< ! --Digital window verification 001 -->

The new Nikon 1 J3 and S1 cameras do not have an anti-aliasing (AA) filter

Nikon-1-J3-sensor

Nikon 1 J3 sensor

A reader pointed me to this article by Fotosidan - according to a Nikon rep, the new Nikon 1 J3 and S1 mirrorless cameras do not have an anti-aliasing (AA) filter, just like the D800E:

"A little unexpected technological innovation is that both the new cameras do not have anti-aliasing filter, which should provide some sharper detail. According Matte Haglund, Nikon has learned much of D800E and now feel ready to skip the AA filter in these models."

The lack of AA filter will deliver sharper images with more details and better resolution and add moire patterns in certain situations.

This entry was posted in Nikon 1 and tagged . Bookmark the permalink. Trackbacks are closed, but you can post a comment.
  • Fry

    the Expeed 3(A) processor has probably enough processing power to fix moire as well..

    • Ola

      Since it theoretically impossible to do, you are wrong.

      • Fry

        fixing color aberration and distortions is “theoretically impossible” as well

        • Ola

          No, it’s not.

          • Wandrade

            Who is right? Please…make a point…

            • Micah Goldstein

              I’ll try to explain in plain English, but forgive me if this is still vague. Basically, you can’t reproduce the effects of an AA filter in post.

              Consider this: four pixels from a bayer array

              BG
              GR

              At Nyquist and above, it is possible that light hitting each of these elements it of completely different color and brightness. In such a case, information from adjacent pixels will not give you valid information about each other.

              However, with an AA filter in place, there is less chance of light only hitting one pixel. It’s not a 0% chance, because the birefringent materials used are not 100% efficient, and do let some light straight through (you still see false color moire and moire on cameras with AA filters, just less of it), And the materials are not equally efficient at all frequencies.

              Does this help?

              There will always be the potential for issue of moire patterns when using a pixel pattern with any sort of symmetry, when trying to capture an image of similar symmety. In other words, if you created a sensor with a radial pattern and used it to take a picture of something with radial symmetry at a slightly finer frequency than the sensor could reproduce, you’d get aliasing and moire.

              In the case of a Bayer pattern, you have a color patterned grid. So if you take a picture of a grid at near and above Nyquist, you’ll see color patterns, whether they’re there or not. If it had not color filter array, you’d just get black and white whorls.

              AA filters are good. But they ain’t cheap. But these cameras are cheap and they want to put crappy and or diffraction limited lenses on them, so there’s less of an issue. It’ll work just fine. Especially with higher resolution sensors at this size.

        • Ola

          Color aberration and geometrical distortions are theoretically 100% correctable.

          Morie comes from improper sampling, that is, false information is added, and that is impossible to remove.

          • Fry

            geometrical distortion is false information as well and correction requires interpolation – ie arbitrary pixel creation. In this sense it’s just as “uncorrectable” as moire.

            • http://www.facebook.com/jiri.dobry Jiří Dobrý

              You can measure the distortion and then use reverse function on the image. Moire is effect of measurement itself Ola is right.

              Sorry my bad english

            • Fry

              the point is – if we consider a low pass filter (AA) is a solution to moire, we might as well use blur in post processing to accomplish this same “fix”.

            • OhYeah

              It just can’t work that way … maybe the wikipedia can help you out:

              http://en.wikipedia.org/wiki/Aliasing

              The bandwidth of the input signal has to be lower than the bandwidth of the sampling device to avoid artifacts. The AA filter in a digital camera is there to make sure that’s true. No amount of “bluring” is going to fix what’s going on in the picture of a brick wall at the top of the article.

              Aberration is an entirely different phenomena. Again, maybe the wikipedia can help: http://en.wikipedia.org/wiki/Optical_aberration

              Or if you don’t have the time, this paragraph is closer to your point: http://en.wikipedia.org/wiki/Chromatic_aberration#Image_processing_to_reduce_the_appearance_of_lateral_chromatic_aberration

              The bottom line is that when the software knows the characteristics of the lens, aberration can be (partially) “fixed”. There’s no way for the software to know all the details about what caused a particular image to have moire

              Note to Jiri, you’re english is nearly perfect!

            • davidpeter

              No, we can’t. After we added the false information (overlapped the spectra), there is no mathematical way to distinguish it from the original spectrum. The only way to remove moire is to take a picture without moire and use that information…

            • Fry

              http://unsharp-mask.blogspot.sk/2007/08/moir-removal-in-photoshop.html

              anything Photoshop can do, expeed 3(A) can do as well. At least to a degree which is fully acceptable to the target consumers of the Nikon 1 system.

  • Maji

    I guess J2 is a still born then… I think Nikon is taking the mirrorless market seriously.

  • http://kyleclements.com/ Kyle Clements

    Question for the more technically minded photographers: Can moire be reduced/eliminated by simply de-focusing the shot subtly?

    • http://www.facebook.com/people/Florian-Hossfeld/100000909554473 Florian Hossfeld

      Yes, its actually a technique I use for videography, you can try defocusing subtely or using a high F stop in landscape / full,scenes situations, a high F stop (f16-22) will soften the image thanks to difraction and remove moire

    • Jer

      Another way to avoid moire is to either increase or decrease the distance to the subject.

  • AlphaTed

    Or target audience for these cameras won’t care about moire.
    Or won’t even know it’s there.

    • gsum

      You’ve either done the analysis and are sharing your wisdom or you’re making a fatuous comment. Somehow I think it’s the latter.

      • fancy

        Fatuous:
        fat·u·ous
        /ˈfaCHo͞oəs/

        Adjective
        Silly and pointless.Synonyms
        foolish – goofy – witless – idiotic

  • Spy Black

    Hmmm, so have they learned enough to release a DX body with no AA filter? ;-)

    • RMJ

      I cannot imagine such thing but D400E sounds quite possible.

      • Spy Black

        Why can’t you imagine that, considering they just did it with a crappy pocket camera.

        • RMJ

          Because D400E would bring more money in the house.

          • Spy Black

            I think you’re missing the point LOL

            • RMJ

              Propably. Care to share ?

            • Spy Black

              The point is apparently Nikon has decided they’ve advance this technology to the point that they can do this this to any camera they feel like, considering they just did so for a consumer P&S camera.

              And aside from that, considering they’re putting the technology in their hot selling P&S camera, it’s more likely an AA-less D7000 derivative would come out before a “D400″, because they’ll sell more of those than they would any “D400″ camera. After all, how many D300s were sold versus D7000s? There’s safety in numbers.

              So I would expect to see an “E” D7000 derivative if they ever decide to expand the AA-less product line. While they’re at it, a D600E would pretty cool too. ;-)

      • MJr

        Why a E at all, they might manage to just not need it.

  • ShowMeTheFuture

    Awesomeness; very cool that Nikon’s engineers are willing to take this gamble. I have a feeling that in 5 or 10 years we’ll all be shooting with either 1″ sensors or 8×10 inch sensors. Or, of course, Google glasses.

  • Up $#!t’s creek

    its on the lower end models… so i’d say its a cost saving measure thats been spun around

  • OhYeah

    Maybe you should think of the lens as the AA filter? When the sensor can out resolve the lens, there’s no aliasing, no moire. This is a sign that pixel density is reaching a practical limit.

    • thomasverbeke

      that could be a good reason indeed; I think it is also going to put an upper limit on the number of megapixels we will see on dx bodies because of the glass; or it would just be marketing

    • Ola

      Indeed it would, IF it could. But the sensor does not out resovle the lens.

  • HotDuckZ

    “Nikon has learned much of D800E and now feel ready to skip the AA filter in these models” sound like an idiot.

    • j v

      They can have learned a lot on how to process the raw data to avoid moire, and perhaps found ways of applying similar algorithms using lower spec’d processors. The research at the time was focused on the D800E, but some braekthroughs or accidental discoveries may have paved the way to also apply it in lower end.

    • An anonimous Photographer

      It’s a bit silly to say that the D800E has no AA filter, because it actually has one. it’s only an adjusted one. Besides in-camera processing is also ‘masking’ moire patterns

  • wincros

    Coolpix cameras and small sensor cameras from other makers have never had an antialiasing filter. There is no conclusion to leap to except that it is a small sensor camera.

  • French Fries

    Moire does not seem to be any problem on my D800E – Even not with wedding dresses.

    • Joe

      True, after 8 months and over 10.000 images with various topics from people and wedding to product shots and architecture with the D800E I’m still waiting to recognize my first moiré.

  • http://www.facebook.com/barefootintactivist Barefoot Intactivist

    The smaller the pixel pitch, the less need for an AA filter, right?

    • rkas

      No not really, it all depends on the number of megapixels actually.

      • MJr

        Chicken and the egg.

    • OhYeah

      You are right. Pixels/area defines the problem.

  • Eric Calabos

    2.7×2.7×14.2=103.5 megapixel.
    In such high density of pixels, lens is the AA filter

    • gsum

      The old Kodak SLRN was 14mp full frame. Nikon’s primes could easily out resolve that sensor (which didn’t have an AA filter) but moire wasn’t a significant problem as it only appears under very specific circumstances i.e. when a repeated pattern is resolved at or near the sensor resolution. Move the camera away from or towards the pattern and moire rapidly disappears.
      Pixel peeping the N1 images, I suspect that the lenses might be able to out resolve the sensor but either way I don’t think that moire will be a problem.

      • Micah Goldstein

        Diffraction limit is probably 2.8 with a 14mp sensor at this size. With kit lenses around 3.5-5.6, I doubt there’s an issue at all.

  • thomasverbeke

    no AA filter; why is that a good thing? I always tough it was there for a reason; don’t see how removing it is a technical innovation unless they have other systems for reducing aliasing.

    • davidpeter

      There is no ideal AA (low pass) filter. If you would like to remove all the moire, you have to sacrifice some sharpness, what would be achievable otherwise. If you take out the AA filter, you have more details, but in some cases, artifacts too. Some cares about aliasing, some not. It’s good to have a manufacturer, who tries to satisfy both camps.

    • MJr

      Someday cars will fly, what o what will we do with the wheels !?

  • Richard S

    The D800E, we are told, does have a filter, coupled to an un-filter, as it were, presumably so that the same geometry as the D800 is maintained. Does the new arrival really have NO AA filter at all?

    • joshimi

      Excellent question: how the flange focal distance is maintain if the sensor have less layer than usual ?

      • Greg H

        you two guys must be really bored-go shoot some pictures and dont lose any more “un-sleep” the sleep you dont get when you stay awake slamming your head against the wall wondering about stuff like this-THERE’S NO AA FILTER IN THE NEW NIKON-why do you need to know all the technicalities? keep it simple you two and go out and smell the roses or even better-photograph a few of them!!!
        geeeeez!!!

        • OhYeat

          I’m starting to wish I could un-read these comments :)

          We’re all so far removed from photography (the taking pictures part) now … I think it’d be a step in the RIGHT direction for some of us to go take pictures of brick walls and look for real moire instead of just talking about it. Once that’s done we get back to the real work of pixel peeping on our 80″ hdtvs. :)

    • Greg H

      WTF?

  • FrenchGuy

    So now mister Nikon I want my D400 without AA filter!

    • KnightPhoto

      Right you are, bring on the D400E. And really, fellow togs, there is no need anymore for a non-E model. Even if Moire can be found in an image, it lives at such tiny places within the overall image that it isn’t consequential. Plus don’t forget to include the dual XQD slots this time! ;-)

  • Gordon

    Nikon has learned much from the market’s response to the D800E.

  • RxGus

    Sounds OK to me… AA is for quitters.

  • nawab

    is that news? It is like-
    hey, you see that person? He is not wearing spectacles.

  • cvbcbcvb

    D800E still have AA filter… please check carefully. it uses two AA filter to to negate the effect.

  • Angler

    Post 1: I have more knowledge and arrogance.
    Post 2: NO, I have more knowledge and arrogance.
    The internet is full of douche bags. Including on this site.

  • Back to top