Moiré, Z6 + 70-200mm/f2.8 S

Moiré has not been an issue I encountered very often. Previously, once I was in Brazil photographing a blue and gold macaw and ran into this problem on the feathers, but that was way back in 2007, with a D2X and a 70-200mm/f2.8 AF-S VR.

 

Yesterday, I was photographing a company Christmas party and a gentleman's shirt turns out to be very moiré prone. However, it was fine from a few specific angles. This time I was using a Z6, which has an anti-aliasing filter, and the Z version of the 70-200mm/f2.8 S. I was using that lens at 200mm and wide open at f2.8.

 

 

 

Blue and Gold Macaw. D2X + 70-200mm/f2.8 AF-S VR, version 1

 

I got this once with my Z6 on a photo of a barn that had a screen in a window. It took me a bit to notice it as it wasn't as pronounced as yours. I think I used a fringe filter in Affinity Photo that took care of most of it.

What happens if converted to B&W? (just tried it on your photo, gets worse)

What happens if converted to B&W? (just tried it on your photo, gets worse)

Converting to B&W will not remove the patterns, although it could be less noticeable without the weird colors.

 

Can you see the patterns in viewfinder or rear screen?

I managed to get a good mono from a similar fabric as Shun's shirt via the PS Black and White conversion set-up. As you can change the tone of each moire colour separately, you can balance it.

Too sharp a lens. One may need a 100 MP or maybe 200 MP sensor to avoid moire ... :-)

I didn't notice anything abnormal until I reviewed the images on a larger monitor. You can see the moiré on the rear LCD if you pay attention to the image review. Again, this is all on a Z6 with the AA filter, which I have been using for three years. What is new to me is the 70-200mm/f2.8 S, and I was using moderate shutter speeds, 1/125 and 1/250 or so. Maybe that lens is indeed too sharp and VR/IBIS is working well.

 

Here is one more example. Interestingly, in this case the moiré only appears in the front part of the sleeve. The rest of the shirt, while still in focus, is not affected.

 

The rest of the shirt, while still in focus, is not affected.

I guess the sensor only 'makes' the moire at/on the plane of focus and the shirt infront and behind are technically within the Dof but not so razor sharp. Maybe?

 

It's very curious that the 3rd shot has NO moire at-all. Obviously the angle of difference between the weave alignment and the sensor were perfectly out of phase.

 

Looking here,

 

Nikon Nikkor Z 70-200 mm f/2.8 VR S review - Image resolution - LensTip.com

 

..it's a good job you weren't at 70mm!

I think Shun's 3rd shot shows no moire because the shirt isn't in focus anywhere.

Interesting. That's not good. I haven't noticed this in my limited experience with my 70-200 yet but then I haven't shot similar subjects at similar exposure. Maybe I should test it out. Edited December 17, 2021 by Mary Doo

In the 2nd image in my opening post, the gentleman was a bit at an angle to the camera, but yet moiré is very serious on the entire left side of his shirt.

 

In any case, as I said, the last time I had serious moiré was way back in 2007 with a completely different subject, although interestingly I was also using another 70-200mm/f2.8 Nikkor zoom. The main issue in common is that both subjects have some very fine patterns. As long as I won't see this problem in another 14 years, I am not going to be too concerned about it. Back in 2012 with my brand new D800E, I tried all sorts of subjects such as feather, fabric, architecture ... and I had a hard time producing moiré on purpose.

 

The 70-200mm/f2.8 S is certainly a very sharp lens. It also reminds me that depth of field at 200mm, f2.8 is quite shallow in the middle distances.

 

Edited December 17, 2021 by ShunCheung

The 70-200mm/f2.8 S is certainly a very sharp lens

I think it looks a bit soft at the top left corner.....:p :D

I think it looks a bit soft at the top left corner.....:p :D

I am afraid that the anit-moiré software I used is very effective. o_O

Hi, I don't really see a specific question, but there seems to be a lot of misunderstanding about what's going on.

 

First, this is not really about a specific lens, it's about a (any) high-quality lens at f/2.8 (or wider), AND a specific magnification of that shirt.

 

Back near 20 years ago I was doing some pretty substantial testing of this sort of thing, on a number of digital cameras that were being considered for the place where I worked. We did mass market portrait work, so obviously every type of fabric was gonna appear sooner or later, and there were a variety of magnifications, so we were especially concerned about moire (aliasing). And because of the very-high volume, virtually no moire would be acceptable - the costs of hand-reworking are too high.

 

What I largely ended up doing was to use a blue pinpoint Oxford shirt as the base test target. (If a camera was gonna produce moire, this was nearly infallible as a test target.) What I would do is to first measure the thread pattern of the fabric (using a magnifier with measuring reticle), then get the pixel pitch of the sensor (from specs). Then I'd estimate how large the shirt must be, roughly, in the frame, in order for the thread pattern to roughly match the pixel pitch.

 

Then, a series of test shots scattered around that size range to see if moire would show up. (This was done in the aperture range that we might foreseeable use.) In fact, the tests were not unlike Shun's, except probably better controlled with respect to magnification. As a note, having the shirt wrap around the body, or to have wavy folds in it is helpful in that this gives some variety in the spacing (aka frequency) of the thread pattern. This means that you don't need an exact magnification; just get close and the variability in the wrapping and tilt of the fabric will take care of the rest.

 

From what Shun is seeing, I'd guess that the moire is just showing up in marginal situations. Many photographers today realize that stopping their lens down can limit the maximum resolving power (read up on Airy disc diameter, related to f-number). So in this case stopping down the lens a bit will most likely prevent this moire completely.

 

In this case I'm making an estimate that stopping down to about f/8 will likely get rid of any possible moire. F/5.6, maybe, maybe not. (I'm presuming your camera has about 24 MP on a "full-size" sensor, pixel pitch around 6 microns.) But if you carry out some more testing you can probably establish a breakpoint which will be "safe" for any potential moire situation. This should work for any lens - only the f-number is significant (provided the lens is high quality).

 

Fwiw the pixel pitch varies diagonally, meaning that it is possible to eliminate aliasing on vertical and horizontal architectural components, yet still get it on diagonal patterns, as in a conventional layout the diagonal pixel pitch is a little longer. And, on a Bayer-pattern array, the red and blue pixels are spaced wider and more sparsely. So if you photographed a detailed pattern in relatively "pure" red or blue colors, moire could still show up.

Stopping down to f/8 may not be sufficient. This is an MTF50 test of the Z 70-200/2.8 S with the Z5 body (which is 24 MP FX, like the Z6).

 

Nikon Nikkor Z 70-200mm f/2.8 VR S Review

 

The MTF50 values at f/8 are very similar to f/2.8 (though the highest frequencies may be attenuated more at f/8). What might work here is the use of a higher-resolution body (Z7/II/Z9) and stopping down to f/11. In studio lighting this is possible, but not in typical documentary / candid / low light situations. I'd still argue that we need higher-resolution sensors to use the new generation lenses correctly (by which I mean in such a way that the imaging process is faithful to the subject).

Edited December 18, 2021 by ilkka_nissila

This test

 

Nikon Z 70-200mm f2.8 VR S review | Cameralabs

 

shows that by stopping down from f/8 to f/11, the moire is significantly reduced in the star (test camera is 45 MP).

So, are you actually better off, in a known Moire scenario, focusing off and relying on DoF to make it look sharp enough?

Shoot at f/11 or f/16 and diffraction will blur photosite-level detail sufficiently that any moire disappears, or at least is greatly suppressed.

 

You can see some cross-hatching and noticeable colour banding in this map copy, due to interference between the pixel spacing and the half-tone dots.

 

Reducing the aperture a few stops gives a cleaner result with no hatching and less obvious banding.

 

Or, you could get the guy to stand still for a few seconds to take 4 pixel-shifted shots and get co-sited colour. :cool::p

Or, you could get the guy to stand still for a few seconds to take 4 pixel-shifted shots and get co-sited colour. :cool::p

@ say, 1/250th 'each', I guess the 4 shots, via mirrorless is what, 1/60th?

This test

 

Nikon Z 70-200mm f2.8 VR S review | Cameralabs

 

shows that by stopping down from f/8 to f/11, the moire is significantly reduced in the star (test camera is 45 MP).

That is not a practical solution. I certainly didn't pay for an f2.8 70-200 and carry that weight in order to use it at f8 and f11, which is way too slow indoors.

That is not a practical solution. I certainly didn't pay for an f2.8 70-200 and carry that weight in order to use it at f8 and f11, which is way too slow indoors.

That's why certain 'studios' forbid the wearing of any fabric that might moire.....!!

That is not a practical solution. I certainly didn't pay for an f2.8 70-200 and carry that weight in order to use it at f8 and f11, which is way too slow indoors.

So use a less sharp lens...?

 

I think that's nutts!

 

Use an old soft lens?

Moiré has been a problem for ages. What to do is either not care, or do what it takes to minimize, if not avoid it. And that means that you indeed avoid clothing that causes problems, and such. Or/and swallow your pride in owning an excellent lens or sensor, and do whatever it takes. As if there is any pride gained by producing horrible looking images using a large, heavy and expensive lens, right?

 

Avoiding it is, of course, best. So if possible, ask to change shirt, or put a jacket over it, or something. Much less trouble than having to deal with moiré in captured images. Getting rid of it in post is nigh impossible.