Hands on Field Test of the Canon EOS 6D Mark II

[dak723]

Blue-sky was a prime area where Canon files did not look as good because of noise.

What planet are you on ? Here's a typical blue sky shot on the 5DII at 100 ISO with zero noise reduction. 100% crop, the whole image is shown for reference. Where is the noise ? It's as clean as anything else from Nikon or Sony - certainly cleaner than MFT.

I just don't get this "colour noise" and "FPN" crap that you throw at the Canons, especially the 5DII.

All the "crap" that Aglet throws at Canon is just that - crap.

He's a TROLL. IGNORE HIM. He's pretty much hijacked every 6D II thread, so please everyone IGNORE HIM.

 

[Aglet]

Blue-sky was a prime area where Canon files did not look as good because of noise.

What planet are you on ? Here's a typical blue sky shot on the 5DII at 100 ISO with zero noise reduction. 100% crop, the whole image is shown for reference. Where is the noise ? It's as clean as anything else from Nikon or Sony - certainly cleaner than MFT.

I just don't get this "colour noise" and "FPN" crap that you throw at the Canons, especially the 5DII.

My 5d2 was one of the early ones, early adopter here, back when. You know, swayed by the fanboy enthusiasm when it first arrived.

The noise mine produced was more obvious to me than your example but pattern noise is there in your example too. I can see it even on my everyday machine with my low end Dell display.
... just have to look a bit closer for it.

Here, I've popped it up a bit to make it easier for you to see it.
I just took a crop from the upper left of your crop so no cloud fragments, enlarged it 4x and did a quick de-sat and contrast kick with Preview to make the patterning more obvious.

Can you NOW see all the stripey vertical noise and pixel non-uniformity?
I don't want that kind of junk in my i mages. I find it visible on-screen and in moderate (12-18") size prints.
Others may not notice it, most don't, but I see it and I'm the one who needs to be satisfied with my images.

So no, it's not as clean as from ABC, sorry. Even MFT with it's extra shot noise has less or no pattern issues.
Those old Canons were rife with pattern noise. Whether from multi-channel read noise issues or pixel non-uniformity as I've learned from Bill Claff's articles on his website. Net result is the same.
My 5d2 had channel read noise issues AND significant non-uniformity issues as a picket fence right across the middle third of the sensor. Yup, sure would have returned it if I'd have figured it out in time.
My 5d2 - most disappointing camera I ever owned.
I shouldn't have listened to the raving fanboys back then.

 

[Aglet]

Can we get back on topic now.

Move your Olympus stuff to the Third Party section, sick of seeing it.

I agree, Tom.
care to post a 100% crop of blue sky at 100 iso from your shiny new toy?

 

[Aglet]

What planet are you on ? Here's a typical blue sky shot on the 5DII at 100 ISO with zero noise reduction. 100% crop, the whole image is shown for reference. Where is the noise ? It's as clean as anything else from Nikon or Sony - certainly cleaner than MFT.

I just don't get this "colour noise" and "FPN" crap that you throw at the Canons, especially the 5DII.

There's nothing wrong with blue sky on 6DII, either. It's just a blue sky.

I'd like to see it...
post an unscaled crop of the upper right corner of your example and I can evaluate it for myself.
can't tell anything from such a reduced size example.

 

[Don Haines]

Blue-sky was a prime area where Canon files did not look as good because of noise.

What planet are you on ? Here's a typical blue sky shot on the 5DII at 100 ISO with zero noise reduction. 100% crop, the whole image is shown for reference. Where is the noise ? It's as clean as anything else from Nikon or Sony - certainly cleaner than MFT.

I just don't get this "colour noise" and "FPN" crap that you throw at the Canons, especially the 5DII.

OMG! Look at all the noise! Canon is doomed!!!!!!

BTW, got the 6D at work and the 6D2 at home. I don't care what the biased tests say, the 6D2 is better than the 6D. Period!

My SO has a 5D2 and is amazed at what the 6D2 does at high ISO....

Most of the negative bias seems to be coming from a very few, very suspect sources......

 

[Aglet]

Blue-sky was a prime area where Canon files did not look as good because of noise.

What planet are you on ? Here's a typical blue sky shot on the 5DII at 100 ISO with zero noise reduction. 100% crop, the whole image is shown for reference. Where is the noise ? It's as clean as anything else from Nikon or Sony - certainly cleaner than MFT.

I just don't get this "colour noise" and "FPN" crap that you throw at the Canons, especially the 5DII.

OMG! Look at all the noise! Canon is doomed!!!!!!

BTW, got the 6D at work and the 6D2 at home. I don't care what the biased tests say, the 6D2 is better than the 6D. Period!

My SO has a 5D2 and is amazed at what the 6D2 does at high ISO....

Most of the negative bias seems to be coming from a very few, very suspect sources......

Yes, those cat-in-a-sack shots were actually pretty impressive hi ISO performance.
AF still quick to lock in those conditions?

 

[Talys]

@Anglet -

Top right 750 pixels@100%

Respectfully, I think if you're going to the amount of effort that you did to try to find if something is flawed, you're spending too much time trying to analyze a photo. I mean, I look at the final image on its intended size and medium (print, screen) and ask, "do I like it?". I think that's infinitely more important than whether you can see an unpleasing pattern if you apply desaturation, magnification, and apply contrast.

To put it another way, if the overall image is not pleasing, who cares if 4 millimeters in the corner exhibit a pattern; and if the overall image looks great, who cares if it does?

And I'm not talking about the photographer's skill -- I mean, the same subject taken with two different cameras. If one image is holistically better than the other (ie judged in its entirety), I'm happy to say so. But to try to say that there is a pattern that no human will ever be able discern with their naked eye is unpleasing is not meaningful to me. For example, that sigma 100-400 to Canon 100-400 comparison indicated that the Sigma was actually holistically better or indistinguishable in image quality to the much more expensive Canon counterpart, but several factors make it less suitable (or unsuitable) for certain types of photography like birding. That's a useful comparison.

 

[Don Haines]

Blue-sky was a prime area where Canon files did not look as good because of noise.

What planet are you on ? Here's a typical blue sky shot on the 5DII at 100 ISO with zero noise reduction. 100% crop, the whole image is shown for reference. Where is the noise ? It's as clean as anything else from Nikon or Sony - certainly cleaner than MFT.

I just don't get this "colour noise" and "FPN" crap that you throw at the Canons, especially the 5DII.

OMG! Look at all the noise! Canon is doomed!!!!!!

BTW, got the 6D at work and the 6D2 at home. I don't care what the biased tests say, the 6D2 is better than the 6D. Period!

My SO has a 5D2 and is amazed at what the 6D2 does at high ISO....

Most of the negative bias seems to be coming from a very few, very suspect sources......

Yes, those cat-in-a-sack shots were actually pretty impressive hi ISO performance.
AF still quick to lock in those conditions?

AF was not as fast as normal to lock, but I am amazed that it did lock!

 

[Mikehit]

My 5d2 was one of the early ones, early adopter here, back when. You know, swayed by the fanboy enthusiasm when it first arrived.

The noise mine produced was more obvious to me than your example but pattern noise is there in your example too. I can see it even on my everyday machine with my low end Dell display.
... just have to look a bit closer for it.

Here, I've popped it up a bit to make it easier for you to see it.
I just took a crop from the upper left of your crop so no cloud fragments, enlarged it 4x and did a quick de-sat and contrast kick with Preview to make the patterning more obvious.

Can you NOW see all the stripey vertical noise and pixel non-uniformity?
I don't want that kind of junk in my i mages. I find it visible on-screen and in moderate (12-18") size prints.
Others may not notice it, most don't, but I see it and I'm the one who needs to be satisfied with my images.

So no, it's not as clean as from ABC, sorry. Even MFT with it's extra shot noise has less or no pattern issues.
Those old Canons were rife with pattern noise. Whether from multi-channel read noise issues or pixel non-uniformity as I've learned from Bill Claff's articles on his website. Net result is the same.
My 5d2 had channel read noise issues AND significant non-uniformity issues as a picket fence right across the middle third of the sensor. Yup, sure would have returned it if I'd have figured it out in time.
My 5d2 - most disappointing camera I ever owned.
I shouldn't have listened to the raving fanboys back then.

You are the one making claims of inadequacy of the 5D against MFT. It is therefore beholden on you to show a comparison of the two systems to demonstrate that superiority - and until you do all you are trying to convince us of is blather.

 

[stevelee]

Here, I've popped it up a bit to make it easier for you to see it.
I just took a crop from the upper left of your crop so no cloud fragments, enlarged it 4x and did a quick de-sat and contrast kick with Preview to make the patterning more obvious.

Can you NOW see all the stripey vertical noise and pixel non-uniformity?
I don't want that kind of junk in my i mages. I find it visible on-screen and in moderate (12-18") size prints.
Others may not notice it, most don't, but I see it and I'm the one who needs to be satisfied with my images.

When the sky looks like that, I don't go out taking pictures. In the Midwest, I would seek shelter.

 

[Aglet]

Yes, those cat-in-a-sack shots were actually pretty impressive hi ISO performance.
AF still quick to lock in those conditions?

AF was not as fast as normal to lock, but I am amazed that it did lock!

and with a long-ish f/4 lens yet. that's pretty good low light AF!

 

[Aglet]

@Anglet -

Top right 750 pixels@100%

Respectfully, I think if you're going to the amount of effort that you did to try to find if something is flawed, you're spending too much time trying to analyze a photo. I mean, I look at the final image on its intended size and medium (print, screen) and ask, "do I like it?". I think that's infinitely more important than whether you can see an unpleasing pattern if you apply desaturation, magnification, and apply contrast.

To put it another way, if the overall image is not pleasing, who cares if 4 millimeters in the corner exhibit a pattern; and if the overall image looks great, who cares if it does?

And I'm not talking about the photographer's skill -- I mean, the same subject taken with two different cameras. If one image is holistically better than the other (ie judged in its entirety), I'm happy to say so. But to try to say that there is a pattern that no human will ever be able discern with their naked eye is unpleasing is not meaningful to me. For example, that sigma 100-400 to Canon 100-400 comparison indicated that the Sigma was actually holistically better or indistinguishable in image quality to the much more expensive Canon counterpart, but several factors make it less suitable (or unsuitable) for certain types of photography like birding. That's a useful comparison.

Thanks for providing that sample.

I don't have to expend any effort to see pattern noise, to me it's obvious and distracts from whatever image it may be part of.

I do seem to have to accentuate it so that others can see and understand that it's there.

So your sky sample is considerably darker than Sporgon's sample.
I stiill did pretty much the same thing; 4x scale, de-sat, change overall exposure and contrast to enhance the variability in what should be a really smooth part of the scene and selected a crop from near the middle.

I think had to use more overall gain to bring out the "grit" than I did with the previous sample so the amplitude of the noise will appear greater in your sample than it otherwise would.

However, what's quite noticeable is that there's almost no fixed pattern noise on a fine vertical scale here like there was on the 5d2 sample. The noise is mosly random so looks more natural, even with some de-mosaicing type artifacts caused by some pixel level noise.
There is still some slight larger scale non-uniformity (horizontal and vertical) but that's harder to notice in real images than fine repetitive patterns.

So in this repect, it looks like the 6d2 is actually delivering a good performance compared to previous off-sensor-ADC bodies.

It's still rather noisey but your camera is not showing any significant pattern noise in this sample.
There can be sample variation, however, so it looks like you have a decent one.

Some of the samples posted by DPreview had some obvious banding and other artifacts and some others have also mentioned banding in moderately lifted shadow areas too. I'd not be surprised if this is still the case but at least your 6d2 is not putting stripes in normal exposure range blue sky some some of the older bodies did and that by itself is a good IQ improvement.

 

[Sporgon]

By having to resorting to such extreme manipulation in order to demonstrate any weakness in the 5DII's data Aglet has inadvertently shown my I still use the 5DII for quality work

Do the same thing to an Olympus file and it is much less pleasing than the example from the 6DII. Not that it really matters as no one goes there except Aglet ;D

 

[Aglet]

By having to resorting to such extreme manipulation in order to demonstrate any weakness in the 5DII's data Aglet has inadvertently shown my I still use the 5DII for quality work

Do the same thing to an Olympus file and it is much less pleasing than the example from the 6DII. Not that it really matters as no one goes there except Aglet ;D

You do quality work.
So you could potentially step up the technical quality of it even more by ditching the 5d2/3 and get something that's not adding so much noise to your workflow.
I don't understand why so many creative people choose to work with a dull tool. ???
Do you really need the extra challenge? LOL

 

[Mikehit]

By having to resorting to such extreme manipulation in order to demonstrate any weakness in the 5DII's data Aglet has inadvertently shown my I still use the 5DII for quality work

Do the same thing to an Olympus file and it is much less pleasing than the example from the 6DII. Not that it really matters as no one goes there except Aglet ;D

You do quality work.
So you could potentially step up the technical quality of it even more by ditching the 5d2/3 and get something that's not adding so much noise to your workflow.
I don't understand why so many creative people choose to work with a dull tool. ???
Do you really need the extra challenge? LOL

This part of the discussion arose because of your claims about MFT. So show us a comparison.

 

[Aglet]

Y'all asked for it.

Here's an original image; scaled from a full rez Olympus EM1 mk2 OOC jpg at base iso (200) with no tweak and default sharpening. (a tad too high... slight halo visible)
EDIT - in camera noise filter is set to OFF

I took a 100% crop from the right area and did the same thing I did to samples from Sporgon and Talys; de-sat, crank contrast and exposure amp to make the noise structure visible.

Yup, kinda noisey too.
But very film-grain-like with no significant patterning visible and no worse than Canon's FF with off-sensor ADC.
I even prefer this noise structure to the Canon's and it would clean up, if needed, quite nicely in DXO-pro with one click.

I stand by my claim.
EM1v2 IQ is comparable to previous generation of Canon FF in IQ.

 

[Talys]

You do quality work.
So you could potentially step up the technical quality of it even more by ditching the 5d2/3 and get something that's not adding so much noise to your workflow.
I don't understand why so many creative people choose to work with a dull tool. ???
Do you really need the extra challenge? LOL

I don't know if you realize it, but saying that someone's good work could just be so much better if they just did it your way... is awfully patronizing.


By the way, I see nothing superior (or inferior) about your sky. Congratulations; it's blue. It looks like sky. I think it's a bit noisier than mine, but, whatever. If you think it's a more beautiful sky, all the power to you

 

[Mikehit]

Y'all asked for it.

Here's an original image; scaled from a full rez Olympus EM1 mk2 OOC jpg at base iso (200) with no tweak and default sharpening. (a tad too high... slight halo visible)

I took a 100% crop from the right area and did the same thing I did to samples from Sporgon and Talys; de-sat, crank contrast and exposure amp to make the noise structure visible.

Yup, kinda noisey too.
But very film-grain-like with no significant patterning visible and no worse than Canon's FF with off-sensor ADC.
I even prefer this noise structure to the Canon's and it would clean up, if needed, quite nicely in DXO-pro with one click.

I stand by my claim.
EM1v2 IQ is comparable to previous generation of Canon FF in IQ.

Thankyou. So it seems no superiority but equivalence which is somewhat different to you original statements - and to be honest to be expected at pixel level nowadays for all manufacturers.

The difference is, of course, if you view the MFT and FF images at the same viewing size the MFT is being being magnified twice as much and the MFT noise will become more apparent

 

[Talys]

Keeping it real, here's the 3 side by side:

6D2-5D2-EM1MkII

Keeping in mind that this isn't apples to apples, because the lighter the color, it's harder it is to see noise.

 

[Aglet]

You do quality work.
So you could potentially step up the technical quality of it even more by ditching the 5d2/3 and get something that's not adding so much noise to your workflow.
I don't understand why so many creative people choose to work with a dull tool. ???
Do you really need the extra challenge? LOL

I don't know if you realize it, but saying that someone's good work could just be so much better if they just did it your way... is awfully patronizing.


By the way, I see nothing superior (or inferior) about your sky. Congratulations; it's blue. It looks like sky. I think it's a bit noisier than mine, but, whatever. If you think it's a more beautiful sky, all the power to you

I like Sporgon's images and we spar on such topics occasionally. I don't mean to be patronizing and it's not directed at him, really. Many creative people, here and elsewhere, get comfortable using a particular tool, whether it's a camera or a paintbrush, and it seems to become part of their comfort zone or even style.
I like using the best tool I can justify to do the job. I do not like being handicapped in any way by my tools.
HAHA! You should see the welding machine i have!

It may be a bit noisier, I'll grant that.
But it's the kind of noise that cleans up better and easier than stripes from the older systems.
I shot film starting in the 70s so I don't mind an organic film-like grain to an image, I'm not keen on screen-door artifacts. I have the same issue with digital projectors too. HD aint HD enuf.