Pages: [1]   Go Down

Author Topic: Effects of AA-filtering  (Read 3991 times)

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Effects of AA-filtering
« on: January 10, 2010, 04:49:25 am »

Hi,

Considering a recent thread on this forum I tried to dig up some comparable test shots with and without AA-filtering, this is not easy.

I realized that http://www.imaging-resource.com/ has a good archive of test shots of different cameras. Two cameras that have chosen a different approach to OLP filtering were the Canon 1DsII and the Kodak DCS Pro SLR/n. These cameras were both full frame and of similar resolution. I would have preferred to compare images from contemporary cameras like the Leica M9 and the Canon 5DII or the Nikon 3DX, but I have not found relevant and comparable images on the net. The same applies to the Sigma DSLRs with the Foevon sensor. RAW-files would be preferable, but this is the best I could find.

Test target shots can be downloaded for both:

http://www.imaging-resource.com/PRODS/SLRN...SLRNRES2F95.HTM

http://www.imaging-resource.com/PRODS/EOS1...DS2RES4992F.HTM

I downloaded both and adjusted exposure in Lightroom so the white areas were similar. From that I used comparison mode in Lightroom. I made three comparisons:

No extra sharpening on either (Canon to the left):
[attachment=19315:NoExtraSharpen.jpg]

The next I tried was to sharpen the Canon image until some artifacts were showning up, the Kodak image was untouched (Kodak to the left):
[attachment=19316:OnlyCanonSharpened.jpg]

Finally I applied the same sharpening to both images (Canon to the left):
[attachment=19317:SameSharpen.jpg]

Best regards
Erik
« Last Edit: January 10, 2010, 05:21:45 am by ErikKaffehr »
Logged
Erik Kaffehr
 

SeanFS

  • Full Member
  • ***
  • Offline Offline
  • Posts: 114
    • http://www.seanshadbolt.co.nz
Effects of AA-filtering
« Reply #1 on: January 10, 2010, 11:14:28 pm »

Quote from: ErikKaffehr
Hi,

Considering a recent thread on this forum I tried to dig up some comparable test shots with and without AA-filtering, this is not easy.

I realized that http://www.imaging-resource.com/ has a good archive of test shots of different cameras. Two cameras that have chosen a different approach to OLP filtering were the Canon 1DsII and the Kodak DCS Pro SLR/n. These cameras were both full frame and of similar resolution. I would have preferred to compare images from contemporary cameras like the Leica M9 and the Canon 5DII or the Nikon 3DX, but I have not found relevant and comparable images on the net. The same applies to the Sigma DSLRs with the Foevon sensor. RAW-files would be preferable, but this is the best I could find.

Test target shots can be downloaded for both:

http://www.imaging-resource.com/PRODS/SLRN...SLRNRES2F95.HTM

http://www.imaging-resource.com/PRODS/EOS1...DS2RES4992F.HTM

I downloaded both and adjusted exposure in Lightroom so the white areas were similar. From that I used comparison mode in Lightroom. I made three comparisons:

No extra sharpening on either (Canon to the left):
[attachment=19315:NoExtraSharpen.jpg]

The next I tried was to sharpen the Canon image until some artifacts were showning up, the Kodak image was untouched (Kodak to the left):
[attachment=19316:OnlyCanonSharpened.jpg]

Finally I applied the same sharpening to both images (Canon to the left):
[attachment=19317:SameSharpen.jpg]

Best regards
Erik
Thanks for posting,
I think this test is largely historical now. I have owned both cameras and can say that while at the base iso's quality was similar, with an edge to the kodak for colour ( that was something Kodak seemed to have really sorted out well ), the similarities ended there. The Canon's low noise levels and  high iso capabilities, power handling, decent jpegs and design features put it in another league, the dynamic range was a touch better too. I rarely saw moire on the Kodak, although the 14n could be bad , something was improved on the 14nx/slrn - possibly the thickness of the modified IR filter. Rainer V also used the Kodak and did extensive testing of the two cameras and concluded the lack of AA filter gave the kodak similar resolution to the Canon, with the advantage ( of the time ) of better Nikon wide lenses.  I have seen moire in very fine fabrics with the 1ds2, so an AA filter doesn't mean you don't get it, but it was less serious than the slr/n in the same circumstance .
I thought that when I tested my 1ds3 against the two I was seeing the same slight increment in resolution with slightly punchier colour , but canon's own literature said the 1ds3 didn't have quite as much dynamic range as the 1ds 2 at base iso, so it felt in a way one step at least  had been taken back to the 14n, of course software improvements have since improved the out put from both earlier cameras.
Logged

Plekto

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 551
Effects of AA-filtering
« Reply #2 on: January 11, 2010, 12:05:43 am »

But you DO know that black and white resolution charts are worthless, don't you?

You need to compare the same charts in each primary color to compare different cameras, because each behaves differently.  Unfortunately, almost nobody does this sort of test, despite the fact that almost all DSLRs are used in color.
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Effects of AA-filtering
« Reply #3 on: January 11, 2010, 01:25:18 am »

Hi,

The idea was to see what differences there are between two cameras one having an AA-filter and the other not. It's hard to find adequate test shots of cameras without AA-filters made at lab conditions and adequate test targets.

The questions I ask are essentially:

1) Is artifacting visible?
2) What kind of difference is there in sharpness
3) Can the sharpness lost to AA-filtering be regained by normal capture sharpening methods

I'd much prefer to have some newer files, preferably in RAW, but I have seen very few test shots on non AA-filtered cameras (like the Leica M9).

I don't really see anything wrong about using BW test patterns. I can see that test patterns overemphasize aliasing errors.


Best regards
Erik

Quote from: Plekto
But you DO know that black and white resolution charts are worthless, don't you?

You need to compare the same charts in each primary color to compare different cameras, because each behaves differently.  Unfortunately, almost nobody does this sort of test, despite the fact that almost all DSLRs are used in color.
Logged
Erik Kaffehr
 

Bart_van_der_Wolf

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 8913
Effects of AA-filtering
« Reply #4 on: January 11, 2010, 08:26:34 am »

Quote from: Plekto
But you DO know that black and white resolution charts are worthless, don't you?

No, I didn't know that. Apparently the folks at the ISO standards committee also were not informed ..., they mainly recommend achromatic targets.

Quote
You need to compare the same charts in each primary color to compare different cameras, because each behaves differently.  Unfortunately, almost nobody does this sort of test, despite the fact that almost all DSLRs are used in color.

Shooting R/G/B charts is an interesting academic pastime, but it is normally not very useful in real life, unless one shoots objects with primary colors of course. The vast majority of natural colors offer spectral mixes of all sorts of color wavelengths. Interestingly the largest component of color is usually its luminosity. Check your images in e.g. a Lab color mode, and see how low the average resolution in the chroma channels is. The human eye also has a higher acuity/resolution for luminance than for chroma.

Cheers,
Bart
Logged
== If you do what you did, you'll get what you got. ==

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Effects of AA-filtering
« Reply #5 on: January 11, 2010, 09:48:51 am »

Bart,

Thanks a lot, so I don't need to feel like a fool!

Erik



Quote from: BartvanderWolf
No, I didn't know that. Apparently the folks at the ISO standards committee also were not informed ..., they mainly recommend achromatic targets.



Shooting R/G/B charts is an interesting academic pastime, but it is normally not very useful in real life, unless one shoots objects with primary colors of course. The vast majority of natural colors offer spectral mixes of all sorts of color wavelengths. Interestingly the largest component of color is usually its luminosity. Check your images in e.g. a Lab color mode, and see how low the average resolution in the chroma channels is. The human eye also has a higher acuity/resolution for luminance than for chroma.

Cheers,
Bart
Logged
Erik Kaffehr
 

aaykay

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 359
Effects of AA-filtering
« Reply #6 on: January 11, 2010, 12:51:55 pm »

Another equally important factor is the lens employed in either case.  Many of these tests are done on 2 different cameras with 2 different lenses to come up with a judgement call on what the *camera* is capable of.

For instance, comparing a Nikon D3X equipped with say the 135mm Nikkor and comparing its output to say the Sony A900 equipped with say the 135mm Carl Zeiss, is not an apples-to-apples comparison between 2 *cameras*.  It is simply comparing 2 different Camera+lens combinations.

Sure, a comparison that say equips a Nikon D3X with say the Sigma 50mm f/1.4 HSM and then equips the Sony A900 with the same Alpha-mount version of the Sigma 50mm f/1.4 HSM, will at least equalize one of the parameters of the test, thus enabling a comparison between the 2 *cameras*.  There are even stresses that various manufacturers place on specific lens designs, where some of them are designed for great wide-open performance with some degradation when one stops down, while others are designed to improve as one stops down....thus when 2 completely different prime lenses having the same FL, from 2 completely different manufacturers are stopped down to say f/8....we still have additional points of differences between the 2 camera+lens combos that are being compared.

Unfortunately, most tests (specifically from sites like dpr) are doing different camera+lens combinations and then projecting that as results from the *camera* and then making judgment calls on camera A being superior to camera B from a resolution/sharpness standpoint.

However, this AA filter equipped result versus the result without an AA-filter is an interesting one - and controversial.
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Effects of AA-filtering
« Reply #7 on: January 11, 2010, 02:22:36 pm »

Hi,

Thanks for comments!


In general I would not see the fact that different lenses are used as a major issue. Very good lenses tend to be nearly the diffraction limit around f/8. This really means that MTF is pretty close on good lenses around f/8 and on axis, it's simply the diffraction limit.

Regarding the AA vs non AA issue, this is exactly what this discussion is about. There are a few issues with the comparison:

1) They are based on JPEG, raw would be preferable
2) The Canon probably has microlenses while the Kodak may not have microlenses
3) The cameras are pretty old and don't have latest development in demosaicing


Best regards
Erik




Quote from: aaykay
Another equally important factor is the lens employed in either case.  Many of these tests are done on 2 different cameras with 2 different lenses to come up with a judgement call on what the *camera* is capable of.

For instance, comparing a Nikon D3X equipped with say the 135mm Nikkor and comparing its output to say the Sony A900 equipped with say the 135mm Carl Zeiss, is not an apples-to-apples comparison between 2 *cameras*.  It is simply comparing 2 different Camera+lens combinations.

Sure, a comparison that say equips a Nikon D3X with say the Sigma 50mm f/1.4 HSM and then equips the Sony A900 with the same Alpha-mount version of the Sigma 50mm f/1.4 HSM, will at least equalize one of the parameters of the test, thus enabling a comparison between the 2 *cameras*.  There are even stresses that various manufacturers place on specific lens designs, where some of them are designed for great wide-open performance with some degradation when one stops down, while others are designed to improve as one stops down....thus when 2 completely different prime lenses having the same FL, from 2 completely different manufacturers are stopped down to say f/8....we still have additional points of differences between the 2 camera+lens combos that are being compared.

Unfortunately, most tests (specifically from sites like dpr) are doing different camera+lens combinations and then projecting that as results from the *camera* and then making judgment calls on camera A being superior to camera B from a resolution/sharpness standpoint.

However, this AA filter equipped result versus the result without an AA-filter is an interesting one - and controversial.
« Last Edit: January 11, 2010, 02:28:33 pm by ErikKaffehr »
Logged
Erik Kaffehr
 

Plekto

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 551
Effects of AA-filtering
« Reply #8 on: January 11, 2010, 10:25:10 pm »

Quote from: BartvanderWolf
No, I didn't know that. Apparently the folks at the ISO standards committee also were not informed ..., they mainly recommend achromatic targets.
I suspect that the reason they suggest it is to try to rule out as many variables as possible.  And of course, absolute resolution is a good bragging right in the adverts.  But the problem still remains that different sensors see colors differently.   Sometimes it's bad enough to cause sharp and blurry areas in the same picture.  This is especially true of certain flowers and similarly colored objects like street signs and such that are nearly uniform in color.  

Nobody really tests for this, though, which is unfortunate.  They should.

The other comments about lenses are also valid.  Obviously two different DSLR cameras with different lenses will produce essentially meaningless results if you're planning on doing a comparison.  Yet I see just about every review site doing this.  It's sloppy work and they should know better by now.
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Effects of AA-filtering
« Reply #9 on: January 12, 2010, 12:07:07 am »

Hi,

I don't agree, sorry.

1) This discussion is not about testing cameras/lenses but about aliasing effect
2) Good lenses are effectively diffraction limited at around f/8 where most tests are done
3) Don't see that color seeing in sensors can blur a picture, axial (longitudal) chromatic aberration can but it's usually gone at medium apertures.

Admittedly, it seems that lenses play bigger role with shrinking pixel pitch. And yes some lenses are better even when stopped down. Cameras like the Canon 7D push the limits.


This is Nikon 50/1.4G tested at Photozone:

http://www.photozone.de/images/8Reviews/le..._50_14g/mtf.gif

and this is the Zeiss 50/1.4 ZF

http://www.photozone.de/images/8Reviews/le...f_50_14/mtf.gif

The Zeiss is obviously better but they are quite closed when stopped down to f/8 where most tests at both DPReview and Imaging Resource are done.


Best regards




Quote from: Plekto
I suspect that the reason they suggest it is to try to rule out as many variables as possible.  And of course, absolute resolution is a good bragging right in the adverts.  But the problem still remains that different sensors see colors differently.   Sometimes it's bad enough to cause sharp and blurry areas in the same picture.  This is especially true of certain flowers and similarly colored objects like street signs and such that are nearly uniform in color.  

Nobody really tests for this, though, which is unfortunate.  They should.

The other comments about lenses are also valid.  Obviously two different DSLR cameras with different lenses will produce essentially meaningless results if you're planning on doing a comparison.  Yet I see just about every review site doing this.  It's sloppy work and they should know better by now.
« Last Edit: January 12, 2010, 12:20:02 am by ErikKaffehr »
Logged
Erik Kaffehr
 

Plekto

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 551
Effects of AA-filtering
« Reply #10 on: January 12, 2010, 02:14:15 am »

http://www.ddisoftware.com/sd14-5d/

Now take a look at the color wheels and resolution charts about 15-20% the way down the page.  The three paragraphs below it describe what's happening far better than I can here.

Halfway down you see the pictures of the flowers side by side.  Here's the excerpt from the paragraph below that I feel is important:

"Where the SD14 holds consistent sharpness across the frame, the 5D has smudged over a bit of detail in areas notoriously problematic for Bayer sensors such as the red carnation and even the white flower where edge detail is being lost to the AA filter.  Looking at the 5D shot, you'd be tempted to believe that the red flower is just a little out of focus because it's in front of (or behind) the other flower due to it not looking as sharp.  In reality, all the flowers in the above crop are in the same plane relative to the lens. "

Parts of the flowers are in in focus and parts aren't, yet the focus of the lens isn't a factor - the way the sensor, AA filter, and camera are processing and blurring the data is what's causing this effect.  I find it to be very annoying, to be honest.  Parts of the image look sharp and parts look blurry, despite being right next to each other.

I don't see any of the review sites testing for this sort of thing, though.  It honestly bothers me far more than most of the other issues, but perhaps that's because I shoot mostly black and white film where sharp and highly defined results are the norm.  Most Bayer sensors after being put through a consumer grade lens and an overly aggressive AA filter look almost like someone hit the blur feature in Photoshop on purpose on bits and parts of the image.

It's why I like the look of the Sigma cameras a bit more - it looks more like film in how it's consistent from one place to another.  Of course, it's also like dropping Velvia 50 in a half-frame camera.  Beautiful *little* pictures.  Sigh.
« Last Edit: January 12, 2010, 02:15:26 am by Plekto »
Logged

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
Effects of AA-filtering
« Reply #11 on: January 12, 2010, 01:17:03 pm »

Quote from: Plekto
http://www.ddisoftware.com/sd14-5d/
"Where the SD14 holds consistent sharpness across the frame, the 5D has smudged over a bit of detail in areas notoriously problematic for Bayer sensors ...
Pletko,

    the color handling differences you talk about are quite likely in comparisons between X3 and Bayer CFA cameras, but I cannot see how it would be much of an issue when comparing cameras with sensors that use the standard Bayer CFA pattern. Well, maybe a little bit due to different deBayering algorithms, which is part of why RAW files would be better, as Erik said.
« Last Edit: January 12, 2010, 01:17:56 pm by BJL »
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Effects of AA-filtering
« Reply #12 on: January 12, 2010, 01:17:56 pm »

Hi,

Thanks for the article. I'd suggest that the article makes a clear and good argument for the Foveon sensor over Bayer. Regarding the resolution in red I see what you mean. It is quite true that only a fourth of the RGB sensors are red, while the Foevon all sensels are sensitive to color. That said I have seen a similar issue comparing scanned Velvia with my Sony Alpha 900, good resolution in the reds on the Sony but bad resolution on the Velvia, even where Velvia was quite good in the greens.

What I would point out that the original posting was about comparing AA-filtered images with non-AA filtered images. If you check the Siemens stars it is very obvious that there are aliasing effects on the Foveon which by and large are not seen on the 5D images. Another question is of course if the aliasing that can be seen on the Siemens stars would negatively affect real world images. The other question may be how much sharpening a image with AA-filtering takes before artifacts start to show up.

The Foveon concept did not really take on. The only Foevon vendor is Sigma, AFAIK. I guess that we will not have the opportunity to find out which potential the Foevon concept really has.

Best reagrds
Erik


Quote from: Plekto
http://www.ddisoftware.com/sd14-5d/

Now take a look at the color wheels and resolution charts about 15-20% the way down the page.  The three paragraphs below it describe what's happening far better than I can here.

Halfway down you see the pictures of the flowers side by side.  Here's the excerpt from the paragraph below that I feel is important:

"Where the SD14 holds consistent sharpness across the frame, the 5D has smudged over a bit of detail in areas notoriously problematic for Bayer sensors such as the red carnation and even the white flower where edge detail is being lost to the AA filter.  Looking at the 5D shot, you'd be tempted to believe that the red flower is just a little out of focus because it's in front of (or behind) the other flower due to it not looking as sharp.  In reality, all the flowers in the above crop are in the same plane relative to the lens. "

Parts of the flowers are in in focus and parts aren't, yet the focus of the lens isn't a factor - the way the sensor, AA filter, and camera are processing and blurring the data is what's causing this effect.  I find it to be very annoying, to be honest.  Parts of the image look sharp and parts look blurry, despite being right next to each other.

I don't see any of the review sites testing for this sort of thing, though.  It honestly bothers me far more than most of the other issues, but perhaps that's because I shoot mostly black and white film where sharp and highly defined results are the norm.  Most Bayer sensors after being put through a consumer grade lens and an overly aggressive AA filter look almost like someone hit the blur feature in Photoshop on purpose on bits and parts of the image.

It's why I like the look of the Sigma cameras a bit more - it looks more like film in how it's consistent from one place to another.  Of course, it's also like dropping Velvia 50 in a half-frame camera.  Beautiful *little* pictures.  Sigh.
« Last Edit: January 12, 2010, 01:33:58 pm by ErikKaffehr »
Logged
Erik Kaffehr
 

Plekto

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 551
Effects of AA-filtering
« Reply #13 on: January 12, 2010, 08:13:12 pm »

Quote from: ErikKaffehr
That said I have seen a similar issue comparing scanned Velvia with my Sony Alpha 900, good resolution in the reds on the Sony but bad resolution on the Velvia, even where Velvia was quite good in the greens.
Yeah, Fuji's color balance is just plain "off" for many of their films.  Velvia 50 is great quality-wise, though the colors are off.  I suspect your scanner might not be picking up all of the reds, though, since if anything Velvia seems to over-saturate reds and greens.

Quote
What I would point out that the original posting was about comparing AA-filtered images with non-AA filtered images. If you check the Siemens stars it is very obvious that there are aliasing effects on the Foveon which by and large are not seen on the 5D images.

I'm not quite so sure that that's a fair assessment.  I see aliasing effects on all of them, and while the SD14 are probably more noticeable due to their geometric nature and how they create patterns instead of random mush, the other two Bayer sensors just simply blur it all out beyond a certain point.  The image looks fine and then it looks blurry. I suspect there's some pre-processing of the data in-camera going on in these cases.   The SD seems to be doing nothing to the data.  This of course might be a good or a bad thing depending upon your point of view.

As for applying this to different cameras, I think it would be valid.  I can see differences between the two Canon models, and I'm sure the Fuji sensors will behave differently than the Nikon or the Sony designs.  Right now we're assuming that all Bayer patterns are equivalent.  I doubt if they are.

Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Effects of AA-filtering
« Reply #14 on: January 13, 2010, 12:28:54 am »

Quote from: Plekto
Yeah, Fuji's color balance is just plain "off" for many of their films.  Velvia 50 is great quality-wise, though the colors are off.  I suspect your scanner might not be picking up all of the reds, though, since if anything Velvia seems to over-saturate reds and greens.
I don't really think that issue is scanner related, I tried to check it different ways. I also tried to take a photograph with an enlarging lens on bellows. The results were very similar. The result of my testing was essentially that I'm not going to use my Pentax 67 equipment for real shoots. The Pentax is built like a tank, which is very nice, the A900 is also bit like a tank. But the Pentax is like a Sherman and the A900 is more like an Abrams :-)

Regarding aliasing visible on the star patterns it is exactly what this topic is about. It is not about Bayer vs. Non Bayer patterns. The sole idea of the AA-filter is to reduce aliasing effects. Now, with a Bayer pattern a colorful moiré pattern will arise at some frequencies, with the Foevon concept this would not be the case. According to signal processing theory the Siemens star patterns would have aliasing, and they certainly have.

As a side note, then Nikon D70 had a very weak AA-filter, or even possibly none and that camera had also quite excessive aliasing, very visible on Siemens star patterns.

Best regards
Erik
« Last Edit: January 13, 2010, 12:35:25 am by ErikKaffehr »
Logged
Erik Kaffehr
 

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Effects of AA-filtering
« Reply #15 on: January 13, 2010, 01:02:17 am »

Quote from: Plekto
As for applying this to different cameras, I think it would be valid.  I can see differences between the two Canon models, and I'm sure the Fuji sensors will behave differently than the Nikon or the Sony designs.  Right now we're assuming that all Bayer patterns are equivalent.  I doubt if they are.
The major difference between the 5D and the 20D is the pixel pitch. The 5D has a quite large pixel pitch and the 20D has a smaller one. Reducing the pixel size is helpful in reducing aliasing effects at least when lens is stopped down beyond optimum aperture.

Fuji sensors are different. The grid is rotated 45 degrees so it's not aligned with horizontal and vertical lines. In their brochures the pixels are octagonal, which also may be helpful in reducing aliasing as they would have some overlap. Unfortunately it seems that Fuji is also not a major player. It seems that the bigger players all use some Bayer variation.

Best regards
Erik
« Last Edit: January 13, 2010, 01:55:12 am by ErikKaffehr »
Logged
Erik Kaffehr
 

SeanFS

  • Full Member
  • ***
  • Offline Offline
  • Posts: 114
    • http://www.seanshadbolt.co.nz
Effects of AA-filtering
« Reply #16 on: January 14, 2010, 03:03:54 pm »

Quote from: ErikKaffehr
Hi,

I don't agree, sorry.

1) This discussion is not about testing cameras/lenses but about aliasing effect
2) Good lenses are effectively diffraction limited at around f/8 where most tests are done
3) Don't see that color seeing in sensors can blur a picture, axial (longitudal) chromatic aberration can but it's usually gone at medium apertures.

Admittedly, it seems that lenses play bigger role with shrinking pixel pitch. And yes some lenses are better even when stopped down. Cameras like the Canon 7D push the limits.


This is Nikon 50/1.4G tested at Photozone:

http://www.photozone.de/images/8Reviews/le..._50_14g/mtf.gif

and this is the Zeiss 50/1.4 ZF

http://www.photozone.de/images/8Reviews/le...f_50_14/mtf.gif

The Zeiss is obviously better but they are quite closed when stopped down to f/8 where most tests at both DPReview and Imaging Resource are done.


Best regards

The sharpest lens I have on my 1ds3 is a Nikon 50mm 1.4 on an adaptor - testing lenses when I first bought it this came as a huge surprise! I only have a few Canon primes to compare - a 50mm macro, 85mm 1.2 and 100mm macro all excellent lenses. Tested at infinity focus , of course things may change at other distances.

Logged

aaykay

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 359
Effects of AA-filtering
« Reply #17 on: January 14, 2010, 09:49:16 pm »

Quote from: ErikKaffehr
The Foveon concept did not really take on. The only Foevon vendor is Sigma, AFAIK. I guess that we will not have the opportunity to find out which potential the Foevon concept really has.

The Sony alpha rumors site had a posting a while back, about a new patent filed by Sony for Foveon type sensor technology.  Don't know how this is different from Foveon but I would be interested to see where Sony takes this:

http://www.freepatentsonline.com/y2009/0303371.html
Logged
Pages: [1]   Go Up