Sony A7RII Rolling Shutter – Compared to Sony A7s Samsung NX1 Canon 1DC Panasonic GH4

August 4th, 2015
Sony A7RII Rolling Shutter - Compared to Sony A7s Samsung NX1 Canon 1DC Panasonic GH4

Sony-a7rii-rolling-shutter-featured

Over the last days we’ve conducted several scientific lab tests, reviewed and taken the new Sony A7RII out into the field. In this article we are looking at the  Sony A7RII rolling shutter performance and see how good the sensor is in comparison to several other cameras.

Links to our other tests:

Rolling Shutter

The so called “rolling shutter” is a phenomenon that skews a camera image when fast moving objects are recorded or during fast pans and handheld camera movement. The reason for this is that most sensors read out the image line by line via a buffer.

On many CMOS cameras the rolling shutter effect has become a common issue, but some cameras have a stronger rolling shutter than others. When comparing the Sony A7RII to other cameras we can see that the rolling shutter effect is quite severe in 4K (UHD) Crop Mode. See the comparison below:

Sony-a7rii-rolling-shutter

We test the rolling shutter with a rotary chart. It always spins at the same speed and has a scale printed on it. The horizontal shift between the top and bottom line of pixels lets us roughly calculate the rolling shutter latency in milliseconds.

In Crop Mode we measured 29ms of latency on the Sony A7RII whereas in Full Frame Mode we only measured 16ms. In HD (Crop Mode) Rolling Shutter is minimal with 7ms, not far off the famous Arri AMIRA camera that has the lowest rolling shutter rating in our tests. Below you can see a chart comparing rolling shutter between several cameras:

Test-Scores-RS-Sony-A7RII

As you can see the Sony A7RII joins the Samsung NX1 which had the most severe rolling shutter we ever tested. The Canon 1DC and Sony A7s perform a little better and the Panasonic GH4 has the best values among small cinema cameras.

Conclusion

Rolling Shutter is for many not a purchase criteria. The phenomenon is mostly an issue when there is overly fast handheld movement or you film fast moving objects. However a rolling shutter of 29ms raises concerns. It is the highest measured rolling shutter among all cameras we every tested.

As we found out on Friday the Full Frame Mode is quite acceptable in terms of quality and offers a much better rolling shutter behaviour, but we also noticed that it performs badly in lowlight. It might come in handy when a better rolling shutter performance is needed and you have sufficient light available.

Comparing the Sony A7RII HD mode with that of the Sony A7s we can see that the Sony A7RII performs much better.

So rolling shutter on the new Sony A7RII is both good and bad. If you need a camera with good rolling shutter performance in 4K (UHD) you can either resort to the Sony A7RII Full Frame Mode or avoid this camera altogether and go with the Panasonic GH4 instead. The Sony A7s performs better, but the difference is not huge.

Please consider getting your camera and gear through this link. Thank you

67
Leave a reply

guest
1
1
- 2
Filter:
all
Sort by:
latest
 James Cook
Member
December 17th, 2015

I really like the Sony a7rii, but I don’t like the overheating. Thinking about the dianosour Sony nex fs700r. What do you think about its image quality as compared to the a7rii? Which camera between the 2 would you choose for video?

Member
November 4th, 2015

I am using the SONY A7Rii in Super 35mm Video mode at 100mbs. On a news set shooting promos with talent I’m finding the that the camera captures the monitors fine but there’s unacceptable banding /pulsing coming from LED background set elements. It’s ruining the footage I am able to capture. I can not find any shutter sync modes or any adjustments other than changing shutter speed to combat this. The set has all SONY XDCAM EX3 cams with shutters turned off and they do not create this same pulsing issue. I am searching all forums and anything on video related info on the A7Rii and thought I would post here hoping someone may have a suggestion. I need to either shut shutter off if possible or I’m hoping there’s some adjustment available in camera. Thank you!

Michael Carmine
Michael Carmine
Member
October 31st, 2015

Any chance you could roll the A7SII in your rolling shutter tester to get the reading of milliseconds in FF mode?

Oscar M
Member
August 16th, 2015

Is this a fried sensor or what is going on :(

http://tinypic.com/r/14446×5/8

Any answers would be appreciated.

 B Brackets
B Brackets
Member
August 28th, 2015
Reply to  Oscar M

looks like you are using the silent shutter under strobing fluro lights?

Ed Hecht
Ed Hecht
Member
August 14th, 2015

What isn’t listed here is the (supposed) 7.9ms rolling shutter of the NX1 in 1080p. Couple that with seemingly superior CAF in video (YouTube it), outside of low light situations, it appears that the NX1 is the go to film maker’s tool, no?

Felix So
Felix So
Guest
August 27th, 2015
Reply to  Ed Hecht

No? Because the lack of good picture profiles in NX1 and it’s extremely user unfriendly h265 kills it.
For Sony, you have great Profiles such as S-Log2 and the famous Cine2 and Cine4. GH4 have it’s own share of gamma like Cinelike D and V(in October the V-Log Lite). NX1 only have DR1.2 which seriously is just a neutral profile. The worse thing is that NX1’s color science is just pure rubbish(compared with other brands),especially the greens.Plus Samsung’s lens lineup are worse than Sony right now. There is no reason to go to NX1 unless you never grade and edit footage much, and uses the image straight from the camera
I need to stress again, H265 is such a pain to work with.

Ed Hecht
Ed Hecht
Member
August 28th, 2015
Reply to  Felix So

Point taken on H.265. I also just realized that it outputs only an 8-bit 4-2-2 signal via HDMI. I would have been willing to pair it with an external recorder, if it output uncompressed like the A7S/A7R II…

Alexander Tardif
Alexander Tardif
Member
August 7th, 2015

Sebastian – thanks for the tests. Quick question, on a somewhat related topic: APS-C/Super 35mm menu has 3 options: On, Auto, Off. The On/Off are self explanatory, but I’m a bit confused about “Auto” in video mode. It appears that when this setting is set to Auto, the video automatically goes to S35 mode and if I want to shoot in FF, I have to set it to Off. Is there an official explanation on how works somewhere? Again, I understand the stills part of this selection, but not the video.

Thanks!

Jason Varner
Jason Varner
Guest
August 6th, 2015

How is it possible for the shutter readout to be 29ms when at 24p with a 48th shutter the sensor is only “open” for 20ms?

Mel Feliciano
Mel Feliciano
Member
August 6th, 2015

Another byproduct of the CMOS sensor with a “rolling shutter” is the partial exposure in the presence of strobe flashes. You can see that nasty effect on celebrity footage when there are a lot of paparazzi taking pictures (TMZ, Extra, Insider…). It looks even worse in slow motion. Is partial exposure still an issue on newer cameras?

C5D User
C5D User
Member
August 5th, 2015

I know it lives in a different camera space but for my novice needs I’m considering the rx10m2. Would we be safe to assume the rolling shutter issues in the a7r2 could be found/expected in the rx10m2?

Oscar M
Member
August 5th, 2015
Reply to  C5D User

Plenty of rx10II footage on youtube so just check it out :)
like this amateur video – this kind of footage will give you the best answer for novice use :)

https://www.youtube.com/watch?v=gw4rLTFiABw

C5D User
C5D User
Guest
August 5th, 2015

I know it’s in a different camera space (fixed lens, etc…) but would you say we should expect the rx10m2 would have the same rolling shutter issues as the a7r2?

Ronald Vonk
Guest
August 4th, 2015

Ugh. I’ll personally never get used to it though. I’ll be happy when global shutter sensors have trickled down to all budget levels.

B Brackets
B Brackets
Guest
August 4th, 2015

Why no new data on the NX1? Why no FHD RS data for the NX1? (it’s there for the sony), Why no new review for the NX1 with so many improvements with firmware?

tom kelly
tom kelly
Member
August 4th, 2015
Reply to  B Brackets

Why don’t you do it and post your own findings? This article is focusing on the new A7r II and this is a great site, so be grateful for it.

 B Brackets
B Brackets
Member
August 4th, 2015
Reply to  tom kelly

Great reply!!! Sorry if i think that 9 months is enough time to revisit a negative review of a competitor to this camera that has had many serious firmware updates. This would then make this A7RII info more relevant . My question is still valid.

 B Brackets
B Brackets
Member
August 4th, 2015

Sure Sebastian except Austria is close to Australia in spelling only :-)

Nino Leitner
Guest
August 4th, 2015

Good point Andrew. I have the same experience. But audiences are oblivious to many new advances of technology.

Dan Crow
Dan Crow
Guest
August 4th, 2015
Reply to  Nino Leitner

If rolling shutter is a byproduct of large sensor cameras giving us greater exposure latitude for a more modest price, I think I’m ok with that.

It seems more likely that rolling shutter, when noticeable, would be easier to fix than dealing more global issues related to a lower dynamic range. But I may not be seeing the whole picture here.

Anyone have experience on both sides of the issue?

Dan Crow
Dan Crow
Member
August 4th, 2015
Reply to  Dan Crow

and by ‘fix’, I mean fix in post.

Andrew Howe
Guest
August 4th, 2015

I am starting to wonder whether rolling shutter is something audiences worry about or whether they just accept it like backwards turning wagon wheels in old westerns? I see it all the time on TV these days. A part of my brain says “Ooh look rolling shutter” but any attempt on my part to engage my family on the pitfalls of CMOS sensors is greeted with contempt or the occasional well aimed cushion.

Kris
Kris
Member
August 5th, 2015
Reply to  Andrew Howe

That’s not rolling shutter. The spinning-wheel thing is, I think, stroboscopic effect, a.k.a. judder. A larger shutter angle would blur the spokes, though the blurred spokes themselves might be spinning the wrong way. ALL cameras have that issue.

With the A7RII, the spokes on the backward spinning wheels (strobing) would look like they’re made of bendy noodles (rolling shutter.)

I mention this because I think the rolling shutter on the new Sony is a huge problem, given its price point, and shouldn’t be shrugged off.

User Experience
User Experience
Member
August 4th, 2015

Sorry if I missed this in your post but I actually can not find any info regarding the rolling shutter measurements for the Sony A7rII in full frame HD.

You’d think this would have been included in the test?

User Experience
User Experience
Member
August 4th, 2015

Sorry, I mean full frame 1080.

User Experience
User Experience
Member
August 4th, 2015

Thank you Sebastian.
I can imagine most folks want to see what this camera can do in 4k, but it would be great if you folks can perform the same tests in 1080. I say this because there are probably a great many of us who are currently underway with a project and looking to upgrade to a new camera but keep shooting primarily in 1080… and sometimes go to 4k for certain shots.

User Experience
User Experience
Member
August 4th, 2015

For example, how does the camera handle low light at 1080 full frame vs. crop?

tom kelly
tom kelly
Member
August 4th, 2015

That makes no sense as HD Crop result was 7ms so judging by the UHD results HD FF should be even better than that but your saying its 2 x worse?

tom kelly
tom kelly
Member
August 4th, 2015

Thats interesting, thanks for checking. Sounds like this camera can cope with most situations as long as the operator is knowledgeable enough to have it set up correctly for the given situation.

Member
April 1st, 2016

Is the 720p from XAVC-S HD 120p mode?
Thanks!

Gum Bum
Gum Bum
Member
August 4th, 2015

My dream camera at this point: a 4k A7s with a global shutter.

The first of these small camera manufacture to put a global shutter in their product win a price.
I don’t know if it’s technically possible, but I would buy that camera in a heartbeat.

Gum Bum
Gum Bum
Member
August 4th, 2015

The BMC 4k has a global shutter and about 12 stops DR. But maybe there are some physical limitations to implementing that in a smaller body?

Gum Bum
Gum Bum
Member
August 4th, 2015

Oh wauw! Really? Talk about false advertising then.

Member
August 10th, 2015

For the record, the upcoming Blackmagic 4.6K sensor retains full DR when switching between global and rolling shutter modes. A first for any sensor in both regards. #coolfactoid

Mel Feliciano
Mel Feliciano
Member
August 10th, 2015
Reply to  Solorio Marco

Are you sure about that? That would be awesome. Dan May from BMD had said in a couple of interviews that the DR for the 4.6k sensor is 15 stops in Rolling Shutter mode and about 13 stops in Global Shutter mode.

Member
August 10th, 2015
Reply to  Mel Feliciano

Yes, I am sure about this. These are fairly recent findings.

Mel Feliciano
Mel Feliciano
Member
August 11th, 2015
Reply to  Solorio Marco

Fantastic news :)

Member
August 11th, 2015

I started testing with the camera/sensor in beta almost 2 months ago in June.
http://www.onerivermedia.com/blog/?p=1614
Although not in the wild, the findings are still valid.

Mel Feliciano
Mel Feliciano
Member
September 9th, 2015
Reply to  Solorio Marco

Why does Dan May keep saying 12 to 13 stops for the 4.6k sensor on Global Shutter mode?. This is from a resent interview by Noam Kroll at the Blackmagic Day event in LA.

Member
September 9th, 2015
Reply to  Mel Feliciano

Was it in reference to ProRes or RAW? Keep in mind that as President, Dan oversees *all* of BMD’s products for the US market (cameras, switchers, converters, software, etc.), which is a ton to remember. My guess is that he either is (A) referring to ProRes, (B) simply isn’t aware that the DR is maintained in GS mode due to the fact he does not work on the technical side of development, or (C) he just simply forgot while in the moment of being in an interview (it happens to the best of us). Unless something changes from the current prototype/beta sensor, full DR is maintained in either GS or RS modes. I covered this very topic on stage while presenting at the same event (and the NYC event).

Mel Feliciano
Mel Feliciano
Member
September 9th, 2015
Reply to  Solorio Marco

They don’t specify if it is ProRes or Raw.
https://vimeo.com/137554962
The topic comes up at 1:03
Would like to watch a video of your presentation but I can’t find any footage on Vimeo or Youtube.

Member
September 9th, 2015
Reply to  Mel Feliciano

My hunch is that Dan simply might not be aware of the new info (he and I are actually planning to meet for lunch tomorrow so I’ll ask him if he’s aware of the update). During the time of that BMD event, the discovery of equal DR between GS and RS (in RAW) was only a few short weeks prior to the event and was done at HQ in Melbourne, Australia. BMD may have not done a corporate-wide notice on the findings since it’s still prototype/beta. Who knows. The presentations from the BMD events are not streamed (I assume it’s because they want people to show up in person to play hands-on with all the new products).

Mel Feliciano
Mel Feliciano
Member
September 9th, 2015
Reply to  Solorio Marco

Since you guys are having lunch tomorrow, there is something I would like you to bring up in the conversation. I’ve read your blog post (from 2013) about how to rig the BMCC as a shoulder mount ENG camera. Makes me wonder if that’s where they got the idea for the Ursa cameras :). This is probably in the pipeline, but I think that the next step for BMD is to make a proper ENG camcorder. The Ursa mini is pretty close but it needs a shotgun mic holder, built-in ND filters, and accessible audio controls (not hidden behind the flip screen), and maybe a smaller sensor for B4 lenses. They have the cinema and studio cameras, now they need the ENG cameras. Technology is getting more affordable, but those ENG/Shoulder-mount cameras from Sony and Panasonic are ridiculously expensive; even the old SD/HD models. One last thing, BMD products support the latest HD-SDI and Thunderbolt connections, now we need HDMI 2.0. Maybe those will be some of the new announcements for this year’s IBC.

Member
September 10th, 2015
Reply to  Mel Feliciano

Mel, this message system looks like it maximizes reply count and as such I couldn’t reply directly to your comment below regarding BMD building a future ENG camera. They get this request all the time (through their forum, social media, and in person at public events) and wouldn’t be anything new to ask Dan and the others there. Likewise, they do not discuss any future products they’re working on in public outlets. They are very “Apple” like in that regard, and rightly so to protect their developments and IP. And if Dan did mention anything new to me in that regard, I couldn’t discuss it anyway since I’m under NDA with BMD. I can only echo information/product announcements that have already been officially announced publicly by BMD themselves. But I will ask him if he’s learned about the DR advancements!

Mel Feliciano
Mel Feliciano
Member
September 10th, 2015
Reply to  Solorio Marco

No problem Marco. One thing is for sure, BMD is going to bring their A-game to IBC and shake the industry once again with some new announcements. I met Grant Petty 8 years ago at NAB, really cool guy. We even talked about his old company Digital Voodoo (Now Bluefish444). Glad to have some one so close BMD always willing to reply to our comments. We will continue our conversation at the OneRiverMedia blog.

Member
September 12th, 2015
Reply to  Mel Feliciano

So as I suspected (after meeting with Dan yesterday), he did not know or hear of the DR topic. Obviously BMD did not do a corporate-wide notification on the topic and was primarily only tested by a few. This makes sense though since the sensor was/is prototype and anything could change. And in fact, they are STILL optimizing the sensor to this day (I’m a bit surprised by that, but I’d rather have that and maximize its full potential). Likewise, there are reps at the BMD booth at IBC that are confirming that the sensor is still being optimized at this very moment. I’m so impressed with the results of my own testing of the sensor that I’m surprised it’s still being further optimized, which makes me wonder what else they might have up their sleeves with the sensor. My only guess is maybe an electronic ND (like RED Motion Mount, Sony FS5, etc.) on the sensor itself (alluding to the possibility of an internal ND solution was also publicly mentioned by BMD at IBC, so I can vocally speculate on this). Whatever BMD adds to this sensor, will be worth the wait, even if just standard optimization. It’s definitely a thing of beauty.

Ed Hecht
Ed Hecht
Member
September 9th, 2015
Reply to  Solorio Marco

Does this also apply to the 4K sensor of the Ursa Mini?

Member
September 9th, 2015
Reply to  Ed Hecht

Unfortunately not. Only the new 4.6K sensor for the URSA and URSA Mini. It’s an entirely new sensor built from the ground up by BMD.

Jett He
Jett He
Guest
October 31st, 2015
Reply to  Solorio Marco

That is completely FALSE. It’s been stated many times over and over. It looses about 2stops when switching to a Global shutter. This is how misinformation spreads like wildfire. The misinformed sells lies as facts.

Member
October 31st, 2015
Reply to  Jett He

So you’re calling me misinformed and selling lies? For one, I’m not selling anything as I do not sell BMD cameras and am not an employee of the company. Second: I’ve beta tested just about all of the BMD cameras since mid-2012, including the URSA Mini 4.6K EF and PL under the prototype sensor. Third: I received backing information directly from BMD and fellow beta testers directly. Forth: This was done under the prototype sensors and like anything, things can change, but from what was tested, the results stand. I will gladly retract any findings based on the shipping sensor (or an updated prototype), but until then, the findings stand. Have you personally tested the 4.6K sensor? Where or who are you getting the information from that says that it is losing 2 stops when switching between GS and RS? Unless it’s from you testing it or from a BMD beta tester, then you are the one claiming falsities. Sadly, not everyone within BMD even knows of these findings, including BMD reps at events/conferences. THAT does not mean the actually root findings are false, but rather, lack of communication within the company. There is also the possibility BMD in general does not want to officially claim the findings UNTIL the shipping sensor is manufactured and shipped. In the end, these findings were NOT false and definitely NOT lies.

 Daniel Rozsnyo
Daniel Rozsnyo
Member
August 1st, 2019
Reply to  Solorio Marco

Well Marco, its 2019 now.. where is the promised GS 4K6 with not lost stops? All your posts above around this are now confirmed FALSE and LIES. Thats it – people were right, you were NOT.

1
1
- 2
Filter:
all
Sort by:
latest

Take part in the CineD community experience