Pages: [1] 2   Go Down

Author Topic: Leica M8: another perspective  (Read 12153 times)

BernardLanguillier

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 13983
    • http://www.flickr.com/photos/bernardlanguillier/sets/
Leica M8: another perspective
« on: December 09, 2006, 08:56:36 pm »

Hi there,

I found the latest addition to the M8 articles to be interesting, but a bit lacking in substantiated data in a few key places. Just to cite a few:

- "These numbers make clear that the best image quality actually in the market is achieved with a coefficient around 1 (Canon 35mm format cameras), just like it was in the film era."

This is IMHO un-substantiated and clearly overlooks the known quality issues of Canon when using wide angles on their FF sensors. FF sensor have some advantages in terms of DR (slight) and hight ISO noise (significant), but overall sharpness is not one of them with today's technology. I'd be the first to celebrate if tomorrow's technology fixes the current issues.

- Leica provides us with a system level estimation of their achieved resoluton (50 lp/mm for the combo of sensor and lens), but we do not have similar figures for the Canon/Nikon systems, which makes the comparison weak IMHO.

- "To sum-up, the Leica M8 can be surpassed in terms of sharpness in big prints only by a few cameras and lenses, thanks to their bigger sensors. Of course, perceived sharpness is only a part of the tale."

Although I undertand the general direction, the demonstration doesn't explain us why a camera with a smaller sensor but higher pixel density (like a Canon 400d or Nikon D2x) would not be able to outperfom the M8 at large print sizes although they have a smaller sensor than that of the M8. The "thanks to their larger sensors" part is IMHO un-substantiated, or not explained clearly enough.

I'd love to be corrected in case I have overlooked some useful information in the article.

Regards,
Bernard
« Last Edit: December 09, 2006, 08:59:03 pm by BernardLanguillier »
Logged

Nemo

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 276
Leica M8: another perspective
« Reply #1 on: December 10, 2006, 05:25:52 am »

Dear Bernard,

there are several clues for the resolution of this case:


Quote
Leica provides us with a system level estimation of their achieved resoluton (50 lp/mm for the combo of sensor and lens), but we do not have similar figures for the Canon/Nikon systems, which makes the comparison weak IMHO.

Erwin Puts' articles have several resolution tests, including a comparison between the Canon 5D and the Leica M8.

Quote
This is IMHO un-substantiated and clearly overlooks the known quality issues of Canon when using wide angles on their FF sensors. FF sensor have some advantages in terms of DR (slight) and hight ISO noise (significant), but overall sharpness is not one of them with today's technology. I'd be the first to celebrate if tomorrow's technology fixes the current issues.

Resolution decreases at the corners and borders due to several factors. It depends on the lens and the sensor. The 24x36 sensors by Canon are able to avoid additional degradation at the borders if the lens is good. The vignetting in Canon full frame digital cameras is not worse than in film based cameras. The sensor is not a problem here. Sean Reid has a interesting article about wide-angle lenses in Canon cameras (particularly, the 5D).

Quote
Although I undertand the general direction, the demonstration doesn't explain us why a camera with a smaller sensor but higher pixel density (like a Canon 400d or Nikon D2x) would not be able to outperfom the M8 at large print sizes although they have a smaller sensor than that of the M8. The "thanks to their larger sensors" part is IMHO un-substantiated, or not explained clearly enough.

The smaller the sensor is, the larger the multiplication of real detail must be for a particular print. The Leica M8 must resolve 1.33 more detail (in fact, the same detail in a smaller capture area) than the Canon 5D for a similar print. The higher sampling resolution, powerful lenses and abscense of anti-alias filter help, but it is not enough in large prints (see Erwin Puts' comparative results). Smaller sensors need even larger multiplications or enlargements of the real detail captured.

If the camera has a APS-C sensor, 1.5 times more real detail (lp/mm) must be captured in order to put the same detail on paper. Here the limiting factor could be in the lenses: diffraction apart, many manufacturers offer MTF graphs for APS-C sensors with 10 lp/mm and 30 lp/mm lines, but these lines only can be compared with 6,7 (=10/1.5) lp/mm and 20 (=30/1.5) lp/mm lines for 35mm format lenses. When you look at the MTF graphs of many APS-C lenses you realize that they have not 1,5 more resolutive power at a particular level of contrast. Manufacturers do not show lines for 45 lp/mm (=30*1.5) in the MTF graphs of APS-C lenses, but it is easy to estimate that they would be well below the 30 lp/mm lines of 35mm lenses. This limitation is even worse in the 4/3 format.

Leica's MTF graphs show lines for 5, 10, 20 and 40 lp/mm. Many lenses have 40 lp/mm lines over 50% of contrast level on axis. The equivalent would be APS-C lenses with 60 lp/mm lines over 50% of contrast or 4/3 lenses with 80 lp/mm lines over 50% of contrast !!!
.
« Last Edit: December 12, 2006, 07:26:38 am by Nemo »
Logged

Nemo

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 276
Leica M8: another perspective
« Reply #2 on: December 10, 2006, 05:34:45 am »

Oskar Barnack trick was to offer the same quality of bigger formats by means of high resolutive lenses and emultions. It can be done, but: 1) only if the differences between format are not too big; 2) at moderate print sizes; 3) with very expensive (high quality) lenses (and sensors, actually).

Olympus tried to repeat the trick, but it doesn't work because: 1) A3 prints became more and more popular; 2) the differences in format size are too big; 3) Zuiko lenses are not better performers by a factor of 2, 1.6 or 1.5. The same goes for APS-C based formats. In the case of Canon, they do not pretend the APS-C format to be a competitor of the 35mm format. The smaller format is aimed a cheaper cameras. Nikon, Pentax, Sony... they are commited to the APS-C format, but I doubt if they are trying to compete against Canon's professional line at all.

Olympus offers MTF graphs for 20 and 60 lp/mm (similar to 10 and 30 lp/mm for 35mm format):

http://www.olympus-europa.com/consumer/dslr_6756.htm

Nikon shows MTF graphs for 10 and 30 lp/mm (similar to 7 and 20 lp/mm for 35mm format):

http://www.nikon-image.com/eng/Nikkor_Lens...f-sdx17-55a.htm

It is difficult to make comparisons with 35mm format gear based on Nikon's graphs.
« Last Edit: December 10, 2006, 08:13:46 am by Nemo »
Logged

bjanes

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3387
Leica M8: another perspective
« Reply #3 on: December 10, 2006, 08:27:10 am »

Quote
Oskar Barnack trick was to offer the same quality of bigger formats by means of high resolutive lenses and emultions. It can be done, but: 1) only if the differences between format are not too big; 2) at moderate print sizes; 3) with very expensive lenses (and sensors).
Olympus offers MTF graphs for 20 and 60 lp/mm (similar to 10 and 30 lp/mm for 35mm format):

http://www.olympus-europa.com/consumer/dslr_6756.htm

Nikon shows MTF graphs for 10 and 30 lp/mm (similar to 7 and 20 lp/mm for 35mm format):

http://www.nikon-image.com/eng/Nikkor_Lens...f-sdx17-55a.htm
[{POST_SNAPBACK}][/a]

These MTF graphs are very interesting, but one must realize that they apply only to the lens and not the entire imaging chain. Furthermore, such graphs are usually generated by a computer analysis of the design of the lens and not what is found in an actual sample where manufacturing defects may degrade the performance.

The Garcia article quotes that the Leica sensor has a Nyquist of 73.5 lp/mm and an enlargement factor of 11.3 times for an A4 sized print. Using his reciprocal formula (which is accurate only for MTFs of about 10%), system resolution would be 42 lp/mm for a lens with 100 lp/mm and 54 lp/mm for a lens resolving 200 lp/mm. For a MTF of 50% (considered to be the figure that best correlated with perceived sharpness), one must do a Fourier analysis and convolution, and I do not know if the results would be more or less than implied by the reciprocal formula.

[a href=\"http://www.normankoren.com/Tutorials/MTF.html]http://www.normankoren.com/Tutorials/MTF.html[/url]

According to SQF analysis, the most important frequencies for an 8 by 12 inch print with a 35mm camera (m = 8) are between 4 and 16 lp/mm and the corresponding frequencies for m = 11.3 would by 6 and 23 lp/mm.

http://bobatkins.com/photography/technical/mtf/mtf4.html

The Leica sensor has plenty of resolution for an 8 inch wide print, but the apparent sharpness would depend on the system MTF 50 between 6 and 23 lp/mm. We simply do not have these data, and any analysis with questionable data is useless. How the Leica would compare with the Olympus or Nikon cameras can not be predicted with any certainty without additional data.
Logged

BernardLanguillier

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 13983
    • http://www.flickr.com/photos/bernardlanguillier/sets/
Leica M8: another perspective
« Reply #4 on: December 10, 2006, 08:34:03 am »

Quote
Resolution decreases at the corners and borders due to several factors. It depends on the lens and the sensor. The 24x36 sensors by Canon are able to avoid additional degradation at the borders if the lens is good. The vignetting in Canon full frame digital cameras is not worse than in film based cameras. The sensor is not a problem here. Sean Reid has a interesting article about wide-angle lenses in Canon cameras (particularly, the 5D).
[a href=\"index.php?act=findpost&pid=89670\"][{POST_SNAPBACK}][/a]

This is probably true, but the article focuses on comparing digital platforms and concludes that FF delivers a superior image quality compared to APS. This just isn't correct IMHO.

Quote
If the camera has a APS-C sensor, 1.5 times more real detail (lp/mm) must be captured in order to put the same detail on paper. Here the limiting factor could be in the lenses: many manufacturers offer MTF graphs for APS-C sensors with 10 lp/mm and 30 lp/mm lines, but these lines only can be compared with 6,7 (=10/1.5) lp/mm and 20 (=30/1.5) lp/mm lines of 35mm format lenses. When you look at the MTF graphs of many APS-C lenses you realize that they have not 1,5 more resolutive power. Manufacturers do not show lines for 45 lp/mm (=30*1.5) in the MTF graphs of APS-C lenses, but it is easy to estimate that they would be well below the 30 lp/mm lines of 35mm lenses.

Leica's MTF graphs show lines for 5, 10, 20 and 40 lp/mm. Many lenses have 40 lp/mm lines over 50% of contrast level on axis. The equivalent would be APS-C lenses with 60 lp/mm lines over 50% of contrast or 4/3 lenses with 80 lp/mm lines over 50% of contrast !!!
[a href=\"index.php?act=findpost&pid=89670\"][{POST_SNAPBACK}][/a]

Thank you for these details.

Am I correct to write that these MTF charts are valid for the full aperture of the lenses? How does the result change when the lenses are stopped down to their optimal f stop?

I am getting images that are extremely sharp with my d2x when viewed at 100% on screen when using the lenses in their sweet spot and have a hard time imagining that a better result is possible. Besides, all the comparisons that I have seen between the d2x and the 1ds2 show that the d2x delivers a sharpness per pixel that is superior to that of the 1ds2. Is that only the result of the weaker AA filter used on the d2x?

Regards,
Bernard

Nemo

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 276
Leica M8: another perspective
« Reply #5 on: December 10, 2006, 09:24:58 am »

Quote
According to SQF analysis, the most important frequencies for an 8 by 12 inch print with a 35mm camera (m = 8) are between 4 and 16 lp/mm and the corresponding frequencies for m = 11.3 would by 6 and 23 lp/mm.

The Leica sensor has plenty of resolution for an 8 inch wide print, but the apparent sharpness would depend on the system MTF 50 between 6 and 23 lp/mm. We simply do not have these data, and any analysis with questionable data is useless. How the Leica would compare with the Olympus or Nikon cameras can not be predicted with any certainty without additional data.

That's right, but it depends of many other factors: print size, subject, effective distance of observation (of the picture)... SQF analysis is interesting, because it includes the subjective perception as a variable in evaluating image quality. However, that subjective perception depends again on many factors.

I can see the difference, in most cases, between a print made from a 6x6 frame and a print made from a 24x36 frame. It depends, I repeat it, on print size, subjecto, real viewing distance (people tend to approach, looking for small details in a big print)... etc...
Logged

Nemo

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 276
Leica M8: another perspective
« Reply #6 on: December 10, 2006, 09:27:03 am »

Quote
Am I correct to write that these MTF charts are valid for the full aperture of the lenses? How does the result change when the lenses are stopped down to their optimal f stop?


Usually they report MTF curves for different apertures, at least at full aperture and an intermediate aperture.
Logged

bjanes

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3387
Leica M8: another perspective
« Reply #7 on: December 10, 2006, 10:34:41 am »

Quote
That's right, but it depends of many other factors: print size, subject, effective distance of observation (of the picture)... SQF analysis is interesting, because it includes the subjective perception as a variable in evaluating image quality. However, that subjective perception depends again on many factors.

I can see the difference, in most cases, between a print made from a 6x6 frame and a print made from a 24x36 frame. It depends, I repeat it, on print size, subjecto, real viewing distance (people tend to approach, looking for small details in a big print)... etc...
[a href=\"index.php?act=findpost&pid=89693\"][{POST_SNAPBACK}][/a]

SQF is interesting, but I agree that looking at prints optimized for size and viewing conditions is the final arbiter of image quality.

IMHO, the extinction resolution figures of a high contrast test chart with default camera sharpening, such as on DPReview, are practically worthless. Interestingly, they give a resolution of about 80% of Nyquist in most cases, which doesn't tell us much. Raw rendering with "standardized" sharpening is more informative but does not take into account the factors you mention.

That is why Michael's mega pixel shootout is most informative. We are still limited to viewing on screen, and some have criticized his methods, but I think that the tests by those experts are very well done. The differences between the Canon 1Ds M2 and the BetterLight scanning back are not as great as I expected.
Logged

Rob C

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 24074
Leica M8: another perspective
« Reply #8 on: December 10, 2006, 11:54:23 am »

Bernard

The D2x is indeed sharp with the best lenses, but then so too is the D200. I have sent a friend (successful pro living in London) some NEFs from my D200 and he has concluded that there is not a lot to call between my cheaper unit and his D2x. This is not a 'be nice to your pal' relationship - I believe his remarks to be fair and realistic conclusions from studying original files from both cameras.

There is just so much more to photography than mathematics; 'twas ever so, just as much in the heyday of film as in today's world of pixels.

As far as film goes, how can you judge anything called a print unless you use the same enlarger and enlarger lens; with digital how can you compare if using different printers? In the end, if the result pleases the intended client/shooter/viewer that's all that matters.

I would love an M8 but at the same time I know that I didn't own an M Leica (film) because I was not comfortable with the idea of not seeing through the lens, much the same reason why the Rollei TLR had to go when simply keeping the TLR instead of buying the 80mm for the Hasselblad would have allowed me the same focal length at no extra cost. As far as the R Leicas go, they always gave the impression of being over-priced and never quite up to date; perhaps the R6.2 would have been the way to go except for the small matter of cost when compared with Nikon's best...

Ciao - Rob C

BernardLanguillier

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 13983
    • http://www.flickr.com/photos/bernardlanguillier/sets/
Leica M8: another perspective
« Reply #9 on: December 11, 2006, 12:10:31 am »

Quote
Bernard

The D2x is indeed sharp with the best lenses, but then so too is the D200. I have sent a friend (successful pro living in London) some NEFs from my D200 and he has concluded that there is not a lot to call between my cheaper unit and his D2x. This is not a 'be nice to your pal' relationship - I believe his remarks to be fair and realistic conclusions from studying original files from both cameras.

There is just so much more to photography than mathematics; 'twas ever so, just as much in the heyday of film as in today's world of pixels.

[a href=\"index.php?act=findpost&pid=89717\"][{POST_SNAPBACK}][/a]

Hello Rob,

I am sure that the D200 is very good ad well. Even my d80 is not that much worse than the D2x, but I see a clear difference in small details like distant leaves on trees,...

Regards,
Bernard

Ben Rubinstein

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1822
Leica M8: another perspective
« Reply #10 on: December 12, 2006, 07:10:34 am »

Bernard, I think you are missing the point, the FF sensor will give superior results with superior lenses, the sensor is not at fault here, it is the canon WA's which fall flat. It isn't a matter of modern technology either given that the older Contax/leica designs give incredible results on the FF sensors, far surpassing the canon WA's.

I think overall the statement was correct, that canon have crippled it's superior sensor with those lenses does not mean that the format is therefore inferior.

If you are to look for reasons why the sensor may be superior/inferior, you might mention the diffraction limitations of crop sensors, especially that of the D2X, as with everything there will always be a compromise or two but I think that the article was referring to the 'package' as a whole.

I currently own a 20D and a 5D, forgetting the resolution difference, the 5D is far superior with DR, latitude and tonality. The noise difference is like being on another planet and I don't see any noticeable loss of detail as a result.

I know there may be tests showing otherwise, or numerous number and maths games, however I'm shooting a couple of thousand RAW frames a week as a wedding photographer and I see what I see and my conclusions are based on that, not what some article may write. Oh and I'm selling the 20D to make way for the 2nd 5D I just bought...  
Logged

BernardLanguillier

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 13983
    • http://www.flickr.com/photos/bernardlanguillier/sets/
Leica M8: another perspective
« Reply #11 on: December 12, 2006, 08:05:03 am »

Quote
Bernard, I think you are missing the point, the FF sensor will give superior results with superior lenses, the sensor is not at fault here, it is the canon WA's which fall flat. It isn't a matter of modern technology either given that the older Contax/leica designs give incredible results on the FF sensors, far surpassing the canon WA's.

I think overall the statement was correct, that canon have crippled it's superior sensor with those lenses does not mean that the format is therefore inferior.

[a href=\"index.php?act=findpost&pid=90029\"][{POST_SNAPBACK}][/a]

Pom,

I understand what you are saying, but the way the article was structured, the first part started by conclusion on the intrinsic superiority of FF over APS, claiming that the angle at which light is striking the sensor is not an issue.

Even if there are lenses that are better than the Canon wides, I am yet to see an example showing that the image quality in the corner of a 1ds2 with these lenses is superior to what I see on my D2x. Discussing these aspects was suddenly called "pixel peeing" at a certain point of time.

On the other hand, the theoretical limitation in lens resolution don't show yet on the APS sensors. And that is also what I am seeing shooting landscape, even if I am average in the hundereds of images only per week.  I agree that these limitation will start to play at a certain level of resolution, but are we certain that the image quality in the corner of FF sensor will not drop even faster if resolution increases the same way?

I am not denying that FF sensor have some advantage in high iso noise and to some extend DR. But these are not the aspects the article was focussing on.

Don't get me wrong, I have no intention to create an APS sensor fan club and know full well that the Canon 5D and 1ds2 are very good cameras. What I find hard to accept is that the article is claiming to offer a scientific approach to the issue, states some strong conclusions on some controversial topics but doesn't really feed us with the required facts to back up these statements. At least in some places.

Cheers,
Bernard

Nemo

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 276
Leica M8: another perspective
« Reply #12 on: December 12, 2006, 04:09:34 pm »

Dear Bernard,

it is not a question of FF against APS, but a basic principle: the bigger the sensor is, the higher is the resolution/contrast reached on paper.

You are right in one point: if you crop the circle of light of a 35mm lens, using a smaller sensor, the corner resolution/contrast will be higher. However, there exist a counterbalance: the smaller the sensor, the bigger the enlargement needed for a particular print. This affects the whole image, center and corners.

Regards,

Rubén.
« Last Edit: December 12, 2006, 04:12:20 pm by Nemo »
Logged

BernardLanguillier

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 13983
    • http://www.flickr.com/photos/bernardlanguillier/sets/
Leica M8: another perspective
« Reply #13 on: December 12, 2006, 05:19:36 pm »

Quote
Dear Bernard,

it is not a question of FF against APS, but a basic principle: the bigger the sensor is, the higher is the resolution/contrast reached on paper.

You are right in one point: if you crop the circle of light of a 35mm lens, using a smaller sensor, the corner resolution/contrast will be higher. However, there exist a counterbalance: the smaller the sensor, the bigger the enlargement needed for a particular print. This affects the whole image, center and corners.

Regards,

Rubén.
[a href=\"index.php?act=findpost&pid=90110\"][{POST_SNAPBACK}][/a]

Ruben,

I understand your comments and agree with them in theory, but:

- What you write here is different from what was written in the M8 article. My reactions are directed against this article, the way it was structured, and the subliminal messages it is conveying. You might want to re-read it,

- The theory is seriously lacking in experimental back up. The little experimental data we have is, if anything, showing the opposite result at the current point of time.

Cheers,
Bernard

Nemo

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 276
Leica M8: another perspective
« Reply #14 on: December 13, 2006, 12:59:42 pm »

Bernard,

I cannot agree with your last comments. I don't see any "subliminal" messages or contradictions.

My experience backs the predictions of the theory: the bigger the sensor is, the better (in terms of real detail reproduced and contrast).

Bigger sensors have additional advantages, but they also have shortcomings (mostly in terms of versatility).

Best.
« Last Edit: December 13, 2006, 01:07:53 pm by Nemo »
Logged

howiesmith

  • Guest
Leica M8: another perspective
« Reply #15 on: December 13, 2006, 02:11:39 pm »

Quote
Bernard,

... the predictions of the theory: the bigger the sensor is, the better (in terms of real detail reproduced and contrast).

[a href=\"index.php?act=findpost&pid=90290\"][{POST_SNAPBACK}][/a]

Is this similar saying the bigger the piece of film, the better?
Logged

Nemo

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 276
Leica M8: another perspective
« Reply #16 on: December 13, 2006, 02:32:37 pm »

Yes, it is. The same principle holds.
Logged

BernardLanguillier

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 13983
    • http://www.flickr.com/photos/bernardlanguillier/sets/
Leica M8: another perspective
« Reply #17 on: December 14, 2006, 12:28:40 am »

Quote
Yes, it is. The same principle holds.
[a href=\"index.php?act=findpost&pid=90314\"][{POST_SNAPBACK}][/a]

Except that all grains are the same size, whether they belong to a sheet of 4*5 Provia, or to a 24*36 roll of the same film.

Pixel size does obviously differ.

The only question really is "are the lenses able to pass enough information for the sensor photosites for these sites to be able to collect enough quality image information". This should be done at equal pixel count.

The theory says that so-called FF sensor should do this better than APS sensors when resolution reaches a certain level. I can relate to that.

You are saying that this limit has been reached with current generation DSLR, I am saying that I haven't seen any experimental results showing this. I am not even factoring DoF/diffraction in the equation, but this would further level the playing field.

Science is based on the usage of experiment to back up theory. I see the theory, but I don't see the experimental results backing it up.

Even in this Canon dominated world we live in (  ), nobody has been able to come up with a clear demonstration that - at low ISO - a 1ds2 file looks better on screen - or prints better than a D2x file.

I understand that some people might not be interested in this discussion, but if you engage in claiming that one system is better than the other, you owe to intellectual fairness to check both. I did and couldn't see an advantage to FF in terms of actual resolution.

Again, I am not denying the value of FF in terms of high ISO noise and DR and I fully understand that some people see these aspects to be important enough to select FF. I will probably do the same when Nikon releases a FF body. What I cannot accept is the selection of FF on the un-proven assumption that it delivers sharper images.

Cheers,
Bernard

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Leica M8: another perspective
« Reply #18 on: December 14, 2006, 01:37:25 am »

Some more theory...

It would be possible to design lenses specially for APS-C size which are optimized for larger apertures than FF lenses. If you can work with 1:4 instead of 1:5.6 the gain is
one step in ISO, like 400 ISO at 1:4 vs. 800 ISO at 1:5.6.

There seems to be some tendency into this direction, there are some reasonably cheap 17-50/2.8 (or so) lenses from Tamron etc. having very good performance (at least in tests).

Depth of field would be about same at 1:4 on APS-C as 1:5.6 on FF.


Best regards
Erik

Quote
Except that all grains are the same size, whether they belong to a sheet of 4*5 Provia, or to a 24*36 roll of the same film.

Pixel size does obviously differ.

The only question really is "are the lenses able to pass enough information for the sensor photosites for these sites to be able to collect enough quality image information". This should be done at equal pixel count.

The theory says that so-called FF sensor should do this better than APS sensors when resolution reaches a certain level. I can relate to that.

You are saying that this limit has been reached with current generation DSLR, I am saying that I haven't seen any experimental results showing this. I am not even factoring DoF/diffraction in the equation, but this would further level the playing field.

Science is based on the usage of experiment to back up theory. I see the theory, but I don't see the experimental results backing it up.

Even in this Canon dominated world we live in (  ), nobody has been able to come up with a clear demonstration that - at low ISO - a 1ds2 file looks better on screen - or prints better than a D2x file.

I understand that some people might not be interested in this discussion, but if you engage in claiming that one system is better than the other, you owe to intellectual fairness to check both. I did and couldn't see an advantage to FF in terms of actual resolution.

Again, I am not denying the value of FF in terms of high ISO noise and DR and I fully understand that some people see these aspects to be important enough to select FF. I will probably do the same when Nikon releases a FF body. What I cannot accept is the selection of FF on the un-proven assumption that it delivers sharper images.

Cheers,
Bernard
[a href=\"index.php?act=findpost&pid=90442\"][{POST_SNAPBACK}][/a]
Logged
Erik Kaffehr
 

BernardLanguillier

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 13983
    • http://www.flickr.com/photos/bernardlanguillier/sets/
Leica M8: another perspective
« Reply #19 on: December 14, 2006, 06:05:10 am »

Some experiments now...

http://forums.dpreview.com/forums/read.asp...essage=21242713

It would appear that those claiming that FF sensors have no problem with image uniformity beyond those introduced by lenses might be overlooking some phenomenon.  

Cheers,
Bernard
Pages: [1] 2   Go Up