Luminous Landscape Forum
Equipment & Techniques => Cameras, Lenses and Shooting gear => Topic started by: Edalongthepacific on January 01, 2010, 09:02:52 pm
-
I read in Digital PhotoPro about the Leica M9. It seems that this camera uses software rather than a "detail-blurring" filter over the sensor to resolve moire issues. The photos from the M9 I have seen look very good. If this is truly an advancement in technology, why hasn't the camera received more publicity regarding this new (?) technology?
-
Mostly because it is mathematically impossible to eliminate moire via software vs an AA filter. There are various software tricks you can use to reduce the appearance of moire, but you can't eliminate it.
-
Hi,
It seems to me, based on previous discussions on this forums, that Jonathan is absolutely right. There is no way to remove aliasing artifacts in postprocessing. It seems also that these artifacts are seldom visible in real life pictures, except for some very fine detail. The AA-filter does reduce edge contrast but not really resolution. The loss of edge contrast can probably regained with sharpening, AA-filtered images require a lot of sharpening. Large amount and small radius.
Best regards
Erik
Mostly because it is mathematically impossible to eliminate moire via software vs an AA filter. There are various software tricks you can use to reduce the appearance of moire, but you can't eliminate it.
-
I read in Digital PhotoPro about the Leica M9. It seems that this camera uses software rather than a "detail-blurring" filter over the sensor to resolve moire issues. The photos from the M9 I have seen look very good. If this is truly an advancement in technology, why hasn't the camera received more publicity regarding this new (?) technology?
Nothing new about it.
-
That said, some technologies like the one Fuji came up with do manage to limit it greatly. The tradeoff is usually lower absolute resolution(tiny details won't resolve as well, but the overall picture looks cleaner and there's almost no negative effects from the sensor pattern itself, so the resulting prints look nearly as clean.
I just wish they would hurry up and develop these alternative sensors faster. Even jumping to ~12MP would be enough, because that would allow for 35mm equivalency. 6-8 isn't quite cutting it.
-
I don't think for one second there is any software inside the M9 that reduces moire patterns. Lets face it, there is so little software in it that we'd have heard about anything new or advanced before now. What the article may be confusing is the software to reduce vignetting with lenses wider than 24mm.
But what Leica have been brave enough to do is realise that moire only affects a very small percentage of photographs, so they left out the AA filter altogether. It was briefly a hot topic of discussion on Leica forums until it was found to be difficult to find or create an example of it outside of a lab. On a couple of occasions alledged examples of real world moire have been posted that are in fact monitor and image resolution conflicts as would happen with any camera.
Steve
-
I read in Digital PhotoPro about the Leica M9. It seems that this camera uses software rather than a "detail-blurring" filter over the sensor to resolve moire issues. The photos from the M9 I have seen look very good. If this is truly an advancement in technology, why hasn't the camera received more publicity regarding this new (?) technology?
I think that Leica does not use a blur filter by necessity rather than by choice. The typical short back focal distance of 27.80 mm on the Leica cameras posed big problems for Leica designers (see Erwin Puts (http://www.imx.nl/photo/leica/camera/page155/m9part2.html)). This limits the filters that can be placed in front of the sensor. In the original M8 digital camera, there was not even room for an infrared filter, which caused problems in color reproduction. The solution was to add an IR filter on the lens.
Mr Puts (http://www.imx.nl/photo/leica/camera/page165/page165.html) does report on a program which reduces the Moire, but does slightly reduce resolution. The conventional wisdom is that Moire can not be removed in software without degrading the image, but sometimes the conventional wisdom is incorrect. Time will tell.
Moire is usually not apparent in landscapes and most other types of photography in nature, but there will still be aliasing. Some perceive the aliasing as added sharpness but others think that it is an ugly artifact. The controversy on the use of blur filters continues.
-
But what Leica have been brave enough to do is realise that moire only affects a very small percentage of photographs, so they left out the AA filter altogether. It was briefly a hot topic of discussion on Leica forums until it was found to be difficult to find or create an example of it outside of a lab. On a couple of occasions alledged examples of real world moire have been posted that are in fact monitor and image resolution conflicts as would happen with any camera.
What evidence do you have of this conscious design decision? Seems a huge leap of faith on your part.
I'm willing to bet it had a lot more to do with engineering than bravery.
-
Hi,
The program that Mr. Puts mentions uses a "neural network" approach. In the image processed by "SharpRawPro" there is a lot of artifacing but it's all monochrome.
http://www.imx.nl/photo/leica/camera/page1...les/testsrp.jpg (http://www.imx.nl/photo/leica/camera/page165/files/testsrp.jpg)
Regarding the question in general, it seems obvious that photographers object mostly to color moiré and that problem seldom shows up in real life subjects, except probably fashion and architecture.
Best regards
Erik
I think that Leica does not use a blur filter by necessity rather than by choice. The typical short back focal distance of 27.80 mm on the Leica cameras posed big problems for Leica designers (see Erwin Puts (http://www.imx.nl/photo/leica/camera/page155/m9part2.html)). This limits the filters that can be placed in front of the sensor. In the original M8 digital camera, there was not even room for an infrared filter, which caused problems in color reproduction. The solution was to add an IR filter on the lens.
Mr Puts (http://www.imx.nl/photo/leica/camera/page165/page165.html) does report on a program which reduces the Moire, but does slightly reduce resolution. The conventional wisdom is that Moire can not be removed in software without degrading the image, but sometimes the conventional wisdom is incorrect. Time will tell.
Moire is usually not apparent in landscapes and most other types of photography in nature, but there will still be aliasing. Some perceive the aliasing as added sharpness but others think that it is an ugly artifact. The controversy on the use of blur filters continues.
-
Chroma moire is pretty easy enough to remove. Luma moire is practically impossible to remove without simultaneously removing real detail or disturbing the image with artifacts.
-
It's not just about moire. I honestly find it amazing that more photographers don't find color aliasing objectionable. To me nothing ruins the film-like appearance of an image more than color aliasing. And unlike moire, color aliasing does show up in in a variety of real-world shots, especially along highlight edges.
As long as we're using Bayer-filtered sensors, I'd rather have a well-designed AA filter.
-
As long as we're using Bayer-filtered sensors, I'd rather have a well-designed AA filter.
+1.
-
Hi,
My opinion may be that AA-filters are OK and needed. The best solution in my view is probably to increase resolution a bit past the diffraction limit, but that may degrade DR (Dynamic Range). My guess is that this discussion on AA filtering or not is a bit overblown. I don't think that DSLR makers put a quite expensive bit of optical filters in their cameras just to reduce image quality.
On the other hand, some cameras like the M9 and all MFDBs lack AA-filtering and are appreciated for that. My guess is that in both cases it may be a question of better lens quality (in relation to pixel pitch).
In my view the M9 is just a very different camera:
- small
- simple
- using very high quality prime lenses
- using a hopefully exact rangefinder
Lloyd Chambers has some interesting writing on the Leica which I'd suggest everyone prepared to spend like ten grand on a M9 should read.
Best regards
Erik
+1.
-
Mostly because it is mathematically impossible to eliminate moire via software vs an AA filter. There are various software tricks you can use to reduce the appearance of moire, but you can't eliminate it.
Jonathan, it is a commonly held belief that aliasing can never be eliminated under any circumstances. Unfortunately, it is one of those "safe" signal processing approaches taken in may textbooks under "default" conditions. I am just making a general remark here and not implying anything to do with Leica and or any other camera. Under certain conditions it is possible to get rid of aliasing even when it has contaminated the data. Please contact me offline and I shall provide you some information and pointers on this topic.
-
AA filter or not? Where's the 'beating a dead horse' icon when you need one.
Might as well argue about UV filters or (yawn) Nikon vs. Canon.
-
Jonathan, it is a commonly held belief that aliasing can never be eliminated under any circumstances. Unfortunately, it is one of those "safe" signal processing approaches taken in may textbooks under "default" conditions. I am just making a general remark here and not implying anything to do with Leica and or any other camera. Under certain conditions it is possible to get rid of aliasing even when it has contaminated the data. Please contact me offline and I shall provide you some information and pointers on this topic.
I'm aware of several techniques to remove or reduce chroma aliasing, but any technique used to eliminate luminance aliasing will reduce real image detail. In order to filter out something, you have to be able to identify it and reliably distinguish it from from what you want to keep. Chroma aliasing can be identified and distinguished by a repeating RGB color cycle at the sampling frequency, but any criteria used to define luminance aliasing will inevitably erroneously define true signal as aliasing in some circumstances. If you want to discuss this in more detail offline, feel free to send me a PM, but I'm highly skeptical that there is any "magic bullet" that can eliminate aliasing in general in the absence of an AA filter.
-
I'm aware of several techniques to remove or reduce chroma aliasing, but any technique used to eliminate luminance aliasing will reduce real image detail. In order to filter out something, you have to be able to identify it and reliably distinguish it from from what you want to keep. Chroma aliasing can be identified and distinguished by a repeating RGB color cycle at the sampling frequency, but any criteria used to define luminance aliasing will inevitably erroneously define true signal as aliasing in some circumstances. If you want to discuss this in more detail offline, feel free to send me a PM, but I'm highly skeptical that there is any "magic bullet" that can eliminate aliasing in general in the absence of an AA filter.
Jonathan, I mean complete elimination of aliasing under certain conditions. No original signal detail is reduced. However, as I said it only happens if certain sampling conditions are satisfied.
-
What might those conditions be, and how likely are those conditions to be met in real-world digital image capture?
-
What might those conditions be, and how likely are those conditions to be met in real-world digital image capture?
Do I need to post my 100% crops from the DMR again?
As I see it the preference for AA or no-AA is a matter of personal choice, there's no 'right' or 'wrong', there are advantages and disadvantages either way.
-
Do I need to post my 100% crops from the DMR again?
Nope ... but you might want to read the thread again ...
-
What evidence do you have of this conscious design decision? Seems a huge leap of faith on your part.
I'm willing to bet it had a lot more to do with engineering than bravery.
Its a 'brave' decision when they face the potential bollocks written about AA filters in reviews and forums like this. Of course its a freaking engineering decision, crikey what do I need to do to stop the odd ironic word being obssesed over?
Steve
-
What might those conditions be, and how likely are those conditions to be met in real-world digital image capture?
If the sampling rate is greater than max. frequency of the original signal but less than Nyquist rate then aliasing spectrum can be regenerated using an adequate DSB modulator and subtracted from the aliased signal to get rid of aliasing. It is difficult, but perhaps not impossible, to build such systems in the practise.
-
Nope ... but you might want to read the thread again ...
You might want to see the prints made from those files b4 you put too much stock in the comments on that thread.
-
If the sampling rate is greater than max. frequency of the original signal but less than Nyquist rate then aliasing spectrum can be regenerated using an adequate DSB modulator and subtracted from the aliased signal to get rid of aliasing. It is difficult, but perhaps not impossible, to build such systems in the practise.
So if the conditions required to remove the aliasing are not met in any typical digital imaging scenario, why even bring it into the discussion? The aliasing that causes visually unpleasant artifacts is due to input frequencies above the Nyquist rate, so what you're talking about is not only of limited practical value under the best of circumstances, it doesn't make any attempt to solve the real-world aliasing problem--incoming data above the Nyquist rate.
-
I read in Digital PhotoPro about the Leica M9. It seems that this camera uses software rather than a "detail-blurring" filter over the sensor to resolve moire issues. The photos from the M9 I have seen look very good. If this is truly an advancement in technology, why hasn't the camera received more publicity regarding this new (?) technology?
There have been numerous threads on this topic recently, no conclusion.
Some people prefer the apparently sharp look of unsharpened AA filterless sensors, others (like myself) prefer the look of a correctly sharpened image coming from a sensor with an AA filter.
In the end, there is very little difference in print at a given pixel count.
Cheers,
Bernard
-
You might want to see the prints made from those files b4 you put too much stock in the comments on that thread.
This thread ... re-read this thread ...
-
So if the conditions required to remove the aliasing are not met in any typical digital imaging scenario, why even bring it into the discussion? The aliasing that causes visually unpleasant artifacts is due to input frequencies above the Nyquist rate, so what you're talking about is not only of limited practical value under the best of circumstances, it doesn't make any attempt to solve the real-world aliasing problem--incoming data above the Nyquist rate.
Firstly, you are mixing up Nyquist frequency with Nyquist rate. You actually mean Nyquist frequency (f_max) when you mention Nyquist rate (2*f_max) in your comment quote above. Sampling theorem says aliasing will happen for any sampling rate less than Nyquist rate (2*f_max) for a low pass signal. And, I am saying that you can have sampling at f_max < f_sample < 2*f_max = Nyquist, i.e., sampling at less than Nyquist rate and still be able to remove that aliasing resulting from that inadequate sampling. I.e., you can have a lower sampling rate than required by the desired Nyquist rate and still be able to eliminate aliasing.
-
I think you did not get it.
No, you don't get it. The aliasing that causes objectionable visual artifacts is generally caused by input frequencies higher than the Nyquist rate of the sensor. IOW, if the sensor samples the image at 50 lp/mm and the lens has a decent MTF at 60 lp/mm, you get aliasing artifacts at 10 lp/mm. Your theoretical approach is only useful for frequencies between 25 and 50 lp/mm, and by your own admission, has no practical real-world implementation approach yet. And it offers no help, even theoretically, for situations where the sensor receives an input of frequencies higher than the 50 lp/mm sampling rate.
The only thing I don't get is why you're cluttering up this thread with a discussion of theoretical BS of little practical relevance under any circumstances, that offers no benefit whatsoever to the greatest real-world aliasing problem encountered in digital imaging--input frequencies higher than the sampling rate.
-
No, you don't get it. The aliasing that causes objectionable visual artifacts is generally caused by input frequencies higher than the Nyquist rate of the sensor. IOW, if the sensor samples the image at 50 lp/mm and the lens has a decent MTF at 60 lp/mm, you get aliasing artifacts at 10 lp/mm. Your theoretical approach is only useful for frequencies between 25 and 50 lp/mm, and by your own admission, has no practical real-world implementation approach yet. And it offers no help, even theoretically, for situations where the sensor receives an input of frequencies higher than the 50 lp/mm sampling rate.
Firstly for the second time: you are mixing up Nyquist frequency with Nyquist rate. You actually mean Nyquist frequency (f_max) when you mention Nyquist rate (2*f_max) in your comment quote above.
Secondly, here is what I am saying: Suppose in a system the max analog frequency is 10 Hz (Nyquist frequency). You then need to sample at 20 samples/sec (Nyquist rate) at least. However, if you sample between 10 and 20 you get aliasing. However, with certain additional mechanism you may still be able to get rid of that aliasing even if you sampled below 20 but over 10. That is all I am saying.
The only thing I don't get is why you're cluttering up this thread with a discussion of theoretical BS of little practical relevance under any circumstances, that offers no benefit whatsoever to the greatest real-world aliasing problem encountered in digital imaging--input frequencies higher than the sampling rate.
Jonathan, I offered you to take it offline.
-
If the sampling rate is greater than max. frequency of the original signal but less than Nyquist rate then aliasing spectrum can be regenerated using an adequate DSB modulator and subtracted from the aliased signal to get rid of aliasing. It is difficult, but perhaps not impossible, to build such systems in the practise.
Joofa,
the problem is that some lenses have enough resolution to deliver signals to the sensor with maximum frequencies higher than the sampling rate (whatever you call it) of the sensor, and these high frequency signal will thus be completely indistinguishable in the sensor's output from a certain lower frequency signals: this is what is meant by "aliasing", and no post-processing can completely eliminate it. For some painfully obvious examples, look at images from Foveon X3 sensors of resolution test charts, where the number of line pairs changes as the line pairs get closer together, while still showing sharp black and white alternation instead of more gracefully and honestly blurring to gray. Gray is an honest "I don't know" whereas aliasing can answer a confident "black" at a place where the correct answer is "white": sharpness with no connection to reality.
It is another question, more esthetic than technical, how often and how badly this aliasing actually harms the final "print image quality".
-
Firstly for the second time: you are mixing up Nyquist frequency with Nyquist rate. You actually mean Nyquist frequency (f_max) when you mention Nyquist rate (2*f_max) in your comment quote above.
Secondly, here is what I am saying: Suppose in a system the max analog frequency is 10 Hz (Nyquist frequency). You then need to sample at 20 samples/sec (Nyquist rate) at least. However, if you sample between 10 and 20 you get aliasing. However, with certain additional mechanism you may still be able to get rid of that aliasing even if you sampled below 20 but over 10. That is all I am saying.
OK, great. Explain how, if you are sampling the image at 50 lp/mm (Nyquist frequency) or 100 pixels/mm (Nyquist rate), you can sample a 75 lp/mm input signal (which is going to alias down to 25 lp/mm) and reliably distinguish the 25 lp/mm aliased signal from a non-aliased signal that was 25 lp/mm before sampling? How you can transform an aliased 25 lp/mm signal back to 75 lp/mm without taking any un-aliased 25 lp/mm signal along for the ride? How does a DSB (dual-sideband modulator?) help solve this problem when the only data you have to work with is the sampled signal?
If what you are saying is valid, I should be able to take an un-aliased image 1000 pixels square, downsample it to 750 pixels using nearest-neighbor resampling (so that no anti-aliasing filtration is done), and then run some kind of software transform on the image to expand it back to 1000 pixels with little or no distortion or loss of detail.
-
Here's a test image, downsampled from 512 to 362 pixels square using nearest-neighbor interpolation, so that it contains aliasing only between the Nyquist frequency and Nyquist rate.
[attachment=19160:Noise_A.bmp]
If your assertions are valid, you should be able to upsample this image back to 512x512 pixels, and your upsampled image and the original image should be essentially identical.
-
It is another question, more esthetic than technical, how often and how badly this aliasing actually harms the final "print image quality".
Exactly. For those who care to go beyond theory to prints, go to Appel Gallery in Sacramento and ask to see my prints. Take a loupe for all I care. Examine the prints at your leisure and see how "bad" it gets. All of my prints in the gallery were made with the Leica DMR (no AA filter), and most were made with a lens that resolves far more than the sensor's sampling rate (Leica 280mm f/4 APO). The aliasing in these prints is so "bad" that skeptical gallery owners can't resist them.
-
Exactly. For those who care to go beyond theory to prints, go to Appel Gallery in Sacramento and ask to see my prints. Take a loupe for all I care. Examine the prints at your leisure and see how "bad" it gets.
Whether aliasing is visually objectionable depends heavily on the subject matter. For most natural subjects (like the stuff you shoot, judging by your web site), a modest amount of aliasing can be visually indistinguishable from enhanced sharpness and detail. But for some subjects, particularly cloth and other man-made objects that have regular, repeating patterns, aliasing an be a serious problem.
[attachment=19161:2004_07_02_0017_a.jpg]
This image is an excellent example. It has been resized to 25% of its original dimensions using nearest-neighbor resampling. Even though the "detail" in the grass is really mostly aliasing, it looks OK because we see roughly what we expect to see there. The aliased false detail is similar to what true detail might look like, so the aliasing is not an issue in that area.
But the barn roof is a completely different story. It is made of corrugated sheet metal that has rusted in stripes along the corrugations. The corrugations are vertically oriented, going perpendicularly from the eaves to the peak of the roof, as shown in this crop:
[attachment=19162:2004_07_...0017crop.jpg]
Obviously, there is a major difference between the aliased image and the original subject in this area. The bottom line is that even if aliasing is advantageous for what you typically shoot, that does not mean that it is a good thing for everyone, particularly portrait/fashion shooters where aliased cloth textures can be a huge problem.
-
Jonathan,
This is a very good example of the problem.
So you essentially "emulate undersampling" by doing nearest neighbor interpolation to 25% scale. This undersampling results in a visible pattern in the 45 degree direction although the real pattern is vertical. The pattern is visible on the roof. Did I get it right?!
Best regards
Erik
Whether aliasing is visually objectionable depends heavily on the subject matter. For most natural subjects (like the stuff you shoot, judging by your web site), a modest amount of aliasing can be visually indistinguishable from enhanced sharpness and detail. But for some subjects, particularly cloth and other man-made objects that have regular, repeating patterns, aliasing an be a serious problem.
[attachment=19161:2004_07_02_0017_a.jpg]
This image is an excellent example. It has been resized to 25% of its original dimensions using nearest-neighbor resampling. Even though the "detail" in the grass is really mostly aliasing, it looks OK because we see roughly what we expect to see there. The aliased false detail is similar to what true detail might look like, so the aliasing is not an issue in that area.
But the barn roof is a completely different story. It is made of corrugated sheet metal that has rusted in stripes along the corrugations. The corrugations are vertically oriented, going perpendicularly from the eaves to the peak of the roof, as shown in this crop:
[attachment=19162:2004_07_...0017crop.jpg]
Obviously, there is a major difference between the aliased image and the original subject in this area. The bottom line is that even if aliasing is advantageous for what you typically shoot, that does not mean that it is a good thing for everyone, particularly portrait/fashion shooters where aliased cloth textures can be a huge problem.
-
So you essentially "emulate undersampling" by doing nearest neighbor interpolation to 25% scale. This undersampling results in a visible pattern in the 45 degree direction although the real pattern is vertical. the pattern is visible on the roof,. Did I get it right?!
That is exactly the point of the exercise. Downsizing with nearest-neighbor causes the same kind of aliasing as one sees when shooting without an AA filter. With the right subject matter, a bit of aliasing can increase perceived detail/sharpness. But with the wrong subject matter, it's just ugly.
-
Jonathan,
What's your view on oversampling, like ignoring diffraction and sampling with say 3 micron pitch? The image would obviously have a lot of non-resolved detail if not a truly excellent lens would be used at a very large aperture. We would obviously down sample using something better than nearest neighbor.
Best regards
Erik
That is exactly the point of the exercise. Downsizing with nearest-neighbor causes the same kind of aliasing as one sees when shooting without an AA filter. With the right subject matter, a bit of aliasing can increase perceived detail/sharpness. But with the wrong subject matter, it's just ugly.
-
Jonathan,
What's your view on oversampling, like ignoring diffraction and sampling with say 3 micron pitch. The image would obviously have a lot of non-resolved detail if not a truly excellent lens would be used at a very large aperture.
IMO, the loss of DR from halving pixel size/pitch is offset by the downsampling, so that's a wash. You could do noise reduction and demosaic prior to downsampling, so that each output pixel would be true RGB (having real sampled data in each color channel), and demosaic and noise reduction artifacts would be minimized by the resampling. I think a credible case could be made for doing most of this in-camera (except for noise reduction) and outputting a downsampled RGB RAW.
-
With the right subject matter, a bit of aliasing can increase perceived detail/sharpness. But with the wrong subject matter, it's just ugly.
That's something everyone can agree with. I would add that some people find aliasing to be ugly even the "right" subject matter.
Some prefer aliased images, describing them as sharp, crunchy, high microcontrast, with lots of fine detail. Others would see it as fake-looking, with harsh transitions, jagged edges, and lots of false detail.
Others would prefer anti-aliased images, describing them as smooth, natural, with the appropriate amount of detail for their size. Some would see them as mushy, hazy, low contrast, and lacking in fine detail.
-
That's something everyone can agree with. I would add that some people find aliasing to be ugly even the "right" subject matter.
Some prefer aliased images, describing them as sharp, crunchy, high microcontrast, with lots of fine detail. Others would see it as fake-looking, with harsh transitions, jagged edges, and lots of false detail.
Others would prefer anti-aliased images, describing them as smooth, natural, with the appropriate amount of detail for their size. Some would see them as mushy, hazy, low contrast, and lacking in fine detail.
Yes indeed. I would add one thing to this, images from AA filterless cameras typically have undergone some form of in camera processing to reduce artifcats (the M9 and S2 come to mind), that are also easy to recognize and can be perceived as being pleasing or not. I would describe this look as micro-level painterly.
Cheers,
Bernard
-
Whether aliasing is visually objectionable depends heavily on the subject matter. For most natural subjects (like the stuff you shoot, judging by your web site), a modest amount of aliasing can be visually indistinguishable from enhanced sharpness and detail. But for some subjects, particularly cloth and other man-made objects that have regular, repeating patterns, aliasing an be a serious problem.
Fine feather detail has many regular repeating patterns. The most difficult bird to photograph for me have been the California Quail and Gambel's Quail; their back and chest feathers cause the wildest aliasing and color moire I've yet seen. However the AA filter doesn't solve the problem. These birds' feathers also cause color moire and aliasing when using a camera with an AA filter.
-
Hi,
If you look at test images with USAF type test targets it's quit obvious that there is aliasing even in AA-filtered images. The amount can vary with converter. Like most things in life the AA-filter is a compromise.
Erwin Puts compares some raw converters here:
http://www.imx.nl/photo/leica/camera/page165/page165.html (http://www.imx.nl/photo/leica/camera/page165/page165.html)
Erwin Puts also have some samples showing color Moiré
here: http://www.imx.nl/photo/leica/camera/page1...s/m9100sela.jpg (http://www.imx.nl/photo/leica/camera/page157/files/m9100sela.jpg) (Leica M9)
and here: http://www.imx.nl/photo/leica/camera/page1...les/d3xsela.jpg (http://www.imx.nl/photo/leica/camera/page157/files/d3xsela.jpg) (Nikon D3x, still has some)
With the full article here: http://www.imx.nl/photo/leica/camera/page157/page157.html (http://www.imx.nl/photo/leica/camera/page157/page157.html)
Interestingly he has not seen any Moiré on the Leica M8 in the same test. Basically the M8 and the M9 share sensor, so it's a bit odd that the Moiré is not showing up. I'm often not certain what Erwin Puts writes about his experiments, like distance and cropping.
Best regards
Erik
Fine feather detail has many regular repeating patterns. The most difficult bird to photograph for me have been the California Quail and Gambel's Quail; their back and chest feathers cause the wildest aliasing and color moire I've yet seen. However the AA filter doesn't solve the problem. These birds' feathers also cause color moire and aliasing when using a camera with an AA filter.
-
Sorry Bernard, which images?
I presume that M9 and S2 users are expected to shoot "raw", if you say that "raw" images are manipulated than the value of having "raw" would be much reduced. Post processing is another issue. In the Video on S2 they photographers discuss that the S2 images are less tolerant of sharpening than the corresponding Phase One back.
One issue here is microlenses. Microlenses increase the fill factor and the fill factor is also effecting aliasing. Making the "sensel" larger increases the probabilty that a ray of light is affecting more than one pixel.
Best regards
Erik
Yes indeed. I would add one thing to this, images from AA filterless cameras typically have undergone some form of in camera processing to reduce artifcats (the M9 and S2 come to mind), that are also easy to recognize and can be perceived as being pleasing or not. I would describe this look as micro-level painterly.
Cheers,
Bernard
-
... look at images from Foveon X3 sensors of resolution test charts, where the number of line pairs changes as the line pairs get closer together, while still showing sharp black and white alternation instead of more gracefully and honestly blurring to gray. Gray is an honest "I don't know" whereas aliasing can answer a confident "black" at a place where the correct answer is "white": sharpness with no connection to reality.
It is another question, more esthetic than technical, how often and how badly this aliasing actually harms the final "print image quality".
That's the interesting thing, though. Our eyes hate random noise and digital artifacts more than they dislike slightly blurry and off-color images. This is also why tube amps sound better than transistors to most people - the nature of the distortion, when it does happen, is more pleasing to our senses. When it doesn't, it's a moot point in either case(and it won't 95%+ of the time). But when it does, well, things get hairy fast with digital, be it audio or optical.
I find that non-Bayer type sensors produce a better image because they cause errors that our brains process as "correct" looking. In that resolution test, for instance, our brains expect there to be detail and a line still present and not random gray. I find this to be most obvious in grass and other repeating semi-random patterns found in nature.
That said, the Fuji method as well as their diagonal alignment creates far less issues than either the Foveon or Bayer designs. I just with it had a few more pixels in HDR/high quality mode. Because what it does with the 6MP it has is amazing.
-
OK, great. Explain how, if you are sampling the image at 50 lp/mm (Nyquist frequency) or 100 pixels/mm (Nyquist rate), you can sample a 75 lp/mm input signal (which is going to alias down to 25 lp/mm) and reliably distinguish the 25 lp/mm aliased signal from a non-aliased signal that was 25 lp/mm before sampling? How you can transform an aliased 25 lp/mm signal back to 75 lp/mm without taking any un-aliased 25 lp/mm signal along for the ride? How does a DSB (dual-sideband modulator?) help solve this problem when the only data you have to work with is the sampled signal?
A sampled data system has periodicities that can be taken advantage of here. I already said that aliasing spectrum can be canceled using these periodicities.
If what you are saying is valid, I should be able to take an un-aliased image 1000 pixels square, downsample it to 750 pixels using nearest-neighbor resampling (so that no anti-aliasing filtration is done), and then run some kind of software transform on the image to expand it back to 1000 pixels with little or no distortion or loss of detail.
Software transform? When did I say that? You need a little background in communication systems to understand some of this stuff.
-
A sampled data system has periodicities that can be taken advantage of here. I already said that aliasing spectrum can be canceled using these periodicities.
Yeah, you keep SAYING that, but you have yet to offer one shred of evidence that it can be done in real-world conditions that have any relevance whatsoever to digital imaging.
Software transform? When did I say that? You need a little background in communication systems to understand some of this stuff.
You're right, you didn't mention software, I did. I have a lot deeper background in the subject than you might think; I used to be very active in CB radio and understand the difference between frequency modulation, amplitude modulation, single sideband modulation, etc. and I've designed and built a variety of electronic devices. I'm also a fairly decent musician; I've been playing various instruments for about 30 years, and have been recording and mixing sound for about 15. I'm familiar with modulation effects, and know that frequency modulation, amplitude modulation (including single- and double-sideband variants) and all kinds of similar things can be implemented in software as well as hardware. You can get all kinds of stuff like that as plug-in modules for audio editing software like ProTools, Adobe Audition, etc.
IF your theory is applicable to digital imaging, than it can be implemented in software, and that software can be used to process images from digital cameras without AA filters (or at least having weaker-than-normal filters) to remove aliasing artifacts without disturbing image detail. To quote your esteemed scientific colleague Warren Mars, "put up or shut up"--either explain HOW aliasing might be removed from a digital image under the conditions you described earlier according to your theory, or STFU and quit wasting LL's bandwidth. Don't be afraid to use big words, I can handle it.
I've already posted an image that contains the kind of aliasing you described; the aliasing it contains exactly splits the difference (logarithmically) between the Nyquist rate and Nyquist frequency. If you can reconstitute it back to its 512x512-pixel size in a manner that closely matches the original image, then congratulations; you have something that can be used to approximately double the linear resolution of any digital sensor. If not, then I'm calling BS on yet another internet crackpot.
-
To quote your esteemed scientific colleague Warren Mars, ... or STFU and quit wasting LL's bandwidth. ... I'm calling BS on yet another internet crackpot.
I'm sorry Jonathan the communication is over. You have esteemed qualifications as you report but I have to spend hours sometimes researching some topic before I summarize it in a few lines for internet forums and I just don't feel compelled to do that right now.
-
I'm sorry Jonathan the communication is over. You have esteemed qualifications as you report but I have to spend hours sometimes researching some topic before I summarize it in a few lines for internet forums and I just don't feel compelled to do that right now.
Why don't you simply post a link to a scientific paper describing this de-aliasing technique if this has been published somewhere? I don't need a layman's summary; for most stuff I can look at the nitty-gritty technical details and figure it out with a bit of study. Surely that wouldn't require hours of effort on your part...
-
[crickets]...
-
I just bracket and blend with the Zero Noise program that's linked in the discussion in this forum. It solves most of the ills in one step.
-
Given that Joofa has posted at length on the subject of aliasing removal on several fora without ever posting anything of substance to support those claims, and hasn't bothered to respond since I challenged him to prove his claims (even though he's posted several times since then), I'm going to go ahead and conclude he can't back up his claims with any substance.
-
... I'm going to go ahead and conclude he can't back up his claims with any substance.
Your assumption is going to make an ass out of u (but not me). Have you ever actually seen a print made from an M8, M9, or DMR? Look at my web images for the last 3.5 years and tell me which one would have been better if I had used a camera with an AA filter. I know of one photo on my website in that time period that is visibly degraded by aliasing. The fine feather detail of that same bird also causes aliasing when photographed with a camera that has an AA filter.
Of course aliasing is false data. So is the AA filter's blur, and the USM required to restore microcontrast to an AA-filtered photos. It's all an illusion - the question is, if you have to be an anal-retentive pixel-peeper to see the difference, does the question deserve the bandwidth that has been devoted to it here? Look at my prints and tell me the lack of an AA filter is a problem. You know where to find them.
-
Your assumption is going to make an ass out of u (but not me).
The issue of whether aliasing is visually tolerable in an image is entirely separate from the question of whether it can be removed without destroying real image data. I've already posted examples of both cases--where aliasing looks OK in an an image, and where it looks really bad. I have no argument with your contention that you can make visually attractive images with captures from a camera with no AA filter.
But as I said before, that is not a universal truth. In some circumstances, aliasing can cause visually jarring artifacts that are not acceptable, and notwithstanding Joofa's assertions to the contrary, there is no way to remove them without taking out real image detail. There some situations where an AA filter is not necessary, and other situations where it is.