Pages: [1] 2   Go Down

Author Topic: Is the "MP race" a good thing?  (Read 9848 times)

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Is the "MP race" a good thing?
« on: August 21, 2016, 04:37:10 am »

Hi,

The term megapixel race is often used. But are small pixels good or bad?

What I would state is small pixels are mostly advantageous as long as they are within optimum of the presently used pixel designs.

It could be argued that we don't need a lot of megapixels, and in many cases it is quite true. Full screen HD is just two megapixels. So, we don't need high MP count cameras for web size images. At least not until 8K TV is here…

On the other hand, pixels scale nicely down. So there is very little disadvantage to high MP images.

Let's look at some examples, the first one is one of my own. I made two similar shots with a Hasselblad Planar 100/3.5, using the same camera position (or so I believe). The first image is with the P45+ back and it produces around 39 MP. The second one was a shift stitch using the same lens with the Sony A7rII. The Sony A7rII is a 42 MP camera, but the stitched image was around 82 MP and the surface are was about the same as on the P45+. The link below goes to the full size image, the downscaling of the image in the forum software creates some artefacts of it's own.
http://echophoto.dnsalias.net/ekr/Articles/MPRace/Planar_100.jpg
This a crop from the original screendump, just showing the most interesting part of the original screen dump


This the full image, but downscaled by forum software:


The images are all actual pixels, the first one is the P45+, the second one is the stitched image from the A7rII, downscaled to 39 MP. The third one is the original stitched image.

Check out the sign "Centralstationen" on the P45+ it is almost unreadable. The back resolves the text, but demosaic really struggles. Colour aliasing error are quite abundant, with the moiré on the blinder perhaps being most typical. The image at the center is downsampled to 39 MP. So, resolution is exactly the same, but the downsampled image is much cleaner. The image on the right shows the full resolution of the 80 MP image and it has no aliasing. Why? I don't know. Some possible explanations:

  • Sensor outresolves the lens, that is the lens can deliver very little contrast at the pixel level.
  • Sensor outresolves the subject, that is the subject lacks fine detail that would not be resolved by the sensor
  • Microlenses on the sensor reduce aliasing as sampling is done over a larger area (area sampling vs, point sampling

So a high resolution sensor will deliver a cleaner image than a lower resolution sensor of similar size.

The next point is that it is often said that small pixels are noisy. Let's compare some high ISO low MP cameras and some of the high MP champs. In the four corners: The Sony A7sII is a 12 MP camera mostly intended for motion and high ISO shooting, we can pair it with the Sony A7rII, the resolution champ within the A7 family.

The Canon 1DXII is Canon's low light offering and the 5DsR is the resolution king from Canon.

We can compare the four using a very nifty feature from DPReview that offers different kinds of comparisons. Let's compare the four at 12800ISO and actual pixels:
http://echophoto.dnsalias.net/ekr/Articles/MPRace/12800ISO_actual_pixels.jpg


The Sony A7sII is arguably the cleanest while the 5DsR may be the most noisy. So, high resolution goes hand in hand with high noise at actual pixels.

The same tool at DPReview can compare images at normalised size, so all images are sized to same resolution:
http://echophoto.dnsalias.net/ekr/Articles/MPRace/12800ISO_normalised.jpg


Voilá, all images are pretty much the same! So, once we compare the images at the same size the high MP images will be comparable to the low MP images.

Now, we switch to 100 ISO and compare gray text on black background at actual pixels.
http://echophoto.dnsalias.net/ekr/Articles/MPRace/ActualPixels.jpg


Here the Sony A7sII fully falls apart, it has far to weak AA-filtering to handle pixel level colour aliasing. The higher resolution of Canon 1DXII paired with proper AA-filter does a much better job. The Canon 5DsR delivers the cleanest image.

I would probably argue that even the Sony A7sII can make very good A2 (16"x23") prints as I have been shooting a lot with 12MP cameras and made a log of good A2-size prints. The 39-50 MP cameras tested here allow for really large print sizes.

A clear disadvantage with high MP cameras is the increase in image size. But, raw images are often quite effectively compressed. The 39 MP images from the P45+ are around 50 MP in size and so are the Sony A7rII images using DNG-compression. On the other hand, full size TIFFs will be large.

Best regards
Erik
« Last Edit: August 21, 2016, 05:08:55 am by ErikKaffehr »
Logged
Erik Kaffehr
 

Rhossydd

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3369
    • http://www.paulholman.com
Re: Is the "MP race" a good thing?
« Reply #1 on: August 21, 2016, 05:27:09 am »

On the other hand, pixels scale nicely down. So there is very little disadvantage to high MP images.
You may be correct in just the context of image quality, but there's also other more practical issues. More pixels than you really need make image processing slow (eg http://forum.luminous-landscape.com/index.php?topic=111999.0 ), cost more to store and archive, slow down reviewing in camera, require large storage cards when shooting etc.
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: Is the "MP race" a good thing?
« Reply #2 on: August 21, 2016, 07:36:01 am »

Hi,

That is a very valid point.

On the other hand MP-count increases slowly, on the Sony side
YearModelMP
2008Alpha 90024
2011A7r36
2013A7rII42

When I startted working with digital back in 2004 I think my hard drives were 250 MB, the ones I have now are 6TB.

My first camera was 6MP and the latest is 42 MP, so MP has gone up sevenfold and disk size 24-fold.

The thread you are referring to discussed performance with 4K and 5K screens. With higher screen resolutions the computers have quite a bit more processing to do.

Best regards
Erik


You may be correct in just the context of image quality, but there's also other more practical issues. More pixels than you really need make image processing slow (eg http://forum.luminous-landscape.com/index.php?topic=111999.0 ), cost more to store and archive, slow down reviewing in camera, require large storage cards when shooting etc.
Logged
Erik Kaffehr
 

Rhossydd

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3369
    • http://www.paulholman.com
Re: Is the "MP race" a good thing?
« Reply #3 on: August 21, 2016, 07:55:22 am »

The thread you are referring to discussed performance with 4K and 5K screens.
Not really.
To quote "When I process lower res files up to say 24mp all is fine but by the time you hit 40-50mp files the performance takes a massive step down.  Too much in my opinion.  To me it feels like the whole architecture of the software just can't handle these high res files."
Quote
My first camera was 6MP
A late arrival ;-) My first was 2.1mp then > 3 > 6 > 8 >16 > 21
Looking at the prints on my walls, most about A3, I need to look extremely closely to see any failings in the 8mp (EOS1Dii) images, so everything after that has really just allowed more selective enlargement.
Logged

Bart_van_der_Wolf

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 8913
Re: Is the "MP race" a good thing?
« Reply #4 on: August 21, 2016, 08:56:41 am »

Not really.
To quote "When I process lower res files up to say 24mp all is fine but by the time you hit 40-50mp files the performance takes a massive step down.  Too much in my opinion.  To me it feels like the whole architecture of the software just can't handle these high res files."

Hi,

But the point Erik is making is that the slow down is largely caused by the time required for generating the preview. As some users have reported, pre-building large previews speeds things up, and by reducing the size of the application window (and thus the preview area), can also help to speed things up.

Quote
Looking at the prints on my walls, most about A3, I need to look extremely closely to see any failings in the 8mp (EOS1Dii) images, so everything after that has really just allowed more selective enlargement.

That may also have to do with the subject matter. When 'details' are recognizable enough, an image could look 'good enough'. But when it comes to meeting the visual acuity limits, for a given viewing distance, we often do need more pixels for larger output sizes.

I've published the following rules of thumb before, but it might help to repeat them:
Quote
Rule of thumb for the required PPI for viewing (large format) output at a given distance:
• 1 arc minute at 1 metre equals 87.32 PPI.
• 0.4 arc minutes at 1 metre equals 218.3 PPI.
Divide by the viewing distance in metres to find the required minimum PPI.

For the metrically challenged this would become:
• 1 arc minute at 1 foot equals 286.48 PPI.
• 0.4 arc minutes at 1 foot equals 716.2 PPI.
Divide by the viewing distance in feet to find the required minimum PPI.

This is before upsampling to the printer driver's native resolution (to avoid low quality printer driver interpolation, and to allow output sharpening at the printer's native output resolution).

So for an A3 (420 x 297 mm) output size, and viewing it at 12 inches reading distance, we need 4737 x 3350 pixels (=15.9 MP after cropping for aspect ratio) to match average visual acuity.

The choice between 1 arc minute or 0.4 arc minute, as angular resolution is dictated by the visual acuity of the viewer. On average, for 20/20 vision, 1 arc minute is used, but many people have better than average acuity (especially at young age). For the rest, distance makes the difference. If an image is never viewed up close (at say reading distance of 30.48 cm or 12 inches), then we can settle for lower resolution originals.

What's often underestimated, is that with proper software (e.g. Topaz 'Detail') we can boost the visibility of the finest detail that our eyes can see (with low contrast), and that will add a lot of 'realism' to surface structure. We will be able to more clearly see what that stuff is made of; e.g. is it cardboard or is it leather?

In many art objects, surface structure and material choice are part of the expression of the creative intent, and I think photography is no exception. Of course, like in painting, there are impressionistic styles and there are realistic styles, but without enough pixels, realistic becomes hard if at all possible.

Cheers,
Bart
« Last Edit: August 22, 2016, 09:13:06 am by BartvanderWolf »
Logged
== If you do what you did, you'll get what you got. ==

Rhossydd

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3369
    • http://www.paulholman.com
Re: Is the "MP race" a good thing?
« Reply #5 on: August 21, 2016, 11:07:42 am »

But the point Erik is making is that the slow down is largely caused by the time required for generating the preview.
The technicalities don't really matter. The OP in that thread is quite clear in his opening post that small files work OK, big ones don't. It's pretty clear.
Quote
So for an A3 (420 x 297 mm) output size, and viewing it at 12 inches reading distance, we need 4737 x 3350 pixels (=15.9 MP after cropping for aspect ratio) to match average visual acuity.
Proves the point perfectly. You don't need more than 16mp for a 'perfect' A3, you'd generally view anything bigger from further away too. Not that many homes have space for anything much bigger than framed A3, even galleries don't often show prints much bigger than that IME.

Logged

Bart_van_der_Wolf

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 8913
Re: Is the "MP race" a good thing?
« Reply #6 on: August 21, 2016, 11:45:59 am »

Proves the point perfectly. You don't need more than 16mp for a 'perfect' A3, you'd generally view anything bigger from further away too. Not that many homes have space for anything much bigger than framed A3, even galleries don't often show prints much bigger than that IME.

That limitation to A3 may depend on local situations. I'd imagine an average home in Japan having much less wall space to hang things on than the wall space in some other countries. If one, like Erik, aims for A2 sized output (twice as large as A3), then 2x as many pixels (32 MP) are required. At larger sizes I think that indeed viewing distance becomes the mediator, because one needs more distance to take in the composition as a whole, and (except photographers) one rarely looks at the details from reading distance (or if hung behind a couch or some other 'obstruction').

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

Tim Lookingbill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2436
Re: Is the "MP race" a good thing?
« Reply #7 on: August 21, 2016, 03:54:12 pm »

I'm not seeing a significant increase in image quality with the increase of pixels in Erik's OP images where it shows moire in the window blinds which will not benefit from more pixels on a scene by scene basis, a real world shooting situation.

Moire is created by lens to subject distance whose straight line detail just happened to be captured at a Lpi frequency and angle that's near the sensor's pixel cell site grid demosaicing couldn't fix. I've worked with halftone dot 4 color seps for commercial printing and I've seen the same pattern when each color sep halftone grid is NOT angled in relation to each other in order to create a rosette pattern which is a compromise to create a uniform look. Halftone dot size and thus Lpi or dots per inch didn't fix the moire, only angling them mathematically did.
Logged

hjulenissen

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2051
Re: Is the "MP race" a good thing?
« Reply #8 on: August 22, 2016, 01:31:37 am »

...Not that many homes have space for anything much bigger than framed A3, even galleries don't often show prints much bigger than that IME.
I have numerous A2 prints in my home, in addition to a ~1m x 17" * 3 triptych.

-h
Logged

BobShaw

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2218
    • Aspiration Images
Re: Is the "MP race" a good thing?
« Reply #9 on: August 22, 2016, 02:51:55 am »

I have numerous A2 prints in my home, in addition to a ~1m x 17" * 3 triptych.
I agree. I have many 800mm wide prints in my home and run a gallery with most images that size. The reality is that if you sell A3 then you are competing with camera clubs. Large high quality prints make people say wow.
Logged
Website - http://AspirationImages.com
Studio and Commercial Photography

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: Is the "MP race" a good thing?
« Reply #10 on: August 22, 2016, 03:26:11 am »

Hi,

Check out the text "Centralstationen" below, on the P45+ it is essentially unreadable, while readibility is excellent on the downsampled high res image. Both crops have the same resolution. The difference is that the downscaled high res image has proper information and the original image has a lot of fake information, interfering with the good info.


Best regards
Erik


I'm not seeing a significant increase in image quality with the increase of pixels in Erik's OP images where it shows moire in the window blinds which will not benefit from more pixels on a scene by scene basis, a real world shooting situation.

Moire is created by lens to subject distance whose straight line detail just happened to be captured at a Lpi frequency and angle that's near the sensor's pixel cell site grid demosaicing couldn't fix. I've worked with halftone dot 4 color seps for commercial printing and I've seen the same pattern when each color sep halftone grid is NOT angled in relation to each other in order to create a rosette pattern which is a compromise to create a uniform look. Halftone dot size and thus Lpi or dots per inch didn't fix the moire, only angling them mathematically did.
Logged
Erik Kaffehr
 

Bart_van_der_Wolf

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 8913
Re: Is the "MP race" a good thing?
« Reply #11 on: August 22, 2016, 06:47:33 am »

I'm not seeing a significant increase in image quality with the increase of pixels in Erik's OP images where it shows moire in the window blinds which will not benefit from more pixels on a scene by scene basis, a real world shooting situation.

Hi Tim,

Significant may mean different things to different people, but I do see the differences in the images as presented. Of course, there are different renderings possible, and different Raw converters produce somewhat different results as well. For instance, I'd expect the Capture One to have a bit less difficulty with the P45+ image than Lightroom.

And I have to stress it once more, more pixels offer more post-processing possibilities. The human eye has a lower contrast sensitivity near the resolution limits of visual acuity. With enough pixels, those resolution details (such as subtle low-contrast surface structure) can be boosted which will make the image 'pop', and become more lifelike in the final print. Light quality also plays a role, and Clarity can do wonders there if there is enough data to use.

Quote
Moire is created by lens to subject distance whose straight line detail just happened to be captured at a Lpi frequency and angle that's near the sensor's pixel cell site grid demosaicing couldn't fix. I've worked with halftone dot 4 color seps for commercial printing and I've seen the same pattern when each color sep halftone grid is NOT angled in relation to each other in order to create a rosette pattern which is a compromise to create a uniform look. Halftone dot size and thus Lpi or dots per inch didn't fix the moire, only angling them mathematically did.

There is a difference between Moiré and the Aliasing that Erik mentions. As you say, the super-positioning of two regular grids with a different pattern frequency or angle, will cause moiré. The aliasing that Erik points out is due to the relatively coarse sampling of detail that is smaller than the sensor can reliably resolve. It doesn't even require repetitive patterns, just super-fine detail will do. Aliased fine detail will mimic as larger/fatter detail, as an alias for the real stuff that it can't resolve (at least not reliably). This will also affect surface structure and sharp edges, and with a Bayer CFA it will produce False Color artifacts (because of the difference in sampling density between green and Red/Blue filtered sensels).

Of course, when not viewing the different camera resolutions side-by-side, it will be hard to see the differences in actual output, but it will already impede our ability to post-process those details. So, given the choice between a better quality and a lower quality output, I'd know which I want to produce, especially if the cost to get there is modest. Just shoot at a higher resolution, use more Megapixels e.g. by stitching if replacing/upgrading the camera is not feasible.

Cheers,
Bart

P.S. Forgot to mention. A higher MP sensor array, with denser sampling, will extract a higher image quality from a lens than a lower density sampling sensor would from the same lens. Resolution and contrast (MTF response) goes up with sensor resolution.
« Last Edit: August 22, 2016, 09:17:59 am by BartvanderWolf »
Logged
== If you do what you did, you'll get what you got. ==

TonyW

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 643
Re: Is the "MP race" a good thing?
« Reply #12 on: August 22, 2016, 10:56:00 am »

Hi Erik,  interesting observations as usual and my opinion to the question posed in the thread title "Is the MP race a good thing?", would be yes with some qualifications and a lot of guessing on my part about what the customer wants (perhaps contrary to what they may actually need)   ;D

Generally I think that competition between manufacturers vying for our custom should lead to product improvements, starting with the high end products and later trickling down to the lower end.  MP race is IMHO just part of this.

Your own testing pitting your P45+ back against your Sony A7rII is interesting and certainly it is obvious that a downscaled stitch from the Sony is more satisfying regarding resolved detail and less/no artefacts.  But also for me raises some additional questions/ what ifs:

* What happens to image IQ comparisons between MF and FF 35mm if:
   a.  You adjust shooting distance to account for crop factor between sensors on your P45
   and
   b.  Then shoot and stitch with your P45 as with the Sony ?

* How do the images stack up single shot to single shot with distance adjusment made to equal image size ? EDITED

* Does a change of raw editor from ACR offer better/lesser IQ than ACR ?
« Last Edit: August 22, 2016, 11:19:28 am by TonyW »
Logged

Tim Lookingbill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2436
Re: Is the "MP race" a good thing?
« Reply #13 on: August 22, 2016, 11:30:57 am »

Hi,

Check out the text "Centralstationen" below, on the P45+ it is essentially unreadable, while readibility is excellent on the downsampled high res image. Both crops have the same resolution. The difference is that the downscaled high res image has proper information and the original image has a lot of fake information, interfering with the good info.

Best regards
Erik

I see the legibility issue with the 'Central Station' street sign but it doesn't prove your premise. You've chosen a fixed static scene for the lens distance to sensor pixel grid frequency which is extremely sensitive to producing moire from fine line patterns close to the same Lpi that is influenced by variations in distance of lens to scene detail in order to support a thin premise that more pixels will solve this ON ALL SCENES PHOTOGRAPHED and that is not the truth.

Or another way to put it is you've chosen a test target to prove something that can't be proven on a consistent scene by scene basis. How does your setup pass the smell test with regard to implementing the scientific method. To me it doesn't pass from my understanding of it.

What happens to the detail when you move the camera farther back or much closer on that fixed static scene for both cameras?
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: Is the "MP race" a good thing?
« Reply #14 on: August 22, 2016, 02:03:13 pm »

Hi,

Thanks for responding.

Thinks are not that easy. Just to say, I don't  really want to compare Sony to P45+, it is more about having the equipment I happen to have. As an example, I cannot move around a lot, as there is a small river between me and my subject. With the Sony I can do a shift stitch, using my HCam Master TS, but I cannot really do that on the Hasselblad.

The idea here is more about demonstrating the advantage of smaller pixel size.

The way I shoot is a very good approximation of what you would see with a 100 MP digital back, as the 100 MO sensors are made by Sony and probably have characteristics very close to the A7rII.

So, the A7r Image actually indicates what you would get out from a 100 MP back shooting a Zeiss Planar 100/3.5.

In general, there will always be a problem when the sensor cannot handle the information from the lens. With a large pixel sensor there will be a lot of information that the sensor will not resolve. So, instead of fine detail, it will deliver coarse artefacts. Image processing experts often think in the frequency domain. A lens resolving beyond "Nyquist" which is sensor resolution, will cause low frequency artefacts. The better the lens the worse artefacts.

Anyways, thanks for feedback, I will see if I can come up with some more elaborate tests. Just to say, these images are borderline between real world shoots and testing. Mostly, I have made some observations in real world shooting and go back to investigate.

Best regards
Erik


Hi Erik,  interesting observations as usual and my opinion to the question posed in the thread title "Is the MP race a good thing?", would be yes with some qualifications and a lot of guessing on my part about what the customer wants (perhaps contrary to what they may actually need)   ;D

Generally I think that competition between manufacturers vying for our custom should lead to product improvements, starting with the high end products and later trickling down to the lower end.  MP race is IMHO just part of this.

Your own testing pitting your P45+ back against your Sony A7rII is interesting and certainly it is obvious that a downscaled stitch from the Sony is more satisfying regarding resolved detail and less/no artefacts.  But also for me raises some additional questions/ what ifs:

* What happens to image IQ comparisons between MF and FF 35mm if:
   a.  You adjust shooting distance to account for crop factor between sensors on your P45
   and
   b.  Then shoot and stitch with your P45 as with the Sony ?

* How do the images stack up single shot to single shot with distance adjusment made to equal image size ? EDITED

* Does a change of raw editor from ACR offer better/lesser IQ than ACR ?
Logged
Erik Kaffehr
 

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: Is the "MP race" a good thing?
« Reply #15 on: August 22, 2016, 02:14:17 pm »

Hi Bart,

My processing skills with C1 are somewhat limited, but I did a quick and dirty conversion in C1 v9. I left most things at defaults, including that "film curve".

I just attach the resulting image, the C1 processing on the right may be better than the one in LR, and could probably improved quite a bit. But, I would say it has the same issues as the LR conversion.

Another way to put it, "there is no substitute for good pixels" and 6.8 micron non OLP filtered pixels are no good pixels on a Zeiss Planar 100/3.5 at f/f11.

Best regards
Erik


Hi Tim,

Significant may mean different things to different people, but I do see the differences in the images as presented. Of course, there are different renderings possible, and different Raw converters produce somewhat different results as well. For instance, I'd expect the Capture One to have a bit less difficulty with the P45+ image than Lightroom.

And I have to stress it once more, more pixels offer more post-processing possibilities. The human eye has a lower contrast sensitivity near the resolution limits of visual acuity. With enough pixels, those resolution details (such as subtle low-contrast surface structure) can be boosted which will make the image 'pop', and become more lifelike in the final print. Light quality also plays a role, and Clarity can do wonders there if there is enough data to use.

There is a difference between Moiré and the Aliasing that Erik mentions. As you say, the super-positioning of two regular grids with a different pattern frequency or angle, will cause moiré. The aliasing that Erik points out is due to the relatively coarse sampling of detail that is smaller than the sensor can reliably resolve. It doesn't even require repetitive patterns, just super-fine detail will do. Aliased fine detail will mimic as larger/fatter detail, as an alias for the real stuff that it can't resolve (at least not reliably). This will also affect surface structure and sharp edges, and with a Bayer CFA it will produce False Color artifacts (because of the difference in sampling density between green and Red/Blue filtered sensels).

Of course, when not viewing the different camera resolutions side-by-side, it will be hard to see the differences in actual output, but it will already impede our ability to post-process those details. So, given the choice between a better quality and a lower quality output, I'd know which I want to produce, especially if the cost to get there is modest. Just shoot at a higher resolution, use more Megapixels e.g. by stitching if replacing/upgrading the camera is not feasible.

Cheers,
Bart

P.S. Forgot to mention. A higher MP sensor array, with denser sampling, will extract a higher image quality from a lens than a lower density sampling sensor would from the same lens. Resolution and contrast (MTF response) goes up with sensor resolution.
Logged
Erik Kaffehr
 

Bart_van_der_Wolf

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 8913
Re: Is the "MP race" a good thing?
« Reply #16 on: August 22, 2016, 02:25:35 pm »

Hi Bart,

My processing skills with C1 are somewhat limited, but I did a quick and dirty conversion in C1 v9. I left most things at defaults, including that "film curve".

I just attach the resulting image, the C1 processing on the right may be better than the one in LR, and could probably improved quite a bit. But, I would say it has the same issues as the LR conversion.

Another way to put it, "there is no substitute for good pixels" and 6.8 micron non OLP filtered pixels are no good pixels on a Zeiss Planar 100/3.5 at f/f11.

Yes, its resolution is a bit less of a mess, but indeed, nothing beats more high quality pixels for a cleaner result.

In fact, such fat pixels could use an AA-filter, but they are usually unavailable in such large sizes (and counterproductive for large amounts of lens shift). With more and denser sampling sensels, the need for using AA-filters is reduced, or the havoc that an unfiltered sensor wreaks is less obtrusive (due to diffraction and the presence of fewer smaller features).

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

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: Is the "MP race" a good thing? Another way to see it…
« Reply #17 on: August 22, 2016, 02:28:54 pm »

Hi,

Another way to see it, it may make sense to buy a system that makes your lens justice. If you don't plan printing larger than A3 a 4/3 camera from Panasonic or Olympus may be a very good option.

Shooting low light, a larger sensor may make some sense. Shooting full aperture with a full frame body will probably mask weaknesses of the sensor. But, I would still say that a camera like the A7rII will deliver a better print than a Sony A7sII in almost any case.

Yes, the large pixel camera will always look better than a small pixel camera on screen, but that depends on different magnification.

Best regards
Erik
Logged
Erik Kaffehr
 

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: Is the "MP race" a good thing?
« Reply #18 on: August 22, 2016, 03:02:38 pm »

Hi Bart,

Could you elaborate on that one?

Just to say, this seems to be a interesting thread with lots of good inputs. :-)

Another point:

Let's say that I want a world class lens, like the Otus 85/1.4. For me it would make some sense to have a world class sensor like the Canon 5DsR or the Sony A7rII, rather than a sensor of more limited resolution like the A7II.

With the A7rII is still see some colour aliasing artefacts even with middle of the road lenses. So it is in no way immune to aliasing. With P45+ it is more of a generic problem. One of the worst issues I have seen are rippled water surfaces. But, I would say that issues with the A7rII are almost a magnitude less than with the P45+.

My guess is that fill factor plays a very major role, so modern sensors with large fill factors are less prone to aliasing than older sensors with a smaller fill factor.


Best regards
Erik





In fact, such fat pixels could use an AA-filter, but they are usually unavailable in such large sizes (and counterproductive for large amounts of lens shift).
Logged
Erik Kaffehr
 

Tim Lookingbill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2436
Re: Is the "MP race" a good thing?
« Reply #19 on: August 22, 2016, 04:21:35 pm »

I'll prove my case with my Sony sensor 2006 6MP Pentax K100D and 18-55mm $80 kit lens with screenshot courthouse below. Both shot Raw taken at slightly different angles, distances and focal length. The one on the left was taken at 50mm, the one on the right 55mm. This is real world shooting situation involving looking and adjusting to fit subject in frame as any photographer would employ. Both were processed and sharpened in CS5/ACR 6.7.
Logged
Pages: [1] 2   Go Up