Can you remove moire in CS2?

[boxofrocks]

Originally Posted by Joe_Lorenzini

D2X? Tom, did Christmas come a little early this year?

No, I wish. It's an NPS camera. My D200 is in Torrance.

I'm holding out for the D3 (March?).

[boxofrocks]

Originally Posted by huxley

ok so i want to know. is it the weave of his suit that is creating the moire effect? or is it a deadly combo of suit and camera with a weakness to moire? i have never run into this yet, but i know with winter coats and sweaters coming out of the closets, this is going to become a bit more common this season. is there anything that can be done IN CAMERA to minimize this in post production?

I think it's a combination of the pattern of "checks" in the weave, the weave of the fabric, and the lighting conditions.

D2X usually doesn't have moire this bad, to my knowledge--the planets just aligned on a couple of shots. Out of about 300 that night, these were the only 2 (and the only 2 I've ever seen like this).

[simon]

It is impossible to design a digital sensor that won't exhibit this under some conditions, at least not without noticeably adversely affecting IQ for all shots....

[Bartman01]

Originally Posted by huxley

ok so i want to know. is it the weave of his suit that is creating the moire effect? or is it a deadly combo of suit and camera with a weakness to moire? i have never run into this yet, but i know with winter coats and sweaters coming out of the closets, this is going to become a bit more common this season. is there anything that can be done IN CAMERA to minimize this in post production?

It is the interaction with the weave in his suit, the grid pattern of the sensor, and bayer interpolotion of the pixel data. Most sensors are laid out like this:

GBGBGBGB
RGRGRGRG
GBGBGBGB
RGRGRGRG

Where:
R = photosite that captures Red color data
B = photosite that captures Blue color data
G = photosite that captures Green color data.

A process known as bayer interpolation is then used to determine the 'true' color that really exists at each photosite. The problem you have is when the detail and the sensor grid interact in a weird way that produces the color pattern. Foveon sensors eliminate this because they capture R+B+G at EVERY pixel site.

On top of that, you have the interaction between the fine weave in the clothing and the fine grid pattern of the sensor. This is what produces that 'wavy' pattern in the image that remains after the color is corrected.

To see an example of this, hold two window screens together and rotate one of them while looking through it. This is essentially what is happening to cause the 'wavy' pattern.

The only 'fix' is to try different camera to subject distances or different focal lengths to find one that minimizes or eliminates the interaction.

[huxley]

sweet explanation bartman. i like the window screen example. makes sense to my itty bitty brain.