I noticed this strange color banding in some of my recent interior real estate shots and am wondering if you guys knew what it is or why it is there. Look at the ceiling.

Shot with D3, 14-24 f2.8 and the SB800.

The first shot is the jpg made from the raw file right out of the camera, the second is the same shot cropped with the saturation cranked up so the banding is more noticable. Didnt seem to matter if the raw file was converted with Photoshop, Lightroom or Nikon NX.

i've seen this before with all my cameras, looks like some sort of moire like pattern... i have no idea tho.

Not sure, but, could this be caused by ambient light, maybe from outside? I'd re-shoot with blinds closed. This doesn't look like banding to me. I've never seen it curved and that wide.

Looking at it again, I'm wondering if the lighting fixture might be the culprit.

I think what you are seeing is real. It's the combined effects of the artificial lights and the outside light coming in.

James R wrote:
Looking at it again, I'm wondering if the lighting fixture might be the culprit.

I am seeing this in other rooms without the fixture. There is another shot I took wihout the flash and did notice the color bands where the light from a window coming in and hitting the ceiling.

Still doesnt explain what it is and why it is happening.

Is it digital or optical?

digital, not optical. It is posterization.

"portions of the original image that presented gradual transitions are replaced by abrupt changes in shading and gradation from one area of tone to another"

and

"Unwanted posterization, also known as banding, may occur when the color depth, sometimes called bit depth, is insufficient to accurately sample a continuous gradation of color tone. As a result, a continuous gradient appears as a series of discrete steps or bands of color — hence the name".

Yupp'ers!

I think it's posterization too. I've seen it a few times, mostly with long exposure shots in the shadows.

Hmmm .... I think Patrick and Firewire are right.

Was this 12 bit raw? If so, try again with 14 and see if it changes.

Most definitely posterization. 14bit might help.

Erik Moore wrote:
Most definitely posterization. 14bit might help.

It is most definately 14 bit, uncompressed raw.

Shot in Adobe RGB with the same results.

my bet is that it is one color channel in the ceiling (of course not entirely, but dominating the pixel detail will be one channel), and it is lit by primarily tungsten, which will underexpose a cameras meter, and since you are already trying to hold highlights, you are leaving the ceiling to an exposure of somewhere around "zone 3" about 20 percent on the histogram.

This area has a small amount of detail recorded compared to the highlights (see linear vs. non linear recording), and thus has few shades to use, creating, again, posterizing. Don't know what I would do about it, probably sub the lighting in the chandelier with daylight type lighting.

if you wanna send me the Raw i would like to look into it and give my 2 cents

jemcfadden(at)gmail.com


J

agreed

oopaddy@gmail.com

patrickphoto wrote:
digital, not optical. It is posterization.

"portions of the original image that presented gradual transitions are replaced by abrupt changes in shading and gradation from one area of tone to another"

and

"Unwanted posterization, also known as banding, may occur when the color depth, sometimes called bit depth, is insufficient to accurately sample a continuous gradation of color tone. As a result, a continuous gradient appears as a series of discrete steps or bands of color — hence the name".

Yupp'ers!
Yup... Not Banding, It is Posterization

jmcfadden wrote:
if you wanna send me the Raw i would like to look into it and give my 2 cents

jemcfadden(at)gmail.com


J


On the way, thanks.

patrickphoto wrote:
my bet is that it is one color channel in the ceiling (of course not entirely, but dominating the pixel detail will be one channel), and it is lit by primarily tungsten, which will underexpose a cameras meter, and since you are already trying to hold highlights, you are leaving the ceiling to an exposure of somewhere around "zone 3" about 20 percent on the histogram.

This area has a small amount of detail recorded compared to the highlights (see linear vs. non linear recording), and thus has few shades to use, creating, again, posterizing. Don't know what I would do about it, probably sub the lighting in the chandelier with daylight type lighting.


I tried duplicating this in another environment with no incandescant lighting just the flash with the same results.

Is this showing up on your RAW file? If not, it may simply be a result of JPEG compression - especially a posted one that may further compress your JPEG. I don't know how much the FM site compresses images that are posted.

Bill
photo-synthesis

this is in response to the RAW file I recieved from the OP.

This is definatley posturizing. if it were truely banding, then some amount of noise added in PS would minimize the effect. It doesn't.

Unless this issue comes up in IDENTICALLY the same pattern image after image, then it is not the camera being "malfunctioning".

Test different color lights, and see if you can isolate the channel that is weakest here.

OH: are these lights CFL's?

patrickphoto wrote:
this is in response to the RAW file I recieved from the OP.

This is definatley posturizing. if it were truely banding, then some amount of noise added in PS would minimize the effect. It doesn't.

Unless this issue comes up in IDENTICALLY the same pattern image after image, then it is not the camera being "malfunctioning".

Test different color lights, and see if you can isolate the channel that is weakest here.

OH: are these lights CFL's?

The lights dont make a difference and I dont know what they are.

The lights do make a difference if they are cfls cycling at 60 hertz (correct my number if I am wrong, someone). This would effect the light fall off of the bulbs in a crazy way when four are different distances from the wall, and consequently the camera. The light is much weaker from the furthest bulb, and with a flickering, can create a very interesting lighting situation that will effect localized exposure fluctuations from shot to shot, and play havoc on different color channels, especially the yellow green, RIGHT where you are having this problem

it is muddy light , blue light from indirect sunlight casting down on warm , yellow woods , mixed with daylight from a flash and from either tungsten warm yellow lights in the fixture or worse CFLs with their terrible CRI values. And All of it showing up on the ceiling

I took the file into Photoshop and cleaned it up after rendering it in NX and turning off the D-lighting and setting the Picture Control to D2x Mode2 and reducing the saturation a bit and then a setting of 20 on the Shadow Protection slider

then in CS3 i used an adjustment layer on the ceiling a Hue / Saturation adjustment to the Yellows and hit the + on the dropper and moved it around on the ceiling and then reduced the saturation -20 and it is really fine and neutral now with little evidence of the issues

this is not really something any camera can render this mixed light is a bugger , you should have seen what film would have done here to get a real look at Bad

J

emreese wrote:
patrickphoto wrote:
this is in response to the RAW file I recieved from the OP.

This is definatley posturizing. if it were truely banding, then some amount of noise added in PS would minimize the effect. It doesn't.

Unless this issue comes up in IDENTICALLY the same pattern image after image, then it is not the camera being "malfunctioning".

Test different color lights, and see if you can isolate the channel that is weakest here.

OH: are these lights CFL's?

The lights dont make a difference and I dont know what they are.

Oh boy the lights make a HUGE difference..........

yup. exactly as jmcfadden says.

jmcfadden wrote:
emreese wrote:
patrickphoto wrote:
this is in response to the RAW file I recieved from the OP.

This is definatley posturizing. if it were truely banding, then some amount of noise added in PS would minimize the effect. It doesn't.

Unless this issue comes up in IDENTICALLY the same pattern image after image, then it is not the camera being "malfunctioning".

Test different color lights, and see if you can isolate the channel that is weakest here.

OH: are these lights CFL's?

The lights dont make a difference and I dont know what they are.

Oh boy the lights make a HUGE difference..........