D810 flash photos are way too warm

[Cleanimage]

I've had my D810 since 2/15. I also have a D800. I've noticed that when I take flash shots using my SB910, I get a much more neutral WB from my D800. Under the same conditions with the same settings and same lens, the images I get from my D810 can be as much or more than 1,000 higher on the Kelvin scale as measured by NX-D.

I've attached 2 images taken a few minutes for illustration purposes. They were converted to JPEGs using NX-D. Both images are straight out their respective cameras using standard auto white balance, 1/30 sp, 250 ISO, 5.6 aperture, with the picture control set to neutral with +7 sharpening.

The D810 flash files almost always look too warm. Has anyone else run into this situation?

Any feedback would be appreciated.

Subscribe to see EXIF info for this image (if available)

Subscribe to see EXIF info for this image (if available)

 

[wilvoeka]

On my phone, to me the first shot looks much better. The second on seems too blue. I will give them another look when I get home.

 

[Cleanimage]

On my phone, to me the first shot looks much better. The second on seems too blue. I will give them another look when I get home.

Jim - Thanks. I appreciate your feedback. I'm working off a color calibrated 27" monitor. To me the D800 file does look a slightly too blue but not much. On the other hand, the D810 file looks way too warm. Anyway, I'll be interested in any additional feedback you give me once you get home and look at it on your monitor.

 

[Luke_Miller]

I wonder if the D810 is factoring in the illumination from the tungsten table lamp. At some point in my Nikon journey it seems to me that Auto White Balance changed from defaulting to Flash when I mounted the flash and, instead, starting to calculate a white balance based on the combination of the flash output and the ambient light.

 

[frdjohns]

If you shoot both cameras without flash at the same subject and same settings, is there a difference between the results?

 

[wilvoeka]

OK, on my calibrated monitor(a 100% Adobe RGB monitor) It still looks to me that the second is to blue, the first is warmer, but the tones acroos the chair and dogs seem to match all over. Where as the second shot there seems to be areas where the tone shifts to blue.

 

[Cleanimage]

I wonder if the D810 is factoring in the illumination from the tungsten table lamp. At some point in my Nikon journey it seems to me that Auto White Balance changed from defaulting to Flash when I mounted the flash and, instead, starting to calculate a white balance based on the combination of the flash output and the ambient light.

Luke - Thanks for your input. Since 2006 I've had three Nikon digital cameras and the D800 meters differently with flash than the other two cameras. My D200 gave very warm flash results, the D800 is a bit cool, and the D810 is back to warm again but not quite as warm as the D200; go figure. I think that you are right that the 810 metering is picking up the lighting right next to the chair but I think that it may also be influenced by flash feedback from the predominate red/brown colors in the room.

 

[Cleanimage]

OK, on my calibrated monitor(a 100% Adobe RGB monitor) It still looks to me that the second is to blue, the first is warmer, but the tones acroos the chair and dogs seem to match all over. Where as the second shot there seems to be areas where the tone shifts to blue.

Jim - I agree with you that the D800 shot is a bit too cool. On the other hand, the D810 shot looks too warm and I was surprised at the difference. As I mentioned to Luke, maybe I shouldn't have been surprised because my D200's flash output was very warm. When I look at the images I see noticeably warmer tones for the dogs, the shawl, and head and armrests on the chair in the D810 shot. I'm thinking that the things that are illuminated the most by the flash may be rendered more warmly by the D810 meter.

 

[greyflash]

In the first shot it is apparent that the incandescent lamp has an effect on the overall kelvin look of the image. On the second shot and based on the reflection of the flash on the left arm of the chair I would say that the flash was brighter than on the first shot, hence the color would be less warm; i.e., the incandescent would not have the same effect as on the first image. stated in another way, it takes X amount of light to properly expose an image and the balance between incandescent and flash will vary based on the percent of light utilized from the flash to properly expose the image.

For a proper test, turn off the incandescent.

 

[Cleanimage]

If you shoot both cameras without flash at the same subject and same settings, is there a difference between the results?

Jim - The D810 shot is slightly warmer, i.e., +229 on the Kelvin scale as measured by NX-D.

 

[Cleanimage]

In the first shot it is apparent that the incandescent lamp has an effect on the overall kelvin look of the image. On the second shot and based on the reflection of the flash on the left arm of the chair I would say that the flash was brighter than on the first shot, hence the color would be less warm; i.e., the incandescent would not have the same effect as on the first image. stated in another way, it takes X amount of light to properly expose an image and the balance between incandescent and flash will vary based on the percent of light utilized from the flash to properly expose the image.

For a proper test, turn off the incandescent.

Gordon - Thanks for your input. I turned off the all lights in the room and took the following test shots. The camera settings were identical. According to NX-D, the kelvin was plus 1,198 for the D810 shot. With the lights on it was plus 1,115 for the D810 shot. Do you think that the D810 is calculating the exposure a bit differently and that results in a higher flash output and a warmer look?

Subscribe to see EXIF info for this image (if available)

Subscribe to see EXIF info for this image (if available)

 

[TonyW]

I think you also need to take into account the differences between cameras and the fact that there is nothing that says that Nikon management cannot change appearance of images between models if they feel it makes improvement.

Did you set Flash WB on both shots?

It could be any of the factors pointed out relating to flash but in addition and my suspicion is the differences could simply be down to any one or combination of the following:

Different Sensor in D810 to the D800
No Low Pass Filter in D810 yes in the D800 plus
Different processing EXPEED 4 vs EXPEED 3
Different algorithm in NX D to account for new sensor

 

[greyflash]

Your turning off the lights kind of shot my theory down. I am out of ideas other than what Tony said above.

 

[Palouse]

Different algorithm in NX D to account for new sensor

Now that's one I'd never had thought of by myself!

 

[Cleanimage]

I think you also need to take into account the differences between cameras and the fact that there is nothing that says that Nikon management cannot change appearance of images between models if they feel it makes improvement.

Did you set Flash WB on both shots?

It could be any of the factors pointed out relating to flash but in addition and my suspicion is the differences could simply be down to any one or combination of the following:

Different Sensor in D810 to the D800
No Low Pass Filter in D810 yes in the D800 plus
Different processing EXPEED 4 vs EXPEED 3
Different algorithm in NX D to account for new sensor

Tony - For the shots I posted I had the WB set to auto in both cameras. According to an article on the Nikon site about setting WB, "Flash adds a more aggressive touch of warmth to take the edge off the bright light of a flash;" I think that I did test out Flash WB one time and, as I recall, the shots were actually warmer than when I used Auto WB.

Anyway, I think that you are right and that it is just a difference in the models. In this case I don't think NX-D has anything to do with it since I get the same results when I look at the NEF images in LR, i.e., the D810 flash NEF is much warmer than the D800 flash NEF.

Regardless, I always shoot NEFs so it is simple to dial the WB back if I don't like the look. Without the flash, I think that the D810's Auto WB does a pretty good job. Go figure, but since the flash results are consistent at least I know what I'm dealing with when I start editing. Thanks for your input.

 

[Palouse]

This has been a most informative discussion. Thanks.

 

[GonzoBernelli]

Did you make a PRE white balance with flash on each camera?

 

[Cleanimage]

Did you make a PRE white balance with flash on each camera?

Nick - No I didn't. I was just concerned about the differences I was getting with the D800 versus D810. However, I'll take a look at using that technique. Thanks.

 

[Cleanimage]

For those of you who followed this thread and responded, I also contacted Nikon about the results I was getting. This is their response after testing both cameras in their lab using flash. It was nice to get a response from Nikon confirming that the issue I'm seeing is normal. Now I just need to keep working on my technique and composition so I can get some good shots!

"What I have found is that the images from our D810 were slightly warmer, richer and truer in color, slightly darker as well, than my images taken with the D800 in the exact menu settings environment with the flash. I then did a full reset and have taken similar images in standard settings on both cameras with similar results. The D800 lightened the orange color I had focused on more so than the D810 which brought the warm, orange color out, even in standard settings. This would most likely indicate that the results were different between the D810 and D800 cameras with similar settings because I have experienced the same results with our equipment."

 

[RickW]

I was interested in this thread because I often shoot my grandson with the 810 and an SB900. I have been thrilled with the results from this setup and always marvel at just how good the Nikon algorithms are. I shot with an 800 before I traded it for the 810, but never side by side so I never noticed any differences. But since you shoot raw and process the shots through NX (I use Camera Raw), as you suggested, does it really matter? I usually shoot "for fun" shots using Auto WB in the camera and then tweak in post. To be honest, I seldom examine the original too closely since I always set the balance to where I like it when converting to tiff or jpeg. But it does seem that the 810/900 combo hits the mark pretty close SOC. Minor adjustments are generally all that are necessary. By the way, I find the same holds true when shooting outdoors with fill flash. I really don't look at the Kelvin readings - I just adjust the shot until I like what I see. On occasion, since my grandson is very fair with very blue eyes, I find that the shot benefits from some warmth. Other times - not so. Depends on the shot and what moves me at the time.
Some examples below.

bit on the warm side since I wanted the orange and yellow in the truck

Subscribe to see EXIF info for this image (if available)

A bit on the cool side - this is a more true representation of his coloring, but it definitely has a blue cast probably because of the blue sweater and the bluish chair.

Subscribe to see EXIF info for this image (if available)

Outdoor fill - this is exactly what his coloring looks like. And there were almost no tweaks to SOC as far as WB goes. Outdoor sun and bit of fill.

Subscribe to see EXIF info for this image (if available)