Canon EF24-70mm f/2.8 M2 vs Sony Zeiss 24-70 on Sony a7R body

[Neutral]

I was always interested to know what is the resolution power limit of Canon EF24-70mm f/2.8 M2 combined with the high resolution body and how this compares to similar lense type from different vendors.
On Canon 1DX this is just perfect piece of glass.

Few days back I received Matabone III adaptor which gave me ability to mount Canon lenses on Sony a7R body.
With that I was finally able to test Canon EF24-70mm f/2.8 M2 on 36mpx Sony a7R body.
As result there is also comparison of image quality between Canon EF24-70mm f/2.8 M2 and Sony Zeiss 24-70 f/4 OSS.

There are two 100% crops at 24mm focal lens and two 100% crops at 70mm focal lens.
All shots are done hand held using camera autofocus.
Focus speed of Canon EF24-70mm f/2.8 M2 with Matabone III is very slow - few seconds of hunting and searching for focus peak.
Sony Zeiss 24-70 f/4 OSS focuses almost instantly.

RAW files from Sony a7R were process using Phase One Capture One with default settings, no adjustements at all

Test results demonstrate that Canon EF24-70mm f/2.8 M2 is much superior to Sony Zeiss 24-70 f/4 OSS in resolution and color reproduction and overall image quality . Especially this is noticable at 70mm focal length .
From test pictures it is obvious that even on 36mpx Sony a7R body Canon EF24-70mm f/2.8 M2 still does not reach it's limit in resolution power. This combination is perfect one when focus speed is not important.
On the other end Sony a7R with Sony Zeiss 24-70 f/4 OSS is a fairly good combination as general purpose walk-around point and shoot camera.

I also tested Canon TSE 17 on a7R and image quality , color rendering and resulution is also of top possible quality.
It seems that Canon TSE 17 the same as Canon EF24-70mm f/2.8 M2 still does not reach it's resolution power limit on a7R
Canon TSE 17 on Sony a7R looks as perfect combination for Architecture and Landscape photograpy.

 

[neuroanatomist]

Interesting results - thanks for sharing!

 

[mackguyver]

Wow, now I feel better about the money I spent on it and I'm more confident about how it will work on a high MP sensor in the future. Thanks for the test.

 

[Neutral]

Wow, now I feel better about the money I spent on it and I'm more confident about how it will work on a high MP sensor in the future. Thanks for the test.

Yes, it was also WAW for myself.

I was expecting to see some difference between them but no to such huge extent.
Pictures taken using Canon EF24-70F/2.8M2 looks like pictures taken by MF system - not only razor sharp up to single pixel on a7R sensor but also by the quality of color rendering which looks perfect to my eye (the same is applied for Canon TSE17).
Also I was surprised by quality of autofocus using Metabone III adaptor for a7R, I thought that I would have to use manual focus get best focusing results.

Pictures taken using Sony Zeiss 24-70 looks as if they were taken by cheap average camera system, and colors look a somewhat "dirty" - as if there is something not very clean and transparent between the camera and the object so results are not very pleasant to the eye.
By looking at DXOMark test results http://www.dxomark.com/Lenses/Sony/Sony-FE-Carl-Zeiss-Vario-Tessar-T-STAR-24-70mm-F4-ZA-OSS I knew that they are not in the same league as Sony Zeiss 55 F1.8 or Canon EF24-70F/2.8M2 and perceptual resolution is low (15mpx) but I expected better actual performance than I saw in my tests. Even DXO Optic Pro 9 lens correction does not help much to bring image quality close to that of Canon EF24-70F/2.8M2. It makes image a bit better but still far below of IQ when using Canon EF24-70F/2.8 M2.
I was expecting a bit more from Sony Zeiss 24-70 at least in color rendering quality.
This lens doesn't not match a7R sensor performance at all, despite the fact that Sony claimed just the opposite.
Even not up very much to the a7 sensor resolution
So Sony Zeiss 24-70mm-F4-ZA-OSS usability is limited to general walk-around average point and shoot requirements when image quality is not an issue and person just need light camera for easy carry around.

But Canon EF24-70F/2.8M2 on a7R is almost perfect combination when image quality is main concern.
a7R could be considered as high quality digital back for high quality Canon lenses including EF24-70F/2.8M2 and TSE 17 and TSE 24. The only one perfect native lens for a7R is Sony Zeiss 55 1.8 which is extremely good and allow to use all a7R resolution capacity.

 

[Neutral]

I think it might also be interesting to compare image quality of Canon 1Dx with Canon EF24-70F/2.8M2 to that of Sony a7R with Sony Zeiss 24-70mm-F4-ZA-OSS - which would be more pleasant to the eye perceptually.
Which system will win ???

 

[Sporgon]

Interesting. The Sony lens is about 60% of the Canon 2.8 II price but I guess the EF 24-70 f4 IS would stand up better than this.

However beware of critical tests hand held.

 

[Neutral]

Some more interesting test results

Canon EF24-70F/2.8 M2 on Canon 1Dx body compared to Sony Zeiss 24-70 f/4 OSS on a7R body

Again two 100% crops , no adjustements, a7R crop when exported from C1 to JPEG was downscaled to 1DX resolution.
No worry abot handheld as shutter speed in both cases is around 1/1600 so camera vibration does not affect results

At 24 mm images looks close to each other though Canon EF24-70F/2.8 M2 on Canon 1Dx perceptually looks a bit better. When 1Dx image upscaled to a7R resolution the a7R image resolution difference is visible but perceptually 1Dx image overall is still slightly better to me eye

At 70mm Canon EF24-70F/2.8 M2 on Canon 1Dx is significantly better than Sony Zeiss 24-70 f/4 OSS and this difference is very obvious.

This results just illustrates the point that the sensor resolution is only part in overall system performance. If lens is mediocre then even image from higher resolution sensor is less quality than from sensor with twice less resolution but using high quality optic.
These test actually correleate with DXOMark measurement results fo both lenses.

 

[cid]

very interesting results

I already knew 24-70mkII is excellent lens, but I'm very excited about results on A7R. One think I'd like to ask is AF performance. You already mentioned AF is slow, but both pictures look spot on. Were there no trouble with missed focus? Especially handheld?

 

[Sporgon]

No worry abot handheld as shutter speed in both cases is around 1/1600 so camera vibration does not affect results

Fair enough.


How about posting the A7 with EF 24-70 sized the same as the 1Dx.

 

[Neutral]

very interesting results

I already knew 24-70mkII is excellent lens, but I'm very excited about results on A7R. One think I'd like to ask is AF performance. You already mentioned AF is slow, but both pictures look spot on. Were there no trouble with missed focus? Especially handheld?

Canon EF24-70F/2.8 M2 AF with Metabones M3 adapter on a7R is slow but accurate.
I had no missing shots at all.
In fact Metabone focusing algorithm is very simple but solid:
It takes two passes from - to + calibrating lens and finding AF peak and then goes back and does final move to the lens focus peak
I think that ultimate resolution power of Canon EF24-70F/2.8 M2 help much in this process as focus peak could be easily detected by contrast AF system. With blurry lenses this would be a bit more difficult

 

[Neutral]

Fair enough.
How about posting the A7 with EF 24-70 sized the same as the 1Dx.

I already posted a7R with Canon EF24-70F/2.8 M2 which shows how much it is superior compared to Sony 24-70 on a7R
It could be easily downscaled to the 1DX size in any photo editing program to see the difference.
Result of normalizing image to lower resolution is obvious - perceptually it will be the same or bit better IQ only with less resolution.
So it would be a bit better that 1Dx with Canon EF24-70F/2.8 M2

 

[Sporgon]

Fair enough.
How about posting the A7 with EF 24-70 sized the same as the 1Dx.

I already posted a7R with Canon EF24-70F/2.8 M2 which shows how much it is superior compared to Sony 24-70 on a7R
It could be easily downscaled to the 1DX size in any photo editing program to see the difference.
Result of normalizing image to lower resolution is obvious - perceptually it will be the same or bit better IQ only with less resolution.
So it would be a bit better that 1Dx with Canon EF24-70F/2.8 M2

Just seen that you have obliged ! I'd already done it so here's mine. The A7r user is gonna want to print some big pictures to gain an advantage.

 

[Neutral]

...
How about posting the A7 with EF 24-70 sized the same as the 1Dx.

OK, here is comparison of A7R with EF 24-70 sized the same as the 1Dx with Canon EF24-70F/2.8 M2 on 1Dx
a7r with Canon EF24-70F/2.8 M2 looks a bit better even if downscaled to 1Dx size

 

[Neutral]

...
The A7r user is gonna want to print some big pictures to gain an advantage.

This is most common confusion and misunderstanding repeated all over different forums from person to person and this resulted that this is taken as granted by majority of the of people who are arguing about value of high MP sensors

In fact this is totally wrong.

Downscaling high MP image to the lowest resolution camera image in comparison and viewing results will never demonstrate how high MP image will look on real print. Viewing this results on the PC monitor is the same.
This is actually what many people are complaining about DXO sensors comparisons when they are normalizing images to 8mb. At these days it would be better to normalize images to 24mpx which is de facto common reference point now.

Now coming to high MP image printing.

Here are A4 and A3 printouts resolutions at fine 720dpi and superfine 1440dpi

A4 8.3 x 11.7 inches
Pixel size for 720 dpi = 11.7x720 x8.3 x720 equal to 8424 x 5976 = 50.34 MP
Pixel size for 1440 dpi = 11.7x1440 x8.3 x1440 equal to 16848 x11952 = 201.367 MP

A3 11.7 x 16.5 inches
Pixel size for 720 dpi = 11.7x720 x 16.5 x720 equal to 8424 x 11880 = 100.077 MP
Pixel size for 1440 dpi = 11.7x1440 x 16.5 x1440 equal to 16848 x 23760 =400.3 MP

So from above it is clear that in fact even most Buyer sensor MF cameras can not over perform standard quality 720dpi A4 printing resolution. On printout every dot is compete pixel . For Buyer sensor effective resolution is about 60-80% of actual pixel count. So only Phase One IQ180 image could match in resolution to fine quality 720 dpi A4 printing resolution and far below of fine quality 1440 dpi A4 print let alone A3 printouts.

So difference on IQ and resolution between 24mp and 36mp will be clearly seen even on A4 printout.
Even on small screens difference in resolution is very noticeable.
This is why people want to have retina displays on their smartphones and now new trend is 4K screens on the smartphones , let alone TV set.

As for big prints at very high resolution - they give you feeling of the real reality. From far distance you see overall image, when you get closer and closer you see more and more details - the same as in real life. On big landscape print from 400mps image you get very close and you would be able see even bee on the flower and flower petals texture and fur on the bee belly. And you will fill this as it is real reality

 

[meli]

...
Here are A4 and A3 printouts resolutions at fine 720dpi and superfine 1440dpi
...
So from above it is clear that in fact even most Buyer sensor MF cameras can not over perform standard quality 720dpi A4 printing resolution.
...

I dont think you understand what those 720dpi & 1440dpi means

 

[Neutral]

...
Here are A4 and A3 printouts resolutions at fine 720dpi and superfine 1440dpi
...
So from above it is clear that in fact even most Buyer sensor MF cameras can not over perform standard quality 720dpi A4 printing resolution.
...

I dont think you understand what those 720dpi & 1440dpi means

Are you really sure of that ?

DPI means exactly what it means - print spatial resolution - dots per inch)

"Dots per inch (DPI, or dpi)[1] is a measure of spatial printing or video dot density, in particular the number of individual dots that can be placed in a line within the span of 1 inch (2.54 cm)"
http://en.wikipedia.org/wiki/Dots_per_inch

So what I said above is 100% true for BW prints, for color prints it is a bit different depending on printer model, number of different ink cartridges in printer resulting in required number of dots to reproduce full color in single image pixel.
So for color print pixel to dot mapping is a bit different - more printer dots required per color pixel
But actual printer resolution for fine printing is 2880x1440dpi , so you can redo calculations above for this DPI density and see equivalent BW resolution for A4 and A3 prints and estimate color resolution

In general these details do not change much the main the point - modern color printers are outperforming most of the digital cameras in print resolution even on A4 prints, let alone A3 or A2 prints.
So you can see difference between high MP and average MP image printouts even on A4 format, but this difference will be almost unnoticeable if high MP image downscaled to average MP image resolution . There is no lossless downscaling - if you downscale image in resolution you will loose fine image details.

 

[Sporgon]

...
Here are A4 and A3 printouts resolutions at fine 720dpi and superfine 1440dpi
...
So from above it is clear that in fact even most Buyer sensor MF cameras can not over perform standard quality 720dpi A4 printing resolution.
...

I dont think you understand what those 720dpi & 1440dpi means

It's the old dpi/ppi thing again

 

[Neutral]

It's the old dpi/ppi thing again

Yes , agree there are a number of confusions about that

In fact relation between printer output image pixel resolution (PPI_print) and printer Dots Per Inch (DPI) resolution could be expressed in general (not going deep into details) as follows:

PPI_print = Function (DPP, DPI, Paper_Quality)
DPP (Required Dots Per Pixel) = Printer_Driver_Mapping_Function(Required_Pixel_quality, Number_of_inks_in_Printer, Avaiability_of_Variable_Droplet Technology, etc. )

Modern printers (e.g. Epson Stylus Pro 3880 or 4900 and others ) can provide final print resolution above 720 pixel per inch (required output PPI is set in print settings dialog box in imaging editing program).
Here is discussion of printer PPI ( not to be confused with DPI) - printed images final resolution using HP 932c
http://forums.adobe.com/message/4266349
"I could see significant reduction in the finest details in the 300 ppi print vs. the three higher resolution prints, and a slight reduction in the 567 ppi vs. 720."

So the argument that using Canon EF24-70 F/2.8M2 on a7R does not give any benefit for printing compared to Canon EF24-70 F/2.8M2 on Canon 1Dx is not correct.
Though difference would be almost not noticeable when downscaling a7R file to 1Dx size (details are lost in downscaling) but when full-res a7R image would be printed on A4 or A3 then difference will be visible especially on A3 print which could not be considered large - if you want to put picture on the wall in your room the best minimum size would be A2
. Also higher quality higher res image gives more flexibility to cropping.

 

[sdsr]

Test results demonstrate that Canon EF24-70mm f/2.8 M2 is much superior to Sony Zeiss 24-70 f/4 OSS in resolution and color reproduction and overall image quality . Especially this is noticable at 70mm focal length .
From test pictures it is obvious that even on 36mpx Sony a7R body Canon EF24-70mm f/2.8 M2 still does not reach it's limit in resolution power. This combination is perfect one when focus speed is not important.
On the other end Sony a7R with Sony Zeiss 24-70 f/4 OSS is a fairly good combination as general purpose walk-around point and shoot camera.

I also tested Canon TSE 17 on a7R and image quality , color rendering and resulution is also of top possible quality.
It seems that Canon TSE 17 the same as Canon EF24-70mm f/2.8 M2 still does not reach it's resolution power limit on a7R
Canon TSE 17 on Sony a7R looks as perfect combination for Architecture and Landscape photograpy.

Sony/Zeiss don't seem to have figured out how to make a first rate zoom for their A7s yet. If you compared the Sony/Zeiss 55mm 1.8 you would have seen a significant difference too (the 35mm 2.8 as well, probably). But you don't need to attach Canon lens as new/expensive as the Canon 24-70II - you can get fantastic images from many much less expensive Canon lenses on the A7r, even such bargains as the 40mm pancake and 85mm 1.8 - if nothing else, the A7r shows just how good many Canon lenses are!

 

[Neutral]

Sony/Zeiss don't seem to have figured out how to make a first rate zoom for their A7s yet. If you compared the Sony/Zeiss 55mm 1.8 you would have seen a significant difference too (the 35mm 2.8 as well, probably). But you don't need to attach Canon lens as new/expensive as the Canon 24-70II - you can get fantastic images from many much less expensive Canon lenses on the A7r, even such bargains as the 40mm pancake and 85mm 1.8 - if nothing else, the A7r shows just how good many Canon lenses are!

Fully agree that "Sony/Zeiss don't seem to have figured out how to make a first rate zoom for their A7s yet"
Sony 24-70 proved even worse than I saw in my initial tests.
Did some walk around with this lens on a7R, looked at the images and found that at short focal length images get blurry at the sides of the frame though more or less OK closer to the center of the frame.
I am very disappointed with the Sony 24-70 - seems that this lens more or less could be good for NEX7 and similar but not a good fit at all for a7R.
But Sony/Zeiss 55mm 1.8 is just perfect on a7R .

As for using other cheaper than EF24-70 F/2.8 M2 Canon lenses on a7R I agree with you that excellent results could be achieved with cheaper Canon lenses. For many people with a7R as second to Canon body this could be good solution. I just do not have them. Some of the Canon lenses though not expensive but are the best in their class
For me my EF24-70 F/2.8 M2 is the mostly commonly lens that was used on my 1DX recently and now it is also perfect fit for a7R