What causes this?

I've seen this moire effect in clothing a couple of times in the last year:

HOSTED PHOTO
please log in to view hosted photos in full size.

Fingerprint on lens? lol

On a serious note, Did you have her wash the jean jacket before taking a picture of it? lol

Ok, ok on a serious serious note.. I think it has something to do with the pattern on the jean jacket and reflecting back into the lens.

How do we know whether this was a camera image or a scanner image? I'm inclined to think it's from a scanner.

In which case, I'm wondering if the original was a screened print, in which case you could be having problems with moire' patterns.

Not exactly a big surprise, since you have named the file moire.jpg.

Are you trying to play games with us?

Nope, it's definitely a camera image.

I'm in the offset printing industry, so I'm brutally familiar with moire, the need for descreening, etc.

I'm certainly not trying to play games with anybody - if there's a way the photographer (not me) could avoid this, I'd really like to know it.

Taken with a Nikon D2X, at 400ISO, 2 days ago.

There's your problem...nikon. lol.

It's true that I've only ever seen the problem in Nikon images. Do you think that could be the reason?

Maybe its a ghost.........Wait...I think I see the Virgin Mary!!

I've seen this happen with Canon as well...
Shoot less sharp (seriously).
It's caused by the interference of the pattern of the fabric with the pattern of the sensor. In the same way that moire used to be caused by interfering grids. (Or whatever the english word is)...

On some black stockings I have been able to PP it out using channel mixer to get the luminouscity equalised, then use another adjustment layer to locally bring the color to what it should be.
Tough job.

I guess that accounts for why no camera was identified in the EXIF data.

If it really is moire', then Rene' is quite correct. After considerable experimentation with different scanning resolutions, I came to the conclusion that moire' is best avoided by scanning at a pitch that is approximately equal to or less the inherent dot pattern of the subject. In other words, Nyquist criteria (says to sample at twice the maximum frequency in the signal) does not apply and gets you in trouble. So, doing the same thing with a camera as what one would do in scanning ought to work.

I remember this being a 'feature' of Nikon DSLR's a couple of years ago, something to do with the type of anti-alias filter that sits over the CCD chip. I believe the images came out of the camera a bit sharper but this was the trade-off.
I've never had it with my 20D.

I've had this happen with my 1D MKII and 50mm F1.4

I've seen it happen with Oshkosh overalls - those thin blue & white stripes.
Took three shots of the same subject from the same position while zooming to different focal lengths. The moire only happened on the medium focal length image, though all of them were plenty sharp.

So I came to the same conclusion as Bob Lay - a slightly different resolution relative to the width of the stripes will solve the problem. But the problem is, you never see it until the shoot is over and the subject is gone.

Thanks for all your help, everyone. I'll have a talk to the photographers about it.