Analysis of RAW samples at Fred Miranda show weak DR

[Mikehit]

6D II taken images will appear slightly sharper than 5D III or 6D if printed at the same size or viewed on screen at the same resolution.

Point taken.
My question could be rephrased as what is the real-world relevance of pixel sharpness if the image we produce is actually sharper?

Which comes back to my question: what is the relevance of pixel sharpness? What does it tell us about the image that comes out of the 6D2? How would analysis of pixel sharpness affect my purchasing decision?

 

[jeffa4444]

6D II taken images will appear slightly sharper than 5D III or 6D if printed at the same size or viewed on screen at the same resolution.

Point taken.
My question could be rephrased as what is the real-world relevance of pixel sharpness if the image we produce is actually sharper?

Which comes back to my question: what is the relevance of pixel sharpness? What does it tell us about the image that comes out of the 6D2? How would analysis of pixel sharpness affect my purchasing decision?

It doesn't what should effect your purchasing decision is to what size you want to enlarge your shots to, what resolution is my monitor or what are my normal viewing levels.

 

[Deleted member 7828]

Interesting...although it is kind of hard to tell on my phone, looks like the trade off for the lower DR on the 6D2 is a slightly crisper (default) RAW file which is more in line with the previous 5D3 and 6D...the 5D4 and the other recent Canon cameras have a softer output. That being said, you may still be able to squeeze out a tiny bit more detail on the 5D4 files with some USM tweaking. And of course, for those raising shadows you may get more detail with the 5D4 as well in the final image.

 

[SecureGSM]

it feels like at my first Job Interview back in 1989

if you were to observe images taken with the same lens, in same lighting conditions, same framing, same distance to subject, same F-stop, same shutter speed, same ISO on the same size screen and at the same resolution ... theeew 6D II taken images will look sharper than images taken with 6D or 5D III. Same is valid for your large prints.
in similar fashion 5DSR images, due to resolution benifit, do look sharper on screen than images taken with 6D, 6D II, 5D III.

here comes purchasing decision advise: if you ar after sharper on-screen images, then 6D II camera seems like a stronger candidate than 6D original.
in another words: 6D II cropping power is stronger than the one of 6D original. you can crop with 6d II by 14% more than with 6D original

6D II taken images will appear slightly sharper than 5D III or 6D if printed at the same size or viewed on screen at the same resolution.

Point taken.
My question could be rephrased as what is the real-world relevance of pixel sharpness if the image we produce is actually sharper?

Which comes back to my question: what is the relevance of pixel sharpness? What does it tell us about the image that comes out of the 6D2? How would analysis of pixel sharpness affect my purchasing decision?

 

[SecureGSM]

that was my intial impression and the motivation behind running these tests. it is an illusion though. at pixel level all these cameras produced virtually identical sharpness results.

Interesting...although it is kind of hard to tell on my phone, looks like the trade off for the lower DR on the 6D2 is a slightly crisper (default) RAW file which is more in line with the previous 5D3 and 6D...the 5D4 and the other recent Canon cameras have a softer output. That being said, you may still be able to squeeze out a tiny bit more detail on the 5D4 files with some USM tweaking. And of course, for those raising shadows you may get more detail with the 5D4 as well in the final image.

 

[Mikehit]

it feels like at my first Job Interview back in 1989

if you were to observe images taken with the same lens, in same lighting conditions, same framing, same distacne to subject, same F-stop, same shutter speed, same ISO on the same size screen and and the same resolution ... theeew 6D II taken images will look sharper than images taken with 6D or 5D III. Same is valid for your large prints.
in similar fashion 5DSR images, due to resolution benifit, do look sharper on screen than images taken with 6D, 6D II, 5D III.

here comes purchasing decision advise: if you ar after sharper on-screen images, then 6D II camera seems like a stronger candidate than 6D original.
in another words: 6D II cropping power is stronger than the one of 6D original. you can crop with 6d II by 14% more than with 6D original

I get all that. I really do. Honest.
What you have not explained is what is the relevance of 'pixel sharpness'. You have gone to the trouble of assessing it and even gone to the trouble of stating "there is NO sharpness advantage for 6D II at the pixel level, ". Yet I don't understand why...what does it tell us.

 

[SecureGSM]

Impression I was getting from looking at 6D II produced images yesterday, was that they are sharper than the one produced by the 6D original. I promissed Lud34 to run a quick sharpness evaluation test to see if Canon decided to trade lower DR in the sensor in exchange for a better resolving power of 6D II sensor. It turns out that I was mistaken and unfortunatelly there is no sharpness advantage in files produced by 6D II at pixel level.
I cannot provide any better explanation but this one:
Canon improved overal image resolution for 6d II by providing more pixels, performing identically well with pixels of 6D original. you get sharper looking image. if that is not enough, then I do not know how to explain better. ignore me

it feels like at my first Job Interview back in 1989

if you were to observe images taken with the same lens, in same lighting conditions, same framing, same distacne to subject, same F-stop, same shutter speed, same ISO on the same size screen and and the same resolution ... theeew 6D II taken images will look sharper than images taken with 6D or 5D III. Same is valid for your large prints.
in similar fashion 5DSR images, due to resolution benifit, do look sharper on screen than images taken with 6D, 6D II, 5D III.

here comes purchasing decision advise: if you ar after sharper on-screen images, then 6D II camera seems like a stronger candidate than 6D original.
in another words: 6D II cropping power is stronger than the one of 6D original. you can crop with 6d II by 14% more than with 6D original

I get all that. I really do. Honest.
What you have not explained is what is the relevance of 'pixel sharpness'. You have gone to the trouble of assessing it and even gone to the trouble of stating "there is NO sharpness advantage for 6D II at the pixel level, ". Yet I don't understand why...what does it tell us.

 

[Mikehit]

Impression I was getting from looking at 6D II produced images yesterday, was that they are sharper than the one produced by the 6D original. I promissed Lud34 to run a quick sharpness evaluation test to see if Canon decided to trade lower DR in the sensor in exchange for a better resolving power of 6D II sensor. It turns out that I was mistaken and unfortunatelly there is no sharpness advantage in files produced by 6D II at pixel level.
I cannot provide any better explanation but this one:
Canon improved overal image resolution for 6d II by providing more pixels, performing identically well with pixels of 6D original. you get sharper looking image. if that is not enough, then I do not know how to explain better. ignore me

it feels like at my first Job Interview back in 1989

if you were to observe images taken with the same lens, in same lighting conditions, same framing, same distacne to subject, same F-stop, same shutter speed, same ISO on the same size screen and and the same resolution ... theeew 6D II taken images will look sharper than images taken with 6D or 5D III. Same is valid for your large prints.
in similar fashion 5DSR images, due to resolution benifit, do look sharper on screen than images taken with 6D, 6D II, 5D III.

here comes purchasing decision advise: if you ar after sharper on-screen images, then 6D II camera seems like a stronger candidate than 6D original.
in another words: 6D II cropping power is stronger than the one of 6D original. you can crop with 6d II by 14% more than with 6D original

I get all that. I really do. Honest.
What you have not explained is what is the relevance of 'pixel sharpness'. You have gone to the trouble of assessing it and even gone to the trouble of stating "there is NO sharpness advantage for 6D II at the pixel level, ". Yet I don't understand why...what does it tell us.

Thank you. I do find this interesting - there are so many design decisions that these discussions do help but I am one of those that feels the need for context rather than just collection new information.

 

[Deleted member 7828]

that was my intial impression and the motivation behind running these tests. it is an illusion though. at pixel level all these cameras produced virtually identical sharpness results.

Interesting...although it is kind of hard to tell on my phone, looks like the trade off for the lower DR on the 6D2 is a slightly crisper (default) RAW file which is more in line with the previous 5D3 and 6D...the 5D4 and the other recent Canon cameras have a softer output. That being said, you may still be able to squeeze out a tiny bit more detail on the 5D4 files with some USM tweaking. And of course, for those raising shadows you may get more detail with the 5D4 as well in the final image.

Interesting, although I was comparing it more to the 5D4, which I see you did not include in your test. I'd be interested to see how the 6D2 and 5D4 compare...

 

[SecureGSM]

5D IV test results, as you requested:

Canon 5D IV

QoF=1850.3 << approx. 3% better than 6DII, 5D III, or 2% better than 6D Original.

Camera Model Canon EOS 5D Mark IV
Firmware Version Firmware Version 1.0.1
Serial Number XXXXXXXXX
Lens EF85mm f/1.8 USM
Focal Length 85.0mm
Test Aperture f/5.6
Test ISO 100
Distance to Target 7.4m
Shutter Speed 1/5s
EV 7.2
Colour Temperature Unknown
Camera Temperature 36C
Spectral Power (R/G/B) 36/30/34
Red Quality 1807.8
Green Quality 1889.4
Blue Quality 1859.9
HVR -2.9%

that was my intial impression and the motivation behind running these tests. it is an illusion though. at pixel level all these cameras produced virtually identical sharpness results.

Interesting...although it is kind of hard to tell on my phone, looks like the trade off for the lower DR on the 6D2 is a slightly crisper (default) RAW file which is more in line with the previous 5D3 and 6D...the 5D4 and the other recent Canon cameras have a softer output. That being said, you may still be able to squeeze out a tiny bit more detail on the 5D4 files with some USM tweaking. And of course, for those raising shadows you may get more detail with the 5D4 as well in the final image.

Interesting, although I was comparing it more to the 5D4, which I see you did not include in your test. I'd be interested to see how the 6D2 and 5D4 compare...

 

[Deleted member 7828]

Hmm, now THAT is interesting. Thanks for sharing.

 

[SecureGSM]

could be just due to the image to image variation as well.. all it proves that there is no substantial pixel level sharpness variation with these 4 cameras. Ideally I should have ran the test over multiple images for each camera model, however it was never meant to be a comprehensive test anyway.

p.s. just to put this into context: I remember following a heated forum discussion re 5D IV image sharpness not as good as the one of 5D III due to stronger AA filter in 5D IV. Looking at the test results, It turns out to be false positive.

Hmm, now THAT is interesting. Thanks for sharing.

 

[chrysoberyl]

could be just due to the image to image variation as well.. all it proves that there is no substantial pixel level sharpness variation with these 4 cameras. Ideally I should have ran the test over multiple images for each camera model, however it was never meant to be a comprehensive test anyway.

p.s. just to put this into context: I remember following a heated forum discussion re 5D IV image sharpness not as good as the one of 5D III due to stronger AA filter in 5D IV. Looking at the test results, It turns out to be false positive.

Hmm, now THAT is interesting. Thanks for sharing.

I truly appreciate all the work and time you have put into these discussions.

Now please compare the A9 at pixel level to...NO - I'm kidding!

Thanks again!

 

[SecureGSM]

No worries. I am thinking about runing the same test with 5DsR image. Will let you know tomorrow

I truly appreciate all the work and time you have put into these discussions.

Now please compare the A9 at pixel level to...NO - I'm kidding!

Thanks again!

 

[sean0306]

It doesn't what should effect your purchasing decision is to what size you want to enlarge your shots to, what resolution is my monitor or what are my normal viewing levels.

Exactly, as there are other factors to consider. While the 6D II may offer better resolution at most F-stops, the greater pixel density also means diffraction should start to creep in about a stop or so sooner than the original 6D. If you do a lot of macro or landscape work at high f-stops, you may start to lose the benefits of having a higher resolution sensor.

 

[Mikehit]

Exactly, as there are other factors to consider. While the 6D II may offer better resolution at most F-stops, the greater pixel density also means diffraction should start to creep in about a stop or so sooner than the original 6D. If you do a lot of macro or landscape work at high f-stops, you may start to lose the benefits of having a higher resolution sensor.

Rubbish. A higher MP sensor will always give greater resolution at a given aperture.

 

[sean0306]

Exactly, as there are other factors to consider. While the 6D II may offer better resolution at most F-stops, the greater pixel density also means diffraction should start to creep in about a stop or so sooner than the original 6D. If you do a lot of macro or landscape work at high f-stops, you may start to lose the benefits of having a higher resolution sensor.

Rubbish. A higher MP sensor will always give greater resolution at a given aperture.

Resolution, yes. But sharpness will affected sooner at higher pixel density. Again, it depends on what you are trying to do, and what your intended output is. The vast majority of folks won't be affected by it.

 

[LonelyBoy]

Exactly, as there are other factors to consider. While the 6D II may offer better resolution at most F-stops, the greater pixel density also means diffraction should start to creep in about a stop or so sooner than the original 6D. If you do a lot of macro or landscape work at high f-stops, you may start to lose the benefits of having a higher resolution sensor.

Rubbish. A higher MP sensor will always give greater resolution at a given aperture.

Resolution, yes. But sharpness will affected sooner at higher pixel density. Again, it depends on what you are trying to do, and what your intended output is. The vast majority of folks won't be affected by it.

I've never understood the point of pixel-level sharpness over total resolution.

 

[Mikehit]

Exactly, as there are other factors to consider. While the 6D II may offer better resolution at most F-stops, the greater pixel density also means diffraction should start to creep in about a stop or so sooner than the original 6D. If you do a lot of macro or landscape work at high f-stops, you may start to lose the benefits of having a higher resolution sensor.

Rubbish. A higher MP sensor will always give greater resolution at a given aperture.

Resolution, yes. But sharpness will affected sooner at higher pixel density. Again, it depends on what you are trying to do, and what your intended output is. The vast majority of folks won't be affected by it.

At any given output size, sharpness will not be affected sooner because diffraction is a function of the aperture not of the sensor.
The logic of what you are saying is that landscapers will prefer to use a 12MP FF sensor over a 36 MP sensor because diffraction effects will make the 36MP less sharp.

Please show me an example of two shots with the only difference being sensor resolution where the higher resolution is less sharp.

 

[Luds34]

Luds34,

I have ran quick sharpness (QoF) Focal test analysis over individual RAW files I have downloaded on dpreview website. Test was conducted at the pixel level (1:1). here are the results (updated with 5D IV data on request):

Canon 6D II: 1808.4
Canon 5D III: 1806.6
Canon 6D: 1829.7 <<<<< better by 1%
Canon 5D IV: 1850.3 <<<<< approx. 3% better than 6DII, 5D III, or 2% better than 6D Original.

there is NO sharpness advantage for 6D II at the pixel level, images appear sharper due to 6D II files are 30% larger than 6D files. (down sampling improves perceived screen image sharpness).

A little late to the party, but thanks for posting your update.

So in short there is no pixel level difference. However the greater resolution of the sensor as whole leads to better sharpness.

Of course the pixel level sharpness being equal at ISO 100 I suppose isn't all that surprising. It's why crop sensors, even phones look good in great light.

Thanks again.