A7RII initial thoughts and images

[uaiomex]

Thank you much Edmund. Great links. I loved the simplicity and the objectivity of both tests.
Now I wish for long exposure tests at base iso. Something in the 4-6 minute long exps.
Eduardo

There's an A7R2 and 5DsR image quality and noise (second page) comparison that has just gone up..
It has some real world images.
The two cameras seem close.
Anyway, images are better than words.


http://www.cameralabs.com/reviews/Sony_Alpha_A7r_II/outdoor_resolution.shtml
http://www.cameralabs.com/reviews/Sony_Alpha_A7r_II/noise.shtml

Edmund

[Jack Hogan]

No comment

Edmund,

What do you think of this post where the a7RII's color performance appears to be the same or better than most current ILCs?

[synn]

Saw online from an A7R II owner that the color rendering does look better when the Phase One IQ 250 profile is used in C1P, as is the case with the Nikons.
Might be worth exploring.

[synn]

There's an A7R2 and 5DsR image quality and noise (second page) comparison that has just gone up..
It has some real world images.
The two cameras seem close.
Anyway, images are better than words.


http://www.cameralabs.com/reviews/Sony_Alpha_A7r_II/outdoor_resolution.shtml
http://www.cameralabs.com/reviews/Sony_Alpha_A7r_II/noise.shtml

Edmund

I own neither, but looking at the comparison, the Canon images look like they have a 0.5px gaussian blur filter applied in comparison to the Sony.
Standalone, the Canon seems to be resolving details on the same level as the D800 (This one I know first hand), the Sony, quite a bit more than that.

Seems like no AA filter > self canceling AA filter.

[Hans Kruse]

I own neither, but looking at the comparison, the Canon images look like they have a 0.5px gaussian blur filter applied in comparison to the Sony.
Standalone, the Canon seems to be resolving details on the same level as the D800 (This one I know first hand), the Sony, quite a bit more than that.

Seems like no AA filter > self canceling AA filter.

The D800E has the same approach as the 5DsR and only the D810 has not AA filter. The 5DsR clearly has higher resolution than the D810 and I own both! The big questions is: Does it really matter?

[synn]

The D800E has the same approach as the 5DsR and only the D810 has not AA filter. The 5DsR clearly has higher resolution than the D810 and I own both! The big questions is: Does it really matter?

Hi Hans,

Yes, I am aware of the fact that the D800E has the same approach. In my personal tests, I really struggle to find any noticeable difference between the D800 and the D800E in terms of detailing rendered. The 810 might be better as it has no AA filter.
You might be right that the 5DsR has higher resolution than the D810 (I am just working from memory here, seeing them side by side would obviously be better), but it is definitely lacking in comparison tot he Sony, in the quoted example.

As for the "Does it matter"? question, it's very subjective. To me, it does. To someone else, it might not.

[Bart_van_der_Wolf]

I own neither, but looking at the comparison, the Canon images look like they have a 0.5px gaussian blur filter applied in comparison to the Sony.

Hi Sandeep,

They were apparently ("Sharpening at 70 / 0.5 / 36 / 10, Luminance and Colour Noise Reduction both set to zero,") sharpened with mostly USM sharpening, while at these small sensel pitches (and at f/8 !) deconvolution sharpening makes a world of difference.

Cheers,
Bart

[synn]

Hi Sandeep,

They were apparently ("Sharpening at 70 / 0.5 / 36 / 10, Luminance and Colour Noise Reduction both set to zero,") sharpened with mostly USM sharpening, while at these small sensel pitches (and at f/8 !) deconvolution sharpening makes a world of difference.

Cheers,
Bart

Hi Bart,
Indeed, deconvolution sharpening does make a huge difference and is an integral part of my D800 workflow now.
But it is indeed fascinating to see that head to head, the Sony delivers more details per pixel, out of the box.

[Bart_van_der_Wolf]

Hi Bart,
Indeed, deconvolution sharpening does make a huge difference and is an integral part of my D800 workflow now.
But it is indeed fascinating to see that head to head, the Sony delivers more details per pixel, out of the box.

Well, it's physically impossible for the Sony sensor to produce higher resolution with the same lens and proper focus, so it's probably something else (defocus, camera vibration, raw converter, aliasing, lack of deconvolution sharpening) causing that. There is a physical resolution difference of around 10% (sensel pitch), so with the same lens the Canon should be visibly sharper at full output size (and the same detail is larger), even for medium and lower contrast features.

Cheers,
Bart

[Slobodan Blagojevic]

..., so it's probably something else...

Confirmation bias.

[synn]

Confirmation bias.

...or the fact that one camera has on AA filter and the other, an AA-but-not-really filter, as I said in the first post.

[adias]

I own neither, but looking at the comparison, the Canon images look like they have a 0.5px gaussian blur filter applied in comparison to the Sony.
Standalone, the Canon seems to be resolving details on the same level as the D800 (This one I know first hand), the Sony, quite a bit more than that.

Seems like no AA filter > self canceling AA filter.

I see exactly the opposite. The 5DSR images are sharper and have more detail. Make sure you are viewing the comparo on a display where the browser window is large enough. Also... if you wear glasses it is easy to favor one side of the screen to the other... thus proper head/glasses viewing alignment is crucial.

Note: I do not own either camera (at this time).

[synn]

I compared the images on my calibrated 30" Ultrasharp. No glasses either, 20:20 vision here.

[Bart_van_der_Wolf]

I compared the images on my calibrated 30" Ultrasharp. No glasses either, 20:20 vision here.

Do note that the smaller 5DS R pixels magnify the details 10% more, and thus larger detail will seem to be of lower resolution in a side by side comparison, but actually shows more detail (assuming identical focus).

Cheers,
Bart

[adias]

I compared the images on my calibrated 30" Ultrasharp. No glasses either, 20:20 vision here.

Tried on a second screen just to confirm - same result: 5DSR side sharper/more detailed.

[Torbjörn Tapani]

To me that last frame seems sharper for the Sony. Overall it is close. We can stop worrying about resolution and focus on other aspects of sensor performance.

[synn]

Do note that the smaller 5DS R pixels magnify the details 10% more, and thus larger detail will seem to be of lower resolution in a side by side comparison, but actually shows more detail (assuming identical focus).

Cheers,
Bart

Good point. I downloaded some samples for both cameras from imaging-resource, downscaled the %DSR image to the dimensions of the A7RII sample and looked at them side by side. Please see attached.
A lot closer now, witht he Sony image looking a tad more detailed. But that could very well be extra contrast out of the box and can be equalized in post.

[BernardLanguillier]

Both are close enough with Sigma 50mm f1.4 used at f8 for this test that the resolution difference seems mostly irrelevant.

The gap could be a bit larger with the Otus at f4, especially on true AA filter less bodies, but the key takeaway for me here is that 40 vs 50mp is of secondary importance even in these tests and probably even more in real world applications.

Cheers,
Bernard

[eronald]

Both are close enough with Sigma 50mm f1.4 used at f8 for this test that the resolution difference seems mostly irrelevant.

The gap could be a bit larger with the Otus at f4, especially on true AA filter less bodies, but the key takeaway for me here is that 40 vs 50mp is of secondary importance even in these tests and probably even more in real world applications.

Cheers,
Bernard

Just like the eventual gap between the Otus and the much cheaper AF Sony Zeiss 50
I would expect a $200 "plastic fantastic" with similar performance to the Otus to appear as soon as the 40MP sensor moves into the consumer range.

Edmund

[BernardLanguillier]

Just like the eventual gap between the Otus and the much cheaper AF Sony Zeiss 50
I would expect a $200 "plastic fantastic" with similar performance to the Otus to appear as soon as the 40MP sensor moves into the consumer range.

That would be fantastic!

Cheers,
Bernard