Resolution / Aliasing / Comparison: Sony A7 vs. A7r vs. Canon 5D mark III

We’ve been testing the new mirrorless Sony alpha a7 and Sony alpha a7R cameras that come with very promising new features dedication to video. Here’s a quick and dirty resolution comparison of the video function between these two and our all time favourite the Canon 5D mark III.

a7a735d3_1When you watch the video you can clearly see the differences between the cameras.
Some initially thought the higher resolution of the a7R versus the a7 would increase the video quality. This is true, but the test reveals that this difference is ever so tiny that it can hardly be noticed. When compared to the 5D mark III on the other hand we can see that the aliasing and moiré of the a7 as well as the a7R is pretty severe, probably similar to that of Canon DSLR’s like the 7D or TxI line.

On the left you can see the rig I used to make an accurate rolling shutter comparison. This revealed that the rolling shutter on the 5D mark III is much stronger than on the a7’s.
Is this a good thing? Yes it is. But the rolling shutter seems to be something most people take less severe as an argument against a camera.

The settings on both a7’s were the “Neutral” picture profile with everything turned down to -3. The same goes for the 5D mark III which had the “Neutral” picture profile and contrast/saturation/sharpness all the way down to 0.

There was no grading applied except for the last shot.

What I found really great about the Sony a7 was its oled viewfinder. This thing is truly amazing. I wish more cameras had this kind of technology implemented, or the A7’s had a cleaner image. In terms of functionality it’s not a bad camera by Sony. For more info check out Johnnie Behiri’s review.

Song kindly provided by themusicbed
Windmills – Creating Something Beautiful

Watch it on Vimeo

Leave a Comment

You are not subscribed to this post. Follow new comments

Login to comment

AKentPhoto November 22, 2013

Any chance you could throw one of the A7s onto the Atomos? Would love to see the difference between the camera codec and the 4.2.2 coming out of the HDMI.

Thanks! Great stuff here!

Richard November 24, 2013

I second this, would love to see the hdmi video out

Sebastian Wöber November 25, 2013

We’re on it.

Brooke November 24, 2013

We are releasing a setup video for the A7 and RX10 shortly the we filmed at InterBEE.

They both look very promising, and work great with the Ninja and Blade.

Sadly we didn’t have them for long enough to take video samples of internal versus our ProRes recording.


-Brooke | atomos

Joel Richards Reply
Joel Richards November 22, 2013

Thanks for this quick comparison! The difference between the A7 and the A7r doesn’t seem to be that great. The 5D3 certainly have less moire and aliasing but it also seems softer and less detailed?

If the rolling shutter test is any indicator the A7 has the fastest readout of all three sensors. It is disappointing that Sony did not do more the leverage the supposed speed of the Bionz X to tackle some of these issues.

Can you do a rolling shutter test between the A7 and the RX10? If the Boinz X can handle reducing the whole 20MP image of the RX10 I wonder why it can’t do the same for the A7. I suspect the issue is readout speed (or some strange software crippling)

James Villarcorta March 15, 2014

A long time user of the 5Dm3 (for video) and now a A7r owner (for stills). You are correct, the image in video is indeed softer but this is easily correctable in post with a bit of sharpening. Moire and aliasing however cant be corrected.

Phiniox Glade November 23, 2013

I keep wondering if I should of dropped the extra money for the a7r, but I wanted the kit lens. I do wish the image was better

wayne aspley November 26, 2013

A fool and their money are easily parted, or so they say. Why are people raving about full frame? Buy a Nex and a metabones system and save a tonne of cash! And what you save treat yourself to a prime that will give you the return your looking for. Better battery life to boot.

Peter Harris November 28, 2013

For still work the 2.5x sensor area of FF vs. aps is significant. For video less so, unless you want full control of depth of field and better low light capability.

Simone Rekkers January 15, 2014

I happen to shoot stills now and then. There’s no other small mirrorless camera that has this kind of resolution. Every time I look at the output, I’m amazed by the detail and dynamic range. I remember the early Sony Walkmans. They kept making them smaller and gave them an amazing feature set. Auto reverse, equalizer, a build in radio. And those things worked on only one AA battery! Sony did it again. This is an extraordinary camera. A classic. High quality build, amazing image quality and packed with features.

Simone Rekkers January 15, 2014

Yeah, I know that feeling. The most expensive camera is the three cameras you bought prior to buying the one you really want. Return or sell the A7, get an A7r and be happy.

paul chiu November 28, 2013


Please tell us what was done in your “grading” process. Thanks.

Brian Steeves November 29, 2013

I am not a professional shooter; I am just getting into video….mostly out of interest. I have a question concerning rolling shutter.

It is my understanding that this occurs when the camera is panned and the distance between the images and the shutter opening and closing varies with the pan speed.
Q. Will a faster shutter speed help to resolve this?
Q. Will changing the shutter from a vertical move to a horizontal move solve this, or does it matter?

I have two 35mm film still-cameras. The flash sync on one is 1/60 with a vertical shutter and the other at 1/125 with a horizontal shutter.

Q. Would this logically follow for the shutter on a video camera?


Sebastian Wöber November 29, 2013

Q1: No
Q2: You cannot change the shutter angle on a digital camera.

Bahman March 23, 2014

I assume EX3 and EX1 are both digital cameras and they both let you to change the shutter angle!

Joel Richards Reply
Joel Richards December 6, 2013

Perhaps a comparison using the Canon 6D would have been more apropos? After all it is priced more in the same market (the A7 being a little cheaper the A7R being a little more expensive).

rishio December 10, 2013

Are there still plans to do a review of the A7r for video? I’ve been checking the site everyday for a more detailed comparison between the A7 and A7r. If they are not that different I’ll purchase an A7. Would appreciate any thoughts if a review is not happening. Thanks!

Alan December 15, 2013

Me too. Can’t find anything anywhere on the video performance of a7 vs a7r.

Oscar December 18, 2013

Sony´s are def. not for video.
They probably rival a rebel in video q.
So in other words a FF rebel video…

dan December 21, 2013

great camera created for sure, but not for serious filming .. thanks sebastian for the test …
very helpful
eek that moire and aliasing

henri Kartmann January 4, 2014

I have a Sony A7r
What video settings will you recommand on tha A7r
I want to do a video mixing Video and still images and i work on FCP X. The export is in HD Quicktime
Thanks a lot
Henri (France)

henri Kartmann January 4, 2014

I use a Sony A7r
What video settings do you recommand on this camera ?
I want to do videos mixing Videos and Still images.
I work on FCP x and the export is in Quicktime 1080p
Thanks a lot
Henri (France)

John January 24, 2015

Sorry but this test, bar the rolling shutter demo, is close to meaningless given that the lenses on the three cameras are so radically different quality-wise. Come on, a Zeiss Sonnar against a variable aperture kit lens and Canon’s 24-104 L. This test would have more credibility and balance if the same lens (PL mount with an adapter) was used on all three. Moiré and aliasing are a test of the entire imaging chain, not just the sensor. It’s also easy to clip away the Moiré on the Canon just by throwing it a tad out of focus.

Rolf Dergham Reply
Rolf Dergham March 15, 2015

I have noticed that the rolling shutter is usually dependant on the video mode used (resolution, framerate, etc.). I’d have liked to see more details about that, and test at high framerate modes and different resolutions.

Sebastian Wöber Reply
Sebastian Wöber March 15, 2015

Hi Rolf,
We do measure these modes as well, but we think it is less relevant to many readers.
The rolling shutter in slow motion mode was exactly 50% as severe. In crop mode it was about 70% meaning that the sensor readout was the same but only the crop portion was relevant.