My eyes are glazing over looking at uscmatt99's comparison shots because the only thing that interests me is real world photography, and I just have not found this to be a problem with my GXR-M shooting, as I mentioned previously. Sorry, but I don't find this type of comparison of interest.
@uscmatt99: This might be a silly question, but did you actually check the colours of the window pane separators? I admit I'm on holidays at the moment, and where I am staying the window parts are painted different colours!
On the whole, though, if these are artifacts, I'm more in the Mitch camp here: in all the interiors I have shot with the GXR, I have not seen any intrusive problems (apart from the insect screens) and that includes a lot of high-frequency fabrics (furnishings).
Would the 16MP sensor do much better with this kind of image, I wonder?
edwardkaraa wrote:
Very unlikely. This is not a high frequency pattern. Don't know exactly what causes it.
Aliasing doesn't need high frequencies, just sharp high contrast edges. You can get color artifacts when there's only one edge that's projected onto only certain subpixels. When it's a repeating pattern, the artifacts will interfere and create moiré.
Mitch Alland wrote:
Here's a sensible discussion of AA filters by Sean Reid, called The Naked Sensor.
Sounds a rehash of what almost everyone is saying, without really understanding/explaining the issues and making it sound like moiré only is an issue for people photographing fabrics.
"But by stripping away the AA filter (or, in Nikon's case, using a system that is supposed to do the equivalent) they're getting closer to being able to render a lens' drawing without degrading it."
I found Sean's article lacking in information to the point of leaving the reader with the mistaken conclusion that moire can be completely negated in post processing. That's not true. Even in his supposedly corrected example, the maze pattern is visible as are spotty luminosity differences in the fabric. That example is not even close to as bad as things might get.
Tariq Gibran wrote:
I found Sean's article lacking in information to the point of leaving the reader with the mistaken conclusion that moire can be completely negated in post processing...
Tariq what you assert that Sean wrote simply is not correct, as he states, "That makes Nikon's decision to offer the D800/D800E especially interesting because in essence the company is saying to the photographer: "You can have this camera either way. You decide." I, personally, lean towards cameras without AA filters but I hold that preference with the full knowledge that such cameras can sometimes trigger artifacts that require time to correct in post (or which, sometimes, cannot be corrected)."
AhamB wrote:
Sounds a rehash of what almost everyone is saying, without really understanding/explaining the issues and making it sound like moiré only is an issue for people photographing fabrics.
"But by stripping away the AA filter (or, in Nikon's case, using a system that is supposed to do the equivalent) they're getting closer to being able to render a lens' drawing without degrading it."
A dubious claim...
I don't find it being a dubious claim at all, based on my experience with the GXR-M, which I finds is very much transparent to lenses, letting the properties of lenses that I know from use with film simply come through — more so, I find, than the M8 and M9.
Mitch Alland wrote:
Tariq what you assert that Sean wrote simply is not correct, as he states, "That makes Nikon's decision to offer the D800/D800E especially interesting because in essence the company is saying to the photographer: "You can have this camera either way. You decide." I, personally, lean towards cameras without AA filters but I hold that preference with the full knowledge that such cameras can sometimes trigger artifacts that require time to correct in post (or which, sometimes, cannot be corrected)."
Hmm, I don't know if he edited/ added that "(or which, sometimes, cannot be correct)" or I just missed it when the article was originally posted yesterday. In any case, he only addresses partial correction of moire with his example and not pattern/ maze issues which he really should have gone into for the article to be completely objective about the subject. Reading the article gives one the impression that a little spot correction is all that is required to deal with the issue and that is not always the case. He should have shown examples where moire could not be corrected with post processing, such as the Nikon example of the Kimono maze pattern. As it is written, I thought I was reading a one sided argument.
Mitch Alland wrote:
I don't find it being a dubious claim at all, based on my experience with the GXR-M, which I finds is very much transparent to lenses, letting the properties of lenses that I know from use with film simply come through — more so, I find, than the M8 and M9.
—MItch/Bangkok
Mitch, this idea that a bayer filtered camera without an AA filter is somehow allowing "greater fidelity to the image cast by the lens", as Sean says is dubious since the lens does not project the maze patterns or the spotty luminosity found in moire. The reverse argument could be just as well made that the AA filter renders the scene more faithful as captured by the lens because, in most cases, it eliminates these artifacts which were not projected by the lens in the first place. Anyway, unless Fuji has somehow cracked the moire code with their new filter array, one must choose their poison with B&W sensitive sensors - sharper images with possible artifacts that cannot always be corrected after the fact OR softer images that require sharpening.
Tariq Gibran wrote:
Mitch, this idea that a bayer filtered camera without an AA filter is somehow allowing "greater fidelity to the image cast by the lens", as Sean says is dubious since the lens does not project the maze patterns or the spotty luminosity found in moire. The reverse argument could be just as well made that the AA filter renders the scene more faithful as captured by the lens because, in most cases, it eliminates these artifacts which were not projected by the lens in the first place. Anyway, unless Fuji has somehow cracked the moire code with their new filter array, one must choose their poison with B&W sensitive sensors - sharper images with possible artifacts that cannot always be corrected after the fact OR softer images that require sharpening....Show more →
Tariq, this is exactly what I find so unbalanced about the whole discussion in this thread as to be removed from reality, particularly as far as the GXR-M is concerned — more so than for the M8/M9. When I started using the GXR-M, with the very first few pictures I took with the old DR-Summicron I was blown away at how much this camera transmits the properties of the lens, I could recognize characteristics in the image that I was familiar with from having shot with this lens on film. That is the big thing about the GXR-M. The fact that sometimes the GXR-M can produce moiré is a very minor thing, as it happens so little — at least in my experience and apparently also in that of Henrik (kosmoskatten), The statements here on the artifacts of the GXR-M are simply unbalanced, implicitly setting the 500 pound gorilla in the room (resolution, clarity and color accuracy) against the mouse (moiré) — there is a fairy tale about a mouse winning, but that's against an elephant, not a gorilla.
Mitch Alland wrote:
Tariq, this is exactly what I find so unbalanced about the whole discussion in this thread as to be removed from reality, particularly as far as the GXR-M is concerned — more so than for the M8/M9. When I started using the GXR-M, with the very first few pictures I took with the old DR-Summicron I was blown away at how much this camera transmits the properties of the lens, I could recognize characteristics in the image that I was familiar with from having shot with this lens on film. That is the big thing about the GXR-M. The fact that sometimes the GXR-M can produce moiré is a very minor thing, as it happens so little — at least in my experience and apparently also in that of Henrik (kosmoskatten), The statements here on the artifacts of the GXR-M are simply unbalance, implicitly setting the 500 pound gorilla in the room (resolution, clarity and color accuracy) against the mouse (moiré) — there is a fairy tale about a mouse winning, but that's against an elephant, not a gorilla.
I'm not specifically aiming my points about moire at the GXR-M even though it does fall under the AA-less bayer filtered sensors (though I was surprised by the moire shown in the house example). That example goes to show that you really never know when moire is going to show up. I would have expected it on something like a fine pitched screen or a lot of studio subjects (fabrics, hair) but seeing it in the house example actually did surprise me. I think for the kind of everyday subject matter that most of you guys are using the GXR-M with (or even something like an M9), the case can be made that the benefits of loosing the AA filter make sense. For architecture and studio use, I don't think that would be the case.
Edit: The house example showing mostly color moire, I don't find that objectionable to deal with in post processing. It's the severe maze/pattern examples (which I have not seen examples of yet with the Ricoh), which would really concern me.
I'm glad that the posted images have stimulated this discussion. I'm new to the forum in terms of posts, but have lurked here for a couple of years, and gleaned some knowledge over that time, nuch of it from contributors to this thread. I've been using the M-mount to supplement my D700 when I want a smaller package. Most of my shooting is while travelling, but I happened to get this lens in the winter in Michigan, and won't be travelling abroad for awhile. Michigan is full of flatly lit subjects with high contrast edges, and is a breeding ground for aliasing color artifacts, as I see them on every leafless branch. I chose the above subject (my house, which does not have any rainbow colors at the edges Kit ) to systematically demonstrate to myself what I've been seeing because it would be obvious, and I'd know what apertures to use in order to mitigate against it. The artifact is there, and I imagine it would be annoying to anyone photographing architectural subjects with straight lines, or a subject with fine detail using a lens that can easily resolve it. I could post up several other photos and crops of everyday subjects showing the same phenomenon. At least with irregular subjects, noise reduction software mixes up the aliasing with noise and it is reduced, but this does kill some of the fine detail. For straight lines, you're just stuck with it, and it is apparent even with web sizing. This degree of aliasing simply does not show up when I shoot either my ZF 100MP or ZF.2 35/1.4 on the D700, under the same tripod mounted conditions.
Tariq seemed to sum it up best in this statement: "one must choose their poison with B&W sensitive sensors - sharper images with possible artifacts that cannot always be corrected after the fact OR softer images that require sharpening." I chose the GXR system because its drawbacks are far outweighed by its advantages, particularly regarding ergonomics, control, focus assist, lens compatibility, and otherwise neutral RAW files that are quick to process. It just sucks to have to deal with the aliasing for some subjects.
Here are a few photos where the GXR performed admirably, which I wouldn't have captured with a DSLR due to its major drawback-- it's a pain to carry everywhere.
uscmatt99 wrote:
Michigan is full of flatly lit subjects with high contrast edges, and is a breeding ground for aliasing color artifacts, as I see them on every leafless branch.
I would really be curious to see an example of moire which shows up in this situation - fine, irregularly shaped twigs and branches on trees. This is a subject I photograph often and I'm trying to decide between the Nikon D800/ D800E. I have shot this subject matter with an AA-less Leaf 16MP MF digital back in the past and never ran into a moire issue, though I only tested the back for a short time.
Tariq Gibran wrote:
I would really be curious to see an example of moire which shows up in this situation - fine, irregularly shaped twigs and branches on trees. This is a subject I photograph often and I'm trying to decide between the Nikon D800/ D800E. I have shot this subject matter with an AA-less Leaf 16MP MF digital back in the past and never ran into a moire issue, though I only tested the back for a short time.
Here's a RAW from the ZM25/GXR combo converted in Aperture and downsized by Smugmug, followed by a 100% crop. It was handheld at 1/217 second, ISO 200. This fortunately isn't nearly as obvious as the house example. It can be a little more intense than this, but I think the crop is pretty representative. I've found it's more easily seen and distracting with pure white skies, but not really an issue to me with a blue or textured sky.
uscmatt99 wrote:
Here's a RAW from the ZM25/GXR combo converted in Aperture and downsized by Smugmug, followed by a 100% crop. It was handheld at 1/217 second, ISO 200. This fortunately isn't nearly as obvious as the house example. It can be a little more intense than this, but I think the crop is pretty representative. I've found it's more easily seen and distracting with pure white skies, but not really an issue to me with a blue or textured sky.
Thanks for the compliment Edward. These pics were at one of those resorts that send you home fatter and redder than when you arrived, and I'm guessing it's not representative of a day in the life of most native Jamaicans.... The last photo is just a posed shot of my wife and I. She loves that type of candid photography, and we spent a day going all over the resort with the GXR and a table-top tripod doing pseudo-candids like that. Here's one more I liked, then I'm done polluting the thread with images
) to systematically demonstrate to myself what I've been seeing because it would be obvious, and I'd know what apertures to use in order to mitigate against it. The artifact is there, and I imagine it would be annoying to anyone photographing architectural subjects with straight lines, or a subject with fine detail using a lens that can easily resolve it. I could post up several other photos and crops of everyday subjects showing the same phenomenon. At least with irregular subjects, noise reduction software mixes up the aliasing with noise and it is reduced, but this does kill some of the fine detail. For straight lines, you're just stuck with it, and it is apparent even with web sizing. This degree of aliasing simply does not show up when I shoot either my ZF 100MP or ZF.2 35/1.4 on the D700, under the same tripod mounted conditions.