In praise of AA filters

Started Oct 23, 2013 | Discussions
Steen Bay
Veteran MemberPosts: 6,411
Like?
Re: Detail above Nyquist in MTF
In reply to rovingtim, Oct 27, 2013

rovingtim wrote:

Lee Jay wrote:

Lee Jay wrote:

Using the slanted line method of oversampling. That's s special case and no real detail can be found that way.

That has been my point all along. Software is 'creating' the appearance of detail over Nyquist in digital files.

Wrong.

So, what are those black lines in the image that are beyond Nyquist then?

That's in the resolution test chart shots in the reviews, not in the lens tests. In the lens tests the MTF-50 values are far below the Nyguist frequency.

Reply   Reply with quote   Complain
Lee Jay
Forum ProPosts: 45,245Gear list
Like?
Re: Detail above Nyquist in MTF
In reply to rovingtim, Oct 27, 2013

rovingtim wrote:

Lee Jay wrote:

Lee Jay wrote:

Using the slanted line method of oversampling. That's s special case and no real detail can be found that way.

That has been my point all along. Software is 'creating' the appearance of detail over Nyquist in digital files.

Wrong.

So, what are those black lines in the image that are beyond Nyquist then?

Aliasing.  Their spacial frequency is by-definition below Nyquist.

-- hide signature --

Lee Jay
(see profile for equipment)

 Lee Jay's gear list:Lee Jay's gear list
Canon IXUS 310 HS Canon PowerShot SX260 HS Canon EOS 5D Canon EOS 20D Canon EOS 550D +23 more
Reply   Reply with quote   Complain
crames
Regular MemberPosts: 192
Like?
Re: Wait a minute: Foveon example not so aliased after all...
In reply to John Sheehy, Oct 27, 2013

John Sheehy wrote:

crames wrote:

Take a look at the foveon example zoomed 300% with sinc interpolation. There is actually only a few small aliased patches, mostly in the upper left.

Huh? I see moire in the right edge, second patch from bottom, the right bottom of the first full patch in the lower right corner, the patch just a little left and south of center, and I also see inconsistent rendering of the darker spots in the areas where the three-toned lines cross the other lines, due to luck of alignment. This is a clearly aliased capture. I don't see just fabric here; I see poor capture, which ruins the fabric. Face it, we need cameras with a hell of a lot more pixels to properly record things like this. Shortcuts do not work.

Yes, I said "mostly" in the upper left. It's nice that you were able to find some artifacts in other areas, but without knowing the origin and processing history of this image it's hard to say what is the exact cause of some defects. For sure I would say that the moire that occurs at angles that are inconsistent with the direction of the weave is aliasing.

The point is that most of the "aliasing" goes away with a little interpolation, which means that most of the so-called aliasing is in fact not aliasing.

Lee Jay said that the image is "covered with aliasing." "The entire thing is full of jaggies, false detail and patterns that don't exist."

Here is the original X3 image again, this time zoomed with nearest neighbor.

NN 300%

There you go, jaggies, "snap to grid," moire patterns, etc.

Here it is with Photoshop bicubic smoother @300%:

Bicubic smoother 300%

Would you say that it is still covered with aliasing, or would you agree that most of the artifacts have disappeared? True aliases do not go away so easily.

Foveon zoomed 300% with sinc interpolation

You can't really evaluate aliasing without doing some interpolation - has to do with the sampling theorem. If it goes away when you interpolate, it can't be aliasing.

What people often think is aliasing is really caused by the way the image is displayed, rather than the way it is captured.

Not in a 1:1 display; then there has to be a difference in capture for there to be a difference in output. A converter does not randomly increase micro-contrast in some areas and decrease it in others.

I'm not understanding what you are saying here. Reconstruction artifacts are especially visible in a 1:1 display, since no reconstruction occurs, other than representing pixels as little sharp-edged squares or rectangles, a distortion that occurs regardless of whether the image was captured properly or not.

Reply   Reply with quote   Complain
rovingtim
Veteran MemberPosts: 8,231
Like?
Re: Detail above Nyquist in MTF
In reply to Lee Jay, Oct 27, 2013

Lee Jay wrote:

rovingtim wrote:

Lee Jay wrote:

Lee Jay wrote:

Using the slanted line method of oversampling. That's s special case and no real detail can be found that way.

That has been my point all along. Software is 'creating' the appearance of detail over Nyquist in digital files.

Wrong.

So, what are those black lines in the image that are beyond Nyquist then?

Aliasing. Their spacial frequency is by-definition below Nyquist.

If that's true, then it suggests aliasing is destroying detail that would have otherwise been evident if the image had been guarded against aliasing. In other words, a file from sensor without an AA will have less real detail than a file from a comparable sensor with an AA filter.

Reply   Reply with quote   Complain
Reilly Diefenbach
Senior MemberPosts: 8,389Gear list
Like?
Wrong way, Corrigan.
In reply to Tom Axford, Oct 27, 2013

In another couple of years, no camera worth owning will have an AA filter, so I suggest the aliasing police get used to processing the output correctly, and enjoying the superior microcontrast and detail easily available.

All but one manufacturer, from Sony to Fuji to Leica to Pentax to Phase One (Canon has to be thinking about it) have quite correctly eliminated the blur filter, and good riddance to it.  Sony's best ever camera upcoming won't have it.  It will be a resolution monster just like the "e" as well as a huge seller for Sony. 
It's not an artifact, it's a pixel unblurred, just as if you'd put a loupe up to your monitor.   The more pixels, the higher the res.  Pretty simple stuff.  You want blurred pixels, go for it while you can.

Blurred

Not blurred

EM1 not blurred

Here's another shot you lot can just read and weep.  This is simply off the charts for a sub $1000 camera a couple of years ago and is kicking down the door of the medium format club.  With a bit of Photoshopping, this would be absolutely indistinguishable from any of the really high priced items at any sane magnification.  This is resolving power to the people in a big way.  You can all thank Nikon for leading the way.

Nikon test shot D5300/351.8G

Reply   Reply with quote   Complain
Lee Jay
Forum ProPosts: 45,245Gear list
Like?
Re: Wait a minute: Foveon example not so aliased after all...
In reply to crames, Oct 27, 2013

crames wrote:

John Sheehy wrote:

crames wrote:

Take a look at the foveon example zoomed 300% with sinc interpolation. There is actually only a few small aliased patches, mostly in the upper left.

Huh? I see moire in the right edge, second patch from bottom, the right bottom of the first full patch in the lower right corner, the patch just a little left and south of center, and I also see inconsistent rendering of the darker spots in the areas where the three-toned lines cross the other lines, due to luck of alignment. This is a clearly aliased capture. I don't see just fabric here; I see poor capture, which ruins the fabric. Face it, we need cameras with a hell of a lot more pixels to properly record things like this. Shortcuts do not work.

Yes, I said "mostly" in the upper left. It's nice that you were able to find some artifacts in other areas, but without knowing the origin and processing history of this image it's hard to say what is the exact cause of some defects. For sure I would say that the moire that occurs at angles that are inconsistent with the direction of the weave is aliasing.

The point is that most of the "aliasing" goes away with a little interpolation, which means that most of the so-called aliasing is in fact not aliasing.

Lee Jay said that the image is "covered with aliasing."

Here is the original X3 image again, this time zoomed with nearest neighbor.

NN 300%

There you go, jaggies, "snap to grid," moire patterns, etc.

Here it is with Photoshop bicubic smoother @300%:

Bicubic smoother 300%

Would you say that it is still covered with aliasing, or would you agree that most of the artifacts have disappeared? True aliases do not go away so easily.

Newton's rings (purple and green) center-right.  Z-shaped artifacts lower right corner.  Shallow diagonal lines center lower-right, center upper and upper right.  Green and purple diagonal lines center, left and center left lower.  45° upper-left to lower-right white fragments bottom middle.  Steep diagonal lines lower-right along with more purple-green lines.  Blurry splotch lower-left.  Blurry diagonal lines combined with purple in the center.  45° diagonal lines upper left in two places.

Etc.

-- hide signature --

Lee Jay
(see profile for equipment)

 Lee Jay's gear list:Lee Jay's gear list
Canon IXUS 310 HS Canon PowerShot SX260 HS Canon EOS 5D Canon EOS 20D Canon EOS 550D +23 more
Reply   Reply with quote   Complain
Lee Jay
Forum ProPosts: 45,245Gear list
Like?
Re: Detail above Nyquist in MTF
In reply to rovingtim, Oct 27, 2013

rovingtim wrote:

Lee Jay wrote:

rovingtim wrote:

Lee Jay wrote:

Lee Jay wrote:

Using the slanted line method of oversampling. That's s special case and no real detail can be found that way.

That has been my point all along. Software is 'creating' the appearance of detail over Nyquist in digital files.

Wrong.

So, what are those black lines in the image that are beyond Nyquist then?

Aliasing. Their spacial frequency is by-definition below Nyquist.

If that's true, then it suggests aliasing is destroying detail that would have otherwise been evident if the image had been guarded against aliasing. In other words, a file from sensor without an AA will have less real detail than a file from a comparable sensor with an AA filter.

That depends on the AA filter, but assuming a decent AA filter and sufficient detail in the scene and passed by the optics, that is true because aliasing corrupts real sampled detail.

-- hide signature --

Lee Jay
(see profile for equipment)

 Lee Jay's gear list:Lee Jay's gear list
Canon IXUS 310 HS Canon PowerShot SX260 HS Canon EOS 5D Canon EOS 20D Canon EOS 550D +23 more
Reply   Reply with quote   Complain
Lee Jay
Forum ProPosts: 45,245Gear list
Like?
Re: Wrong way, Corrigan.
In reply to Reilly Diefenbach, Oct 27, 2013

Reilly Diefenbach wrote:

In another couple of years, no camera worth owning will have an AA filter, so I suggest the aliasing police get used to processing the output correctly, and enjoying the superior microcontrast and detail easily available.

You cannot fix the aliasing caused by insufficient sampling combined with no optical low-pass filter.  You either have to have a filter or you have to have sufficient pixel density for diffraction to act as the filter at all f-stops.  That's a lot of pixels (hundreds of megapixels to gigapixels depending on sensor size and f-stop required).

-- hide signature --

Lee Jay
(see profile for equipment)

 Lee Jay's gear list:Lee Jay's gear list
Canon IXUS 310 HS Canon PowerShot SX260 HS Canon EOS 5D Canon EOS 20D Canon EOS 550D +23 more
Reply   Reply with quote   Complain
crames
Regular MemberPosts: 192
Like?
Re: Wait a minute: Foveon example not so aliased after all...
In reply to Lee Jay, Oct 27, 2013

Lee Jay wrote:

crames wrote:

John Sheehy wrote:

crames wrote:

Take a look at the foveon example zoomed 300% with sinc interpolation. There is actually only a few small aliased patches, mostly in the upper left.

Huh? I see moire in the right edge, second patch from bottom, the right bottom of the first full patch in the lower right corner, the patch just a little left and south of center, and I also see inconsistent rendering of the darker spots in the areas where the three-toned lines cross the other lines, due to luck of alignment. This is a clearly aliased capture. I don't see just fabric here; I see poor capture, which ruins the fabric. Face it, we need cameras with a hell of a lot more pixels to properly record things like this. Shortcuts do not work.

Yes, I said "mostly" in the upper left. It's nice that you were able to find some artifacts in other areas, but without knowing the origin and processing history of this image it's hard to say what is the exact cause of some defects. For sure I would say that the moire that occurs at angles that are inconsistent with the direction of the weave is aliasing.

The point is that most of the "aliasing" goes away with a little interpolation, which means that most of the so-called aliasing is in fact not aliasing.

Lee Jay said that the image is "covered with aliasing."

Here is the original X3 image again, this time zoomed with nearest neighbor.

NN 300%

There you go, jaggies, "snap to grid," moire patterns, etc.

Here it is with Photoshop bicubic smoother @300%:

Bicubic smoother 300%

Would you say that it is still covered with aliasing, or would you agree that most of the artifacts have disappeared? True aliases do not go away so easily.

Newton's rings (purple and green) center-right. Z-shaped artifacts lower right corner. Shallow diagonal lines center lower-right, center upper and upper right. Green and purple diagonal lines center, left and center left lower. 45° upper-left to lower-right white fragments bottom middle. Steep diagonal lines lower-right along with more purple-green lines. Blurry splotch lower-left. Blurry diagonal lines combined with purple in the center. 45° diagonal lines upper left in two places.

Etc.

-- hide signature --

Lee Jay
(see profile for equipment)

I see, you want to call any image artifact "aliasing" on this jpeg that's been around the block. It's easier to see other artifacts after all the reconstruction error is removed. What happened to the other 90% of the "entire thing is full of jaggies, false detail and patterns that don't exist"?

Reply   Reply with quote   Complain
Lee Jay
Forum ProPosts: 45,245Gear list
Like?
Re: Wait a minute: Foveon example not so aliased after all...
In reply to crames, Oct 27, 2013

crames wrote:

Lee Jay wrote:

Newton's rings (purple and green) center-right. Z-shaped artifacts lower right corner. Shallow diagonal lines center lower-right, center upper and upper right. Green and purple diagonal lines center, left and center left lower. 45° upper-left to lower-right white fragments bottom middle. Steep diagonal lines lower-right along with more purple-green lines. Blurry splotch lower-left. Blurry diagonal lines combined with purple in the center. 45° diagonal lines upper left in two places.

Etc.

I see, you want to call any image artifact "aliasing" on this jpeg that's been around the block.

Those are the ones that look like aliasing to me.  Of course, there's no way to know as aliasing is impossible to separate from real image detail.  That's the problem with it - it can't be removed once it's been sampled because it's mixed in with the real thing.

It's easier to see other artifacts after all the reconstruction error is removed. What happened to the other 90% of the "entire thing is full of jaggies, false detail and patterns that don't exist"?

I think I just pointed out quite a number of them, and I didn't point out the jaggies on nearly every line.

-- hide signature --

Lee Jay
(see profile for equipment)

 Lee Jay's gear list:Lee Jay's gear list
Canon IXUS 310 HS Canon PowerShot SX260 HS Canon EOS 5D Canon EOS 20D Canon EOS 550D +23 more
Reply   Reply with quote   Complain
The Davinator
Forum ProPosts: 13,355Gear list
Like?
So....
In reply to Tom Axford, Oct 27, 2013

.....the next time I decide to photograph high contrast test charts, I'll keep this in mind.  For all other uses where I want the best resolution, I'll use a camera without the AA.

 The Davinator's gear list:The Davinator's gear list
Canon PowerShot G3 Canon PowerShot SX150 IS Canon EOS D30 Canon EOS 10D Nikon D2X +18 more
Reply   Reply with quote   Complain
Vlad S
Senior MemberPosts: 3,116Gear list
Like?
It's not a "yes" or "no" proposition
In reply to Reilly Diefenbach, Oct 27, 2013

Reilly Diefenbach wrote:

All but one manufacturer, from Sony to Fuji to Leica to Pentax to Phase One (Canon has to be thinking about it) have quite correctly eliminated the blur filter, and good riddance to it. Sony's best ever camera upcoming won't have it. It will be a resolution monster just like the "e" as well as a huge seller for Sony.

It's not an artifact, it's a pixel unblurred, just as if you'd put a loupe up to your monitor. The more pixels, the higher the res. Pretty simple stuff.

The companies are eliminating the AA filters because at the current resolutions the chances of moire patterns are very few and far between. Even hair is often outresolved by the current sensor at the typical head and shoulders scale.

Nevertheless, theoretically the moire is still possible, it's just that in practice the number of subjects that can display it has diminished greatly. In your examples you can see stepping edges in the diagobals of the balance scale and in parts of the brush. Foveon is not immune to it either, especially the earlier, 5MP versions.

So it's not a question of whether the method is good or bad, but whether the benefits outweigh the disadvantages for each given sensor resolution. Until the Bayer sensors hit 16-24 MP the benefits of the AA filters were too significant to eliminate them.

Vlad

 Vlad S's gear list:Vlad S's gear list
Olympus OM-D E-M5 Panasonic Lumix G Vario 14-45mm F3.5-5.6 ASPH OIS Olympus M.Zuiko Digital ED 9-18mm 1:4.0-5.6 Panasonic Lumix G 20mm F1.7 ASPH Panasonic Lumix G Vario 100-300mm F4-5.6 OIS +2 more
Reply   Reply with quote   Complain
Reilly Diefenbach
Senior MemberPosts: 8,389Gear list
Like?
Re: It's not a "yes" or "no" proposition
In reply to Vlad S, Oct 27, 2013

Vlad S wrote:

Reilly Diefenbach wrote:

All but one manufacturer, from Sony to Fuji to Leica to Pentax to Phase One (Canon has to be thinking about it) have quite correctly eliminated the blur filter, and good riddance to it. Sony's best ever camera upcoming won't have it. It will be a resolution monster just like the "e" as well as a huge seller for Sony.

It's not an artifact, it's a pixel unblurred, just as if you'd put a loupe up to your monitor. The more pixels, the higher the res. Pretty simple stuff.

The companies are eliminating the AA filters because at the current resolutions the chances of moire patterns are very few and far between. Even hair is often outresolved by the current sensor at the typical head and shoulders scale.

Nevertheless, theoretically the moire is still possible, it's just that in practice the number of subjects that can display it has diminished greatly. In your examples you can see stepping edges in the diagobals of the balance scale and in parts of the brush.

My contention Vlad is that that is okay, it's the actual pixels, not blurred.  The scale is more legible without a doubt.  If you want finer detail, folks, get an 80MP camera.  Just don't blur my pixels!!!

Foveon is not immune to it either, especially the earlier, 5MP versions.

So it's not a question of whether the method is good or bad, but whether the benefits outweigh the disadvantages for each given sensor resolution. Until the Bayer sensors hit 16-24 MP the benefits of the AA filters were too significant to eliminate them.

Vlad

Reply   Reply with quote   Complain
Lee Jay
Forum ProPosts: 45,245Gear list
Like?
Re: It's not a "yes" or "no" proposition
In reply to Reilly Diefenbach, Oct 27, 2013

Reilly Diefenbach wrote:

My contention Vlad is that that is okay, it's the actual pixels, not blurred. The scale is more legible without a doubt. If you want finer detail, folks, get an 80MP camera. Just don't blur my pixels!!!

You'd rather have sharp, but wrong pixels than very slightly blurry, but accurate pixels?

Removing the AA filter gets you, at best, 30% more equivalent pixel count, but it means you can't trust your images to show what was actually there, and it means you might get some very, very ugly and false artifacts that are obvious at any scale.

-- hide signature --

Lee Jay
(see profile for equipment)

 Lee Jay's gear list:Lee Jay's gear list
Canon IXUS 310 HS Canon PowerShot SX260 HS Canon EOS 5D Canon EOS 20D Canon EOS 550D +23 more
Reply   Reply with quote   Complain
John Sheehy
Forum ProPosts: 16,293
Like?
Re: Wait a minute: Foveon example not so aliased after all...
In reply to crames, Oct 27, 2013

crames wrote:

John Sheehy wrote:

crames wrote:

Take a look at the foveon example zoomed 300% with sinc interpolation. There is actually only a few small aliased patches, mostly in the upper left.

Huh? I see moire in the right edge, second patch from bottom, the right bottom of the first full patch in the lower right corner, the patch just a little left and south of center, and I also see inconsistent rendering of the darker spots in the areas where the three-toned lines cross the other lines, due to luck of alignment. This is a clearly aliased capture. I don't see just fabric here; I see poor capture, which ruins the fabric. Face it, we need cameras with a hell of a lot more pixels to properly record things like this. Shortcuts do not work.

Yes, I said "mostly" in the upper left. It's nice that you were able to find some artifacts in other areas, but without knowing the origin and processing history of this image it's hard to say what is the exact cause of some defects. For sure I would say that the moire that occurs at angles that are inconsistent with the direction of the weave is aliasing.

All moire is aliasing.  All aliasing is not moire, though.  Moire is an extra pattern caused by the juxtaposition of existing patterns.

The point is that most of the "aliasing" goes away with a little interpolation, which means that most of the so-called aliasing is in fact not aliasing.

It doesn't go away.  You lower the contrast of everything

Lee Jay said that the image is "covered with aliasing." "The entire thing is full of jaggies, false detail and patterns that don't exist."

Here is the original X3 image again, this time zoomed with nearest neighbor.

Yes, and the aliasing, including the moire, is much easier too see, as original pixel transients are maintained at the edges of the big tiles.

There you go, jaggies, "snap to grid," moire patterns, etc.

Here it is with Photoshop bicubic smoother @300%:

Would you say that it is still covered with aliasing, or would you agree that most of the artifacts have disappeared? True aliases do not go away so easily.

I would say that the contrast of all small detail, real or artificial, is now much lower, but I can still see patches of higher contrast and lower contrast areas, due to luck of alignment between sensor and subject detail.  I also see softened jaggies, not the lumpless near-analog texture we would have had with 9x as many pixels.

Not in a 1:1 display; then there has to be a difference in capture for there to be a difference in output. A converter does not randomly increase micro-contrast in some areas and decrease it in others.

I'm not understanding what you are saying here. Reconstruction artifacts are especially visible in a 1:1 display, since no reconstruction occurs, other than representing pixels as little sharp-edged squares or rectangles, a distortion that occurs regardless of whether the image was captured properly or not.

I'm saying that if we are looking at a 100% view, then the low-frequency moire artifacts could not have come from processing.  You can only have modulating pixel contrast if the capture is aliased.  Perhaps you have a much narrower definition of aliasing than I do.  Do you believe that there is a range of sampling artifacts occurring between oversapmpling and your definition of aliasing, which are caused by undersampling but not "aliasing"?  IMO, every dot, and every edge in an image should have the same visible blur regardless of how the sensor aligns with the subject; anything less than this is "aliasing".  Proper sampling is oversampling, IMO.

Reply   Reply with quote   Complain
John Sheehy
Forum ProPosts: 16,293
Like?
Re: So....
In reply to The Davinator, Oct 27, 2013

Dave Luttmann wrote:

.....the next time I decide to photograph high contrast test charts, I'll keep this in mind. For all other uses where I want the best resolution, I'll use a camera without the AA.

You can use whatever you want, but if you think that there are going to be no artifacts shooting at f/2.8 with a sharp lens at 1/2000, you are not being realistic.  Some people will see the aliasing on non-repeating patterns, even if you don't.

Reply   Reply with quote   Complain
Tony Beach
Senior MemberPosts: 5,073
Like?
Re: It's not a "yes" or "no" proposition
In reply to Lee Jay, Oct 27, 2013

Lee Jay wrote:

Removing the AA filter gets you, at best, 30% more equivalent pixel count, but it means you can't trust your images to show what was actually there, and it means you might get some very, very ugly and false artifacts that are obvious at any scale.

This was argued extensively when the D800 and D800E came out.  There was lots of angst among early adopters about which way to go, but now that people have actually used these cameras the consensus is overwhelmingly in favor of the D800E with no AA filter because its files are just cleaner and show more resolution at wider apertures.

Nikon has seen this play out and that's why you can't get a D7100 with an AA filter -- period.  Likewise Sony has done the same with their A7r.  As sensor resolution increases this is going to be a completely moot discussion and all cameras will no longer have AA filters.

Reply   Reply with quote   Complain
Tony Beach
Senior MemberPosts: 5,073
Like?
Re: Wrong way, Corrigan.
In reply to Lee Jay, Oct 27, 2013

Lee Jay wrote:

Reilly Diefenbach wrote:

In another couple of years, no camera worth owning will have an AA filter, so I suggest the aliasing police get used to processing the output correctly, and enjoying the superior microcontrast and detail easily available.

You cannot fix the aliasing caused by insufficient sampling combined with no optical low-pass filter. You either have to have a filter or you have to have sufficient pixel density for diffraction to act as the filter at all f-stops. That's a lot of pixels (hundreds of megapixels to gigapixels depending on sensor size and f-stop required).

Reilly is absolutely right here, and the writing is on the wall for all to see (see my reply below posted a couple of minutes ago).  Your argument that we need gigapixels of resolution is ludicrous because we are talking about the real world (rather than a fantasy or theoretical world) where lenses, technique, and software make exceeding Nyquist frequency at far lower resolutions then you propose a non-issue.

Reply   Reply with quote   Complain
Lee Jay
Forum ProPosts: 45,245Gear list
Like?
Re: Wrong way, Corrigan.
In reply to Tony Beach, Oct 27, 2013

Lee Jay wrote:

Reilly Diefenbach wrote:

In another couple of years, no camera worth owning will have an AA filter, so I suggest the aliasing police get used to processing the output correctly, and enjoying the superior microcontrast and detail easily available.

You cannot fix the aliasing caused by insufficient sampling combined with no optical low-pass filter. You either have to have a filter or you have to have sufficient pixel density for diffraction to act as the filter at all f-stops. That's a lot of pixels (hundreds of megapixels to gigapixels depending on sensor size and f-stop required).

Reilly is absolutely right here, and the writing is on the wall for all to see (see my reply below posted a couple of minutes ago).  Your argument that we need gigapixels of resolution is ludicrous because we are talking about the real world (rather than a fantasy or theoretical world) where lenses, technique, and software make exceeding Nyquist frequency at far lower resolutions then you propose a non-issue.

I have shot pixel-sharp shots at the full frame equivalent of 184MP, and that's with slow and very long lenses. With faster and shorter lenses it would be much easier and I could do far better.

Until we're above 200MP on full frame, and possibly more, I want an AA filter.
--
Lee Jay
(see profile for equipment)

 Lee Jay's gear list:Lee Jay's gear list
Canon IXUS 310 HS Canon PowerShot SX260 HS Canon EOS 5D Canon EOS 20D Canon EOS 550D +23 more
Reply   Reply with quote   Complain
Lee Jay
Forum ProPosts: 45,245Gear list
Like?
Re: It's not a "yes" or "no" proposition
In reply to Tony Beach, Oct 27, 2013

Lee Jay wrote:

Removing the AA filter gets you, at best, 30% more equivalent pixel count, but it means you can't trust your images to show what was actually there, and it means you might get some very, very ugly and false artifacts that are obvious at any scale.

This was argued extensively when the D800 and D800E came out.  There was lots of angst among early adopters about which way to go, but now that people have actually used these cameras the consensus is overwhelmingly in favor of the D800E with no AA filter because its files are just cleaner and show more resolution at wider apertures.

Nikon has seen this play out and that's why you can't get a D7100 with an AA filter -- period.  Likewise Sony has done the same with their A7r.  As sensor resolution increases this is going to be a completely moot discussion and all cameras will no longer have AA filters.

I agree but we need another factor of 10 more pixel count first.
--
Lee Jay
(see profile for equipment)

 Lee Jay's gear list:Lee Jay's gear list
Canon IXUS 310 HS Canon PowerShot SX260 HS Canon EOS 5D Canon EOS 20D Canon EOS 550D +23 more
Reply   Reply with quote   Complain
Keyboard shortcuts:
FForum MMy threads