consistently and clearly "green"! Below is a version with the windbreaker corrected to the proper eyeball colour.

The hue shift needed is about -30 for the D7000 windbreaker colour to arrive at the eye observed blue green - left image below. For the D7100 the required hue shift is reduced to about -20 - not shown here. To conclude, I like the slight progress in colour differentiation coming with the D7100.

IMO, the image is filled with green which is throwing off the camera's color balance. Shoot the same guy, jacket, etc.  but with a neutral color background and see what happens.

 j_photo's gear list:j_photo's gear list
Nikon D800 Nikon Df
Reply   Reply with quote   Complain
GerardHaines
Regular MemberPosts: 417Gear list
Like?
Color accuracy of the jacket aside...
In reply to Bernard Delley, 7 months ago

...what I find interesting is that the D7000 image on the left actually looks better overall than the D7100 on the right.  Perhaps that's just my perception?

 GerardHaines's gear list:GerardHaines's gear list
Nikon D7000 Nikon D5100 Nikon AF-S DX Nikkor 18-55mm f/3.5-5.6G VR Nikon AF-S DX Nikkor 55-300mm f/4.5-5.6G ED VR Tokina AT-X Pro 11-16mm f/2.8 DX II +1 more
Reply   Reply with quote   Complain
Bernard Delley
Regular MemberPosts: 237Gear list
Like?
Re: Color accuracy of the jacket aside...
In reply to GerardHaines, 7 months ago

GerardHaines wrote:

...what I find interesting is that the D7000 image on the left actually looks better overall than the D7100 on the right. Perhaps that's just my perception?

I agree, I like the skin colour in the left image better. But, that is a WB issue easily fixed in PP. Here the D7000 through ACR looks nicer.  In VX there is much less difference in the skin tone.

Reply   Reply with quote   Complain
chary zp
Regular MemberPosts: 432Gear list
Like?
the nef - what is all this sorcery?
In reply to Bernard Delley, 7 months ago

Bernard Delley wrote:

http://homepage.swissonline.ch/delley/data/DSC_3911.NEF

remember, the colour seen by eye is shown with reasonable approximation in the original post. The manipulated left part of the lower composite image.

I've just opened the nef file in nikon software. I assume the line of colour squares second from the bottom are BGRYMC primary colours.

With WB set by Grey Point to the grey squares on the bottom line, the primary colours look quite believable, with the Cyan square at H° 193-ish and the jacket basically the same.

If the jacket is supposed to be greenish like in your original post in the second picture on the left, why arent the other green squares also deformed in the terms of colour?

 chary zp's gear list:chary zp's gear list
Nikon Coolpix P7700 Nikon D90 Nikon D200 Nikon D7100 Nikon AF-S DX Nikkor 18-105mm f/3.5-5.6G ED VR +8 more
Reply   Reply with quote   Complain
Mako2011
Mako2011 MOD
Forum ProPosts: 15,594
Like?
Adobe ve ViewNX2
In reply to Bernard Delley, 7 months ago

Bernard Delley wrote:

http://homepage.swissonline.ch/delley/data/DSC_3911.NEF

remember, the colour seen by eye is shown with reasonable approximation in the original post. The manipulated left part of the lower composite image.

Thank you very much for the D7K NEEF.  I noticed it was shot in WB Auito A2. The difference in how Adobe ACR and ViewNX2 convert is striking. I also converted using WB set to Shade in both programs and again...a huge color difference, Adobe's inability to interpret the Nikon picture controls makes for completely different renderings. Pretty cool

Do you have the D7100 NEF as well?  Thank you

-- hide signature --

My opinions are my own and not those of DPR or its administration. They carry no 'special' value (except to me and Lacie of course)

Reply   Reply with quote   Complain
Stephan Def
Regular MemberPosts: 287Gear list
Like?
Re: Color change from D7000 to D7100
In reply to Bernard Delley, 7 months ago

There is one very basic question I have to you. Why are you using the Jacket instead of the color checker passport to verify your colors?

It seems to me it is more sound just to see if the squares on the color checker passport have the colors that they should have. This can even be done with computer software by selecting on the squares with color picker software and then checking the resulting values against the RGB color tables.

The jacket is of a somewhat undefined color, and that just makes everything harder.

Furthermore to be sure you have the correct white balance you can spot white balance also on the Color checker passport, the 4th patch from the white one in the bottom row of the ColorChecker classic is 18% gray.

 Stephan Def's gear list:Stephan Def's gear list
Sony Alpha DSLR-A200 Nikon D5100 Nikon AF-S DX Nikkor 55-200mm f/4-5.6G VR Nikon AF-S DX Nikkor 18-55mm f/3.5-5.6G VR
Reply   Reply with quote   Complain
Bernard Delley
Regular MemberPosts: 237Gear list
Like?
Re: Color change from D7000 to D7100
In reply to Stephan Def, 7 months ago

Stephan Def wrote:

There is one very basic question I have to you. Why are you using the Jacket instead of the color checker passport to verify your colors?

It seems to me it is more sound just to see if the squares on the color checker passport have the colors that they should have. This can even be done with computer software by selecting on the squares with color picker software and then checking the resulting values against the RGB color tables.

The jacket is of a somewhat undefined color, and that just makes everything harder.

Furthermore to be sure you have the correct white balance you can spot white balance also on the Color checker passport, the 4th patch from the white one in the bottom row of the ColorChecker classic is 18% gray.

I think the color checker colors are all quite OK. They might come closer by finetuning WB etc.  There are slight color shifts with the usual suspects: skin color, bluish green and red orange.

The fun thing is that the jacket goes wildly off in color. You cannot check this just by the raw. My original post contains an attempt to communicate the actual color as seen by eye. You basically have to take my word, as I cannot show you that jacket. Your store may have something with that pigment. The blue-green patch in the ColorChecker gives a hint on a similar, but less dramatic shift. If you have a ColorChecker, you can verify this part.

Reply   Reply with quote   Complain
Mako2011
Mako2011 MOD
Forum ProPosts: 15,594
Like?
Set
In reply to Bernard Delley, 7 months ago

Bernard Delley wrote:

Stephan Def wrote:

There is one very basic question I have to you. Why are you using the Jacket instead of the color checker passport to verify your colors?

It seems to me it is more sound just to see if the squares on the color checker passport have the colors that they should have. This can even be done with computer software by selecting on the squares with color picker software and then checking the resulting values against the RGB color tables.

The jacket is of a somewhat undefined color, and that just makes everything harder.

Furthermore to be sure you have the correct white balance you can spot white balance also on the Color checker passport, the 4th patch from the white one in the bottom row of the ColorChecker classic is 18% gray.

I think the color checker colors are all quite OK. They might come closer by finetuning WB etc. There are slight color shifts with the usual suspects: skin color, bluish green and red orange.

The fun thing is that the jacket goes wildly off in color. You cannot check this just by the raw. My original post contains an attempt to communicate the actual color as seen by eye. You basically have to take my word, as I cannot show you that jacket. Your store may have something with that pigment. The blue-green patch in the ColorChecker gives a hint on a similar, but less dramatic shift. If you have a ColorChecker, you can verify this part.

You have to set the WB using the 18% Color Checker grey block in the RAW file to accurately set the WB before "The blue-green patch in the ColorChecker gives a hint on a similar, but less dramatic shift" for those who don't own the card.

Again, Adobe seems to treat your WB auto 2 setting dramatically different than ViewNX2 or the in camera JPEG engine does.

-- hide signature --

My opinions are my own and not those of DPR or its administration. They carry no 'special' value (except to me and Lacie of course)

Reply   Reply with quote   Complain
Nexu1
Senior MemberPosts: 1,710
Like?
Re: Set
In reply to Mako2011, 7 months ago

Mako2011 wrote:

Bernard Delley wrote:

Stephan Def wrote:

There is one very basic question I have to you. Why are you using the Jacket instead of the color checker passport to verify your colors?

It seems to me it is more sound just to see if the squares on the color checker passport have the colors that they should have. This can even be done with computer software by selecting on the squares with color picker software and then checking the resulting values against the RGB color tables.

The jacket is of a somewhat undefined color, and that just makes everything harder.

Furthermore to be sure you have the correct white balance you can spot white balance also on the Color checker passport, the 4th patch from the white one in the bottom row of the ColorChecker classic is 18% gray.

I think the color checker colors are all quite OK. They might come closer by finetuning WB etc. There are slight color shifts with the usual suspects: skin color, bluish green and red orange.

The fun thing is that the jacket goes wildly off in color. You cannot check this just by the raw. My original post contains an attempt to communicate the actual color as seen by eye. You basically have to take my word, as I cannot show you that jacket. Your store may have something with that pigment. The blue-green patch in the ColorChecker gives a hint on a similar, but less dramatic shift. If you have a ColorChecker, you can verify this part.

You have to set the WB using the 18% Color Checker grey block in the RAW file to accurately set the WB before "The blue-green patch in the ColorChecker gives a hint on a similar, but less dramatic shift" for those who don't own the card.

Again, Adobe seems to treat your WB auto 2 setting dramatically different than ViewNX2 or the in camera JPEG engine does.

-- hide signature --

My opinions are my own and not those of DPR or its administration. They carry no 'special' value (except to me and Lacie of course)

I downloaded the nef file and imported it into lightroom.  When I use the spot tool for white balance on the gray block the WB changes from 5700 & +40 tint to 6500 and +17 (roughly).  Which is a noticeable change.  But the jacket is still quite blue.  If I go into HSL and change the blue hue to -40 I get a relatively green jacket.  But that seems like a huge change to have to make to a RAW file.  Something seems weird.

Reply   Reply with quote   Complain
Bernard Delley
Regular MemberPosts: 237Gear list
Like?
Re: ACR vs ViewNX
In reply to Mako2011, 7 months ago

Again, Adobe seems to treat your WB auto 2 setting dramatically different than ViewNX2 or the in camera JPEG engine does.

Actually, in looking at the jpg from the nef I did not notice any striking difference between ViewNX and ACR. So for the simplicity of my workflow my assembly was done all in Adobe.

Could you post your two jpg drawn from the same nef and point out the differences?

Reply   Reply with quote   Complain
Keyboard shortcuts:
FForum MMy threads