back alley
IMAGES
isn't the op just asking for an image that has moire in it?
Do we really need the AA filter?
- Thread starter giellaleafapmu
- Start date
- Latest activity Latest activity:
- Replies 35
- Views 3K
semilog
curmudgeonly optimist
willie_901
Veteran
Yes, and the thread's title asks another question. Both questions are answered.
tstermitz
Well-known
Yes, Moire artifacts do occur in some photos. Regularly aligned, Aspen tree trunks provoke rainbow bands of moire on an M9 sensor when using the Elmarit-M 24mm lens. In the future, I'll try to remind myself to use a smaller aperture so that diffraction can mitigate this.
This occurs on the more distant trunks, but the consequence is severe enough that you don't need to zoom in very far to see it. The moire tool in Lightroom helps reduce the effect away from 1:1 pixel peeping.
This occurs on the more distant trunks, but the consequence is severe enough that you don't need to zoom in very far to see it. The moire tool in Lightroom helps reduce the effect away from 1:1 pixel peeping.
semilog
curmudgeonly optimist
Or you could get a properly engineered camera.
😛😛😛
Seriously, though, that's a good solution.
giellaleafapmu
Well-known
Ok, let's see if I can make myself clear. First I didn't like the kind of answers in which things are added to let the question seem silly such as: "Physics is not propaganda". I never said anything vaguely similar to this and I don't see why you are adding the line in your answer. Second, having explained a (already well known) physical reason does not say anything about wheter you see much Moiré in the new cameras without AA filters. I don't have full access to the engeneering of cameras and as far as I know they could suppress Moiré at a later stage with an algoritm instead of a filter for example. Fact is that Moiré was pretty evident in the Kodak cameras, I didn't see that much with the D800e and telling, as someone did, that if you are precise focussing and use a sturdy tripod you should definitively see it is a bit strange, hence my sarcasm about not being able to use a tripod and focus on a light-table, I mean, I might be an incredibly bad photographer but I can safely claim that I can put my camera in a tripod, focus and shot... Why did I ask? Because so far I like the D800e look better than the D800 look but never worried too much about details because I don't own the cameras, I rent them and
if ever Moiré was to appear in a picture I would suppress some way or assume the lost and rent one more day a D800 with AA filter, however now cameras withourt AA filter are appearing more and more and if I ever consider buying a new camera (which will eventually happen, even if one could make a living out of a D5000, another funny comment) I might consider which is better and consider it on facts, not on some braguing about Canon and Nikon. That was all. As said many had answered quite satisfactorly (thanks), then the interenting posts on megapixel wars and other completely unrelated topics appeared...including the "Physics is not propaganda". Really not called for... As a side note I could not see anyone in the thread producing a "normal" picture (not a picture of a net in front of a window or some strange pattern made to produce it) taken with a D800e or a Pentax K-5iie which present Moiré artifacts...
GLF
willie_901
Veteran
I stand by every word I wrote.
semilog
curmudgeonly optimist
Heck, I stand by every word you wrote, too.
GaryLH
Veteran
A lot depends on what was in the picture, if u will notice any moire effect or how close to the main subject the issue that could cause this.
I have the following cameras that I know do not have AA filters
- Ricoh gxr a12m module -- I have seen moire in certain situations (around 4000 shots..haven't used it much lately)
- Fuji xp1 & xe1 -- there is chance for it, but so far have not seen it (over 8000 shots)
- dp 1 & 2 Merrill -- theoretically not possible, I have not seen it after shooting over 5000 shots between the two including in situations where I have seen it on the Ricoh.
The Fuji and Merrill's get most of my use lately.
As other have said moire removal tools or shooting at f11 or better..all help.
Gary
I have the following cameras that I know do not have AA filters
- Ricoh gxr a12m module -- I have seen moire in certain situations (around 4000 shots..haven't used it much lately)
- Fuji xp1 & xe1 -- there is chance for it, but so far have not seen it (over 8000 shots)
- dp 1 & 2 Merrill -- theoretically not possible, I have not seen it after shooting over 5000 shots between the two including in situations where I have seen it on the Ricoh.
The Fuji and Merrill's get most of my use lately.
As other have said moire removal tools or shooting at f11 or better..all help.
Gary
semilog
curmudgeonly optimist
…AND photograph a subject having spatial frequencies close to a multiple of the pixel pitch, AND using an aperture wide enough that diffraction does not act as a de facto AA filter AND using a lens with well-controlled enough aberrations that the aberrations do not act as a de facto AA filter…
Yes, that is a lot of things to have to do for an image to be degraded by moiré and other aliasing effects.
It happens to be the same list of things you have to do to see any difference at all between a camera with a modern AA filter, and a camera without one.
Hence my comment that this is, for most applications, a wildly overblown issue.
semilog
curmudgeonly optimist
Color moiré, not possible. Other aliasing artifacts, definitely possible. Indeed, if the lens and acquisition conditions are right, aliasing of a spatially or temporally discretized signal is inescapable.
For example, aliasing occurs with the monochrome sensors used for high-end microscopy and astronomy. The usual solution is to oversample -- either by using diffraction as an AA filter, or by defocusing slightly.
Aliasing also occurs with one-dimensional signals, such as audio signals. Wikipedia has a good description:
http://en.wikipedia.org/wiki/Aliasing
Be sure you listen to the audio aliasing sawtooth demo at the bottom of the page. It provides a vivid and nasty-sounding example of why we prefer that our signals should be bandwith-limited (= low-pass filtered = anti-aliased) before they are discretized.
GaryLH
Veteran
Hmmm did not know this...Semilog.
However, in a normal everyday photography using a foveon sensor I have never seen it, what would be something I should try to c if I can recreate it.
Thanks
Gary
However, in a normal everyday photography using a foveon sensor I have never seen it, what would be something I should try to c if I can recreate it.
Thanks
Gary
semilog
curmudgeonly optimist
Nice. Having thoroughly beaten the horse's corpse we're wandering a bit off-topic and I cannot resist sharing this throughly mesmerizing and slightly disquieting video: 32 out of sync metronomes end up synchronizing.
GaryLH
Veteran
Opps.. After re-reading let me rephrase differently. Given color cannot happen, is it possible in the converted image (to monochrom)? If so then what should I do to try to recreate it?
Thanks
Gary
semilog
curmudgeonly optimist
Aliasing artifacts usually appear near harmonics of the sampling frequency.
As typical way to see the artifacts is to look at a descending frequency series (picket fences, screen doors, fabrics viewed from a 45° angle, for example).
In photography a lot of the (obvious) artifacts that might otherwise appear with these sorts of subjects are hidden by shallow depth of field, so the frequncy-matchup is effectively anti-aliased by being out of focus.
A careful bracketed focus series of a relatively distant brick wall at a 45° angle might get you there. Don't use too small an aperture (I'd start at f/4) or diffraction will prevent aliasing.
Can't guarantee that this will work (might be an effort for little gain) but it's what I would try.
Worth noting that the tighter the pixel pitch, the harder it is to see aliasing and the less necessary the AA Filter is. Cell phone camera sensors are carefully matched to lenses that are just barely diffraction-limiting for the sensor. For example the current iPhone (1.75 µm pixel pitch) has an f/2.4 lens and is (theoretically) diffraction-limited for that sensor at f/2.8. Aberrations in the lens and focus errors probably produce just enough defocus to remove the need for a separate AA filter.
As typical way to see the artifacts is to look at a descending frequency series (picket fences, screen doors, fabrics viewed from a 45° angle, for example).
In photography a lot of the (obvious) artifacts that might otherwise appear with these sorts of subjects are hidden by shallow depth of field, so the frequncy-matchup is effectively anti-aliased by being out of focus.
A careful bracketed focus series of a relatively distant brick wall at a 45° angle might get you there. Don't use too small an aperture (I'd start at f/4) or diffraction will prevent aliasing.
Can't guarantee that this will work (might be an effort for little gain) but it's what I would try.
Worth noting that the tighter the pixel pitch, the harder it is to see aliasing and the less necessary the AA Filter is. Cell phone camera sensors are carefully matched to lenses that are just barely diffraction-limiting for the sensor. For example the current iPhone (1.75 µm pixel pitch) has an f/2.4 lens and is (theoretically) diffraction-limited for that sensor at f/2.8. Aberrations in the lens and focus errors probably produce just enough defocus to remove the need for a separate AA filter.
GaryLH
Veteran
Thanks
Will give it a try..
Gary
Will give it a try..
Gary
