11-26-2010, 06:33 AM
1. the strength of an AA filter is a matter of taste. There's no wrong or right here. Leica has none at all for instance and that's perfectly fine for what they want to achieve (ultimate sharpness). Kodak did the same. There're also applications where you just don't want/need an AA filter - e.g. astro photography.
2. there're many cameras with weak AA filters (A33, NEX-5, all MFTs, D7000, K10D) and there're next to zero user complaints about this. That's a fact and no theory.
3. yes, a weak AA filter causes moirees in certain situations - that's not a theoretical aspect but a practical one. However, this is often blown way out-of-proportions (-> (2)). Markus showed me some converted (Leica M) RAWs via CaptureOne. C1 seems to have a fairly efficient anti-moiree algorithm (unlike Silkypix) and CaptureOne is one of the sharpest RAW converters out there. It is not able to eliminate the problem altogether of course. It is the question whether this is an issue for your personal photography. It may be fore yours and your taste.
4. the K5 has a strong AA filter. I don't like it personally and it is counterproductive for the lens tests. I don't want to measure blur after all. I don't test the camera here - I test lenses.
5. the difference between the K5 and D7000 is certainly higher than just 4% in RAW mode. It is higher than 4% vs the A33 already.
6. at 16mp many lenses start to act as a low pass filter because they don't reach the sensor performance (due to the high pixel density digicams had no AA filter since the dawn of time). That's certainly true regarding the borders performance of most lenses.
7. the K10D has no vertical filtering. It can reach resolution figures well beyond Nyquist here. Horizontally the filter is about as strong as on the K5. In JPEG mode both you won't notice the difference (because the K10D has a dismal JPEG engine). In RAW mode that's rather easily visible - the K10D delivers pin sharp results here (and yes, sometimes moirees).
[quote name='Class A' timestamp='1290748932' post='4451']
The AA filter in the K-5 is a bit stronger than the one in the D7000 and that's good.
GordonBGood reckons the [url="http://forums.dpreview.com/forums/read.asp?forum=1036&message=37005086"]difference in resolution might be around 4%[/url] but this is hard to judge on the basis of JPEGs where the sharpness parameters are not directly comparable.
Note that the D7000 exhibits colour moiré which is bad and due to an AA filter that is too weak.
You might also be interested in [url="http://www.pentaxforums.com/forums/1280870-post113.html"]Falk Lumo's first assessment[/url] in which he states for the K-5 "A weaker as well as a stronger AA filter wouldn't be good". He also says "[url="http://www.pentaxforums.com/forums/1280938-post118.html"]the K-5 renders tack-sharp images[/url]".
So in conclusion there is nothing wrong with the AA filter of the K-5. Were it any weaker then all you'd get is spurious detail, i.e., detail that wasn't in the scene in the first place. Spurious contrast looks good, but is wrong.
Klaus, I'd be very interested to learn from you why you think that removing the K-5's AA filter makes sense. What it doesn't let through, should not be recorded by the sensor in the first place. Why do you want to record moiré artefacts?
It might be the case that for your special application involving B&W patterns only it could make sense to process the RAW data prior to demosaicing. Are you doing this? If not then I suspect that results based sensors with weak or no AA filters may not be entirely reliable.
It strikes me as being very odd that you think that the difference between the K-5 (2500 LW/PH) and the K10D (2350 = [url="http://forums.dpreview.com/forums/read.asp?forum=1036&message=37005287"]2500-150[/url]) is just 6% (while there is a nominal 26% difference in # of vertical pixels). Nothing that I saw from the K-5 and read about its resolution indicates that it is possible to defend this statement. It seems that different levels of (over-) sharpening are applied.
There might be something about your methods that I don't understand and I'm curious as to how the apparent mismatch between your's and Falk's findings can be reconciled. Maybe there isn't such a disparity in assessment after all, but to me it certainly appears to be like one.
If there is any chance that your initial quick statement needs more rectification than saying that you need a camera for lab tests then I think you should act quickly because your reputation might be at stake. I would truly hate to see your and your fine site's reputation suffer from a quick, loose comment that wasn't followed up properly.
[/quote]
2. there're many cameras with weak AA filters (A33, NEX-5, all MFTs, D7000, K10D) and there're next to zero user complaints about this. That's a fact and no theory.
3. yes, a weak AA filter causes moirees in certain situations - that's not a theoretical aspect but a practical one. However, this is often blown way out-of-proportions (-> (2)). Markus showed me some converted (Leica M) RAWs via CaptureOne. C1 seems to have a fairly efficient anti-moiree algorithm (unlike Silkypix) and CaptureOne is one of the sharpest RAW converters out there. It is not able to eliminate the problem altogether of course. It is the question whether this is an issue for your personal photography. It may be fore yours and your taste.
4. the K5 has a strong AA filter. I don't like it personally and it is counterproductive for the lens tests. I don't want to measure blur after all. I don't test the camera here - I test lenses.
5. the difference between the K5 and D7000 is certainly higher than just 4% in RAW mode. It is higher than 4% vs the A33 already.
6. at 16mp many lenses start to act as a low pass filter because they don't reach the sensor performance (due to the high pixel density digicams had no AA filter since the dawn of time). That's certainly true regarding the borders performance of most lenses.
7. the K10D has no vertical filtering. It can reach resolution figures well beyond Nyquist here. Horizontally the filter is about as strong as on the K5. In JPEG mode both you won't notice the difference (because the K10D has a dismal JPEG engine). In RAW mode that's rather easily visible - the K10D delivers pin sharp results here (and yes, sometimes moirees).
[quote name='Class A' timestamp='1290748932' post='4451']
The AA filter in the K-5 is a bit stronger than the one in the D7000 and that's good.
GordonBGood reckons the [url="http://forums.dpreview.com/forums/read.asp?forum=1036&message=37005086"]difference in resolution might be around 4%[/url] but this is hard to judge on the basis of JPEGs where the sharpness parameters are not directly comparable.
Note that the D7000 exhibits colour moiré which is bad and due to an AA filter that is too weak.
You might also be interested in [url="http://www.pentaxforums.com/forums/1280870-post113.html"]Falk Lumo's first assessment[/url] in which he states for the K-5 "A weaker as well as a stronger AA filter wouldn't be good". He also says "[url="http://www.pentaxforums.com/forums/1280938-post118.html"]the K-5 renders tack-sharp images[/url]".
So in conclusion there is nothing wrong with the AA filter of the K-5. Were it any weaker then all you'd get is spurious detail, i.e., detail that wasn't in the scene in the first place. Spurious contrast looks good, but is wrong.
Klaus, I'd be very interested to learn from you why you think that removing the K-5's AA filter makes sense. What it doesn't let through, should not be recorded by the sensor in the first place. Why do you want to record moiré artefacts?
It might be the case that for your special application involving B&W patterns only it could make sense to process the RAW data prior to demosaicing. Are you doing this? If not then I suspect that results based sensors with weak or no AA filters may not be entirely reliable.
It strikes me as being very odd that you think that the difference between the K-5 (2500 LW/PH) and the K10D (2350 = [url="http://forums.dpreview.com/forums/read.asp?forum=1036&message=37005287"]2500-150[/url]) is just 6% (while there is a nominal 26% difference in # of vertical pixels). Nothing that I saw from the K-5 and read about its resolution indicates that it is possible to defend this statement. It seems that different levels of (over-) sharpening are applied.
There might be something about your methods that I don't understand and I'm curious as to how the apparent mismatch between your's and Falk's findings can be reconciled. Maybe there isn't such a disparity in assessment after all, but to me it certainly appears to be like one.
If there is any chance that your initial quick statement needs more rectification than saying that you need a camera for lab tests then I think you should act quickly because your reputation might be at stake. I would truly hate to see your and your fine site's reputation suffer from a quick, loose comment that wasn't followed up properly.
[/quote]