Rolling Shutter Guide

[Ed_David]

I have always found this thread to be really important in deciding on using certain cameras inside vehicles, with fast motion:

http://www.dvxuser.com/V6/showthread.php?303559-Measuring-rolling-shutter-put-a-number-on-this-issue!

Thanks to Samuel

Going to post text as well, in case dvx100's forum goes under:

RX100 V 250 fps ---- 3.8 ms (3.8-3.8)
BM 4.6K (2K crop)--- 6.3 ms (official)
1DX II 1080p 60&120- 6.7 ms (6.6-6.7)
NX1 1080p ---------- 7.9 ms (7.7-8.0-8.1-7.8)
NX500 2.5k --------- 8.6 ms (8.2-8.8-8.9)
a99 II 1080p ------- 8.7 ms (8.8-8.7)
BM MicroStudio 4K--- 9.9 ms (official)
a7r II 1080p S35--- 10.5 ms (11.0-10.0)
a7r II 1080p FF---- 10.8 ms (11.0-10.5)
1DX II 1080p 24 --- 11.1 ms (11.1-11.2)
a6000 1080p24&30--- 12.2 ms (13.9-11.1-11.5)
BM 4.6K (4K crop)-- 12.6 ms (official)
BMMCC ------------- 13.3 ms (official) my measurement was 13.4 ms (13.7-13.5-12.9)
GH4 1080p --------- 13.7 ms (13.7-13.5-13.1-14.0-13.9-13.5-13.5-13.9)
RX100 V 1080p ----- 13.9 ms (14.4-14.0---14.2-13.1) (first 2 with stabilization, last 2 without)
D750 -------------- 14.5 ms (14.6-14.1-14.6-14.8)
1DX II 4k 24&60 --- 14.7 ms (14.6-14.9-14.9-14.6)
RX10 -------------- 14.8 ms (14.5-14.8-15.0)
GH5 --------------- 15.0 ms (official, preproduction)
BM 4.6K ----------- 15.2 ms (official) my measurement was 16.3 ms (16.3-15.9-16.8)
a6300 1080p24 ----- 15.2 ms (15.8-14.6)
GX85 1080p -------- 15.4 ms (13.8-16.7-15.7)
GH3 --------------- 15.5 ms (15.4-15.7-15.4)
G7 1080p ---------- 16.9 ms (17.5-16.3)
XT2 1080p --------- 16.9 ms (16.9-16.0-17.4-17.3)
RX100 IV 1080p ---- 16.9 ms (16.5-17.3) (without stabilization it's slightly faster: 16.1)
RX100 V 4k -------- 17.5 ms (18.1-16.3-17.7---17.7-16.9-18.5) (first 3 with stabilization, last 3 without)
BMPCC ------------- 17.7 ms (official) my measurement was 17.8 ms (17.7-17.7-20.0-20.0)
a7s APS-C 1080p --- 19.5 ms (20.3-18.4-19.5-19.5)
a7R II 4K FF ------ 19.9 ms (19.3-19.6-19.4-21.2)
5D3 --------------- 20.5 ms (20.7-20.5-20.4)
D5200 ------------- 22.4 ms (22.5-22.1-22.6)
GH4 4K/UHD -------- 22.0 ms (official) my measurement was 22.5 ms(23.2-22.3-22.7-22.7-22.4-21.8-22.4-22.8-22.6)
a99 II 4k --------- 23.4 ms (23.3-23.5)
BMC --------------- 23.6 ms (official) my measurement was 25.0 ms (26.7-24.8-23.5)
5D2 --------------- 25.9 ms (25.5-26.4-25.8)
5Dsr -------------- 27.7 ms (27.5-27.9-27.6)
G7 4k ------------- 28.1 ms (27.8-28.5)
GX85 4k ----------- 28.9 ms (30.2-29.0-28.8-27.5)
NEX-5N ------------ 29.4 ms (28.8-29.6-28.9-29.8-29.8-29.1-29.7)
a7s II FF 1080p --- 30.3 ms (29.5-31.2)
a7s II FF 4K ------ 30.4 ms (30.2-30.6)
a7s FF 1080p ------ 30.5 ms (30.1-32.0-30.5-30.3-29.2-30.9)
XT2 4k ------------ 30.7 ms (30.5-30.7-31.2-31.2-30.8-29.9)
NX1 4K ------------ 30.9 ms (30.6-31.6-31.4-30.7-30.2)
NX1 UHD ----------- 32.6 ms (32.9-32.0-32.9-32.5)
a7R II 4K S35 ----- 33.3 ms (35.6-32.2-32.8-35.3-32.6-31.1)
RX100 IV 4K ------- 36.6 ms (36.4-36.7) (without stabilization it's slightly faster: 35.7)
a6300 4K 24fps ---- 39.0 ms (39.2-38.9)

OTHER NUMBERS, MOST OF THEM FOUND BY SQUIG SOMEWHERE ON THE WEB:

Film --------------- 5 ms? (here)
Scarlet ----------- 14 ms
F65 --------------- 14 ms
AF100 ------------- 14.85 ms
FS100 ------------- 15 ms
FS7 --------------- 15 ms (here)
C300 -------------- 16 ms
Red One MX -------- 16.6 ms
7D ---------------- 21 ms
5D MKII ----------- 25 ms
GH1 --------------- 25 ms
D90 --------------- 33 ms
 

 

I am curious to see how the sony a9, a99, and canon 1dx ii stack up against each other in rolling shutter.

I get asked a lot to do crash cameras in cars - and in the past, I have used the a7s ii in 1080p mode.  But these guys now seem like a better option to prevent jello.

[Andrew Reid]

Just the 4K cams

BM MicroStudio 4K--- 9.9 ms (official)
BM 4.6K (4K crop)-- 12.6 ms (official)
1DX II 4k 24&60 --- 14.7 ms (14.6-14.9-14.9-14.6)
GH5 --------------- 15.0 ms (official, preproduction)
BM 4.6K ----------- 15.2 ms (official) my measurement was 16.3 ms (16.3-15.9-16.8)
RX100 V 4k -------- 17.5 ms (18.1-16.3-17.7---17.7-16.9-18.5) (first 3 with stabilization, last 3 without)
a7R II 4K FF ------ 19.9 ms (19.3-19.6-19.4-21.2)
GH4 4K/UHD -------- 22.0 ms (official) my measurement was 22.5 ms(23.2-22.3-22.7-22.7-22.4-21.8-22.4-22.8-22.6)
a99 II 4k --------- 23.4 ms (23.3-23.5)
G7 4k ------------- 28.1 ms (27.8-28.5)
GX85 4k ----------- 28.9 ms (30.2-29.0-28.8-27.5)
a7s II FF 4K ------ 30.4 ms (30.2-30.6)
XT2 4k ------------ 30.7 ms (30.5-30.7-31.2-31.2-30.8-29.9)
NX1 4K ------------ 30.9 ms (30.6-31.6-31.4-30.7-30.2)
NX1 UHD ----------- 32.6 ms (32.9-32.0-32.9-32.5)
a7R II 4K S35 ----- 33.3 ms (35.6-32.2-32.8-35.3-32.6-31.1)
RX100 IV 4K ------- 36.6 ms (36.4-36.7) (without stabilization it's slightly faster: 35.7)
a6300 4K 24fps ---- 39.0 ms (39.2-38.9)

A few remarks

Don't think the number for the X-T2 is quite right, I'd say in my real world tests it comes out quite a bit better than the NX1 4K

30ms is where it gets bothersome, under that is fine for most stuff. Interesting that the A99 II 4K came out slightly worse than A7R II - it must be doing a better image though, more pixels in the readout.

Fastest stills camera sensors - GH5 and 1D X II... I think Canon could have given us better quality 120fps 1080p given how fast that chip is.

Good to see BM at the top - power hungry though.

Best value for money is GH5 and RX100 V.

[EthanAlexander]

Here they are in alphabetical order  4K(+) Bolded. I also Added a few from C5D numbers, and cleaned up the list.

• RX100 V 250 fps ---- 3.8 ms 
• 1DC S35HD --------- 18ms
• 1DC 4K ------------- 25ms
• 1DX II 1080p 24 --- 11.1 ms 
• 1DX II 1080p 60&120- 6.7 ms 
• 1DX II 4k 24&60 --- 14.7 ms 
• 5D MKII ----------- 25 ms
• 5D2 --------------- 25.9 ms 
• 5D3 --------------- 20.5 ms (18ms C5D)
• 5Dsr -------------- 27.7 ms 
• 7D ---------------- 21 ms
• a6000 1080p24&30--- 12.2 ms 
• a6300 1080p24 ----- 15.2 ms 
• a6300 4K 24fps ---- 39.0 ms 
• a7r II 1080p FF---- 10.8 ms 
• a7r II 1080p S35--- 10.5 ms 
• a7R II 4K FF ------ 19.9 ms 
• a7R II 4K S35 ----- 33.3 ms 
• a7s APS-C 1080p --- 19.5 ms
• a7s FF 1080p ------ 30.5 ms  (29ms C5D)
• a7s II FF 1080p --- 30.3 ms 
• a7s II FF 4K ------ 30.4 ms 
• a99 II 1080p ------- 8.7 ms 
• a99 II 4k --------- 23.4 ms 
• AF100 ------------- 14.85 ms
• Alexa 2K ------------ 6ms
• Amira S35 2K ------- 2ms
• BM 4.6K ----------- 15.2 ms (official) my measurement was 16.3 ms 
• BM 4.6K (2K crop)--- 6.3 ms (official)
• BM 4.6K (4K crop)-- 12.6 ms (official)
• BM MicroStudio 4K--- 9.9 ms (official)
• BMC --------------- 23.6 ms (official) my measurement was 25.0 ms 
• BMMCC ------------- 13.3 ms (official) my measurement was 13.4 ms 
• BMPCC ------------- 17.7 ms (official) my measurement was 17.8 ms 
• C300 -------------- 16 ms
• C300 II 4K --------- 6ms
• D5200 ------------- 22.4 ms 
• D750 -------------- 14.5 ms 
• D90 --------------- 33 ms
• F65 --------------- 14 ms
• Film --------------- 5 ms?
• FS100 ------------- 15 ms
• FS7 4K ----------- 15 ms (here) (14ms C5D)
• G7 1080p ---------- 16.9 ms 
• G7 4k ------------- 28.1 ms
• GH1 --------------- 25 ms
• GH3 --------------- 15.5 ms 
• GH4 1080p --------- 13.7 ms 
• GH4 4K/UHD -------- 22.0 ms (official) my measurement was 22.5 ms (C5D: 14ms)
• GH5 --------------- 15.0 ms (official, preproduction)
• GX85 1080p -------- 15.4 ms 
• GX85 4k ----------- 28.9 ms 
• NEX-5N ------------ 29.4 ms 
• NX1 1080p ---------- 7.9 ms 
• NX1 4K ------------ 30.9 ms 
• NX1 UHD ----------- 32.6 ms
• NX500 2.5k --------- 8.6 ms 
• Red One MX -------- 16.6 ms
• RX10 -------------- 14.8 ms 
• RX100 IV 1080p ---- 16.9 ms  (without stabilization it's slightly faster: 16.1)
• RX100 IV 4K ------- 36.6 ms  (without stabilization it's slightly faster: 35.7)
• RX100 V 1080p ----- 13.9 ms(first 2 with stabilization, last 2 without)
• RX100 V 4k -------- 17.5 ms  (first 3 with stabilization, last 3 without)
• Scarlet ----------- 14 ms
• XT2 1080p --------- 16.9 ms 
• XT2 4k ------------ 30.7 ms 

[no_connection]

I don't see why X-T2 would be incorrect, I think I got about that number when I used the promo clip and calculated it.

[Don Kotlos]

A9 FF 4K should be about ~25ms : http://www.imaging-resource.com/news/2017/08/22/a-new-test-debuts-at-ir-the-sony-a9-rolling-shutter-is-pretty-minimal

There are reports that the S35 has faster readout with not much hit on the quality. 

Such a shame not to implement the ~7ms of stills during video mode. 

I am interesting to see what they did with RX0 that has the same fast readout technology, and is targeted for "action" ... 

The other camera that would be interesting to check is the E-M1 mkII. It will probably be better than all other 4K cameras here. Not the most cinematic look, but coupled with the excellent IBIS it will be very useful for handheld stuff. 

[Ed_David]

2 hours ago, Andrew Reid said:

Just the 4K cams

BM MicroStudio 4K--- 9.9 ms (official)
BM 4.6K (4K crop)-- 12.6 ms (official)
1DX II 4k 24&60 --- 14.7 ms (14.6-14.9-14.9-14.6)
GH5 --------------- 15.0 ms (official, preproduction)
BM 4.6K ----------- 15.2 ms (official) my measurement was 16.3 ms (16.3-15.9-16.8)
RX100 V 4k -------- 17.5 ms (18.1-16.3-17.7---17.7-16.9-18.5) (first 3 with stabilization, last 3 without)
a7R II 4K FF ------ 19.9 ms (19.3-19.6-19.4-21.2)
GH4 4K/UHD -------- 22.0 ms (official) my measurement was 22.5 ms(23.2-22.3-22.7-22.7-22.4-21.8-22.4-22.8-22.6)
a99 II 4k --------- 23.4 ms (23.3-23.5)
G7 4k ------------- 28.1 ms (27.8-28.5)
GX85 4k ----------- 28.9 ms (30.2-29.0-28.8-27.5)
a7s II FF 4K ------ 30.4 ms (30.2-30.6)
XT2 4k ------------ 30.7 ms (30.5-30.7-31.2-31.2-30.8-29.9)
NX1 4K ------------ 30.9 ms (30.6-31.6-31.4-30.7-30.2)
NX1 UHD ----------- 32.6 ms (32.9-32.0-32.9-32.5)
a7R II 4K S35 ----- 33.3 ms (35.6-32.2-32.8-35.3-32.6-31.1)
RX100 IV 4K ------- 36.6 ms (36.4-36.7) (without stabilization it's slightly faster: 35.7)
a6300 4K 24fps ---- 39.0 ms (39.2-38.9)

A few remarks

Don't think the number for the X-T2 is quite right, I'd say in my real world tests it comes out quite a bit better than the NX1 4K

30ms is where it gets bothersome, under that is fine for most stuff. Interesting that the A99 II 4K came out slightly worse than A7R II - it must be doing a better image though, more pixels in the readout.

Fastest stills camera sensors - GH5 and 1D X II... I think Canon could have given us better quality 120fps 1080p given how fast that chip is.

Good to see BM at the top - power hungry though.

Best value for money is GH5 and RX100 V.

Agree with this - I am really eyeing that GH5.  

15ms for me is pretty much as slow as I find acceptable.  Any slower definitely feels like jello.

[no_connection]

I want to know what the E-M10 MIII gets.

[Inazuma]

The XT2 numbers are taken from my shots. https://www.youtube.com/playlist?list=PLERrthnToa6Gws4tfM7VN0hd7jqmhKuKW

I have not done any measurements myself but it feels pretty similar to the G80 and NX1 from my experience. 

I would be interested in other XT2 owners shooting similar tests that can be measured. There may be some discrepancy between units. 

[essbe]

I think the measurement for the a99ii 4k refers to the super 35 mode, which makes it quite a bit faster than the a7rii, Samuel H (who added the figures) refers to the a99ii as much better than the a7rii in his post. Would be interesting to know the numbers for its FF 4k mode.

[Andrew Reid]

No that A99 II measurement will be the full frame pixel binning mode, the Super 35mm mode has worse rolling shutter, more towards 30ms-ish as with the A7R II Super 35mm mode, as it's a 5K full pixel readout. Pixel binning is faster. The more pixels you need to read, the slower it can scan the sensor from top to bottom.

1 hour ago, Inazuma said:

The XT2 numbers are taken from my shots. https://www.youtube.com/playlist?list=PLERrthnToa6Gws4tfM7VN0hd7jqmhKuKW

I have not done any measurements myself but it feels pretty similar to the G80 and NX1 from my experience. 

I would be interested in other XT2 owners shooting similar tests that can be measured. There may be some discrepancy between units. 

Nice. How did Sam do the maths? I take it accounts for different panning speeds and different focal lengths between camera test footage?

[Inazuma]

2 hours ago, Andrew Reid said:

No that A99 II measurement will be the full frame pixel binning mode, the Super 35mm mode has worse rolling shutter, more towards 30ms-ish as with the A7R II Super 35mm mode, as it's a 5K full pixel readout. Pixel binning is faster. The more pixels you need to read, the slower it can scan the sensor from top to bottom.

Nice. How did Sam do the maths? I take it accounts for different panning speeds and different focal lengths between camera test footage?

When I emailed him last year he said this:

There's a post about it somewhere in the thread. Focal length and angular velocity of the pan affect how fast stuff travels on the image, but my measure is relative so it doesn't depend on any of that. I measure the top-down skew, and put that in relation to the frame-to-frame displacement. I know how much time goes on between frames, so I can then use that to calculate how much time passed between the moments when the top and bottom of the frame where captured.

To clarify a bit: the key here is that focal length and angular velocity of the pan have exactly the same effect on skew as they have on frame-to-frame displacement.

[no_connection]

Focal length or pan speed does not really matter as long as you can see and measure the effect.

In short you have a vertical line like a lamp post, when you pan it at a constant speed you take the distance between two frames, then you check how much the bottom part of the line lags behind. Then you calculate the % of the distance that it lagged compared to distance moved. Then you take the frame time, roughly 42ms for 24fps. And you apply your % to that.

X-T2 would be about 72% lag if RS is 30ms. Which when you think about it it is quite a lot, the bottom part of the image is lagging a significant portion of the frame behind.
The a6300 pretty much lags an entire frame behind between top and bottom, which is why 30fps had to be cropped (only 33ms to work with)

[Andrew Reid]

Thanks for the good explanations!

[Dave Maze]

10 hours ago, Ed_David said:

I have always found this thread to be really important in deciding on using certain cameras inside vehicles, with fast motion:

http://www.dvxuser.com/V6/showthread.php?303559-Measuring-rolling-shutter-put-a-number-on-this-issue!

Thanks to Samuel

Going to post text as well, in case dvx100's forum goes under:

RX100 V 250 fps ---- 3.8 ms (3.8-3.8)
BM 4.6K (2K crop)--- 6.3 ms (official)
1DX II 1080p 60&120- 6.7 ms (6.6-6.7)
NX1 1080p ---------- 7.9 ms (7.7-8.0-8.1-7.8)
NX500 2.5k --------- 8.6 ms (8.2-8.8-8.9)
a99 II 1080p ------- 8.7 ms (8.8-8.7)
BM MicroStudio 4K--- 9.9 ms (official)
a7r II 1080p S35--- 10.5 ms (11.0-10.0)
a7r II 1080p FF---- 10.8 ms (11.0-10.5)
1DX II 1080p 24 --- 11.1 ms (11.1-11.2)
a6000 1080p24&30--- 12.2 ms (13.9-11.1-11.5)
BM 4.6K (4K crop)-- 12.6 ms (official)
BMMCC ------------- 13.3 ms (official) my measurement was 13.4 ms (13.7-13.5-12.9)
GH4 1080p --------- 13.7 ms (13.7-13.5-13.1-14.0-13.9-13.5-13.5-13.9)
RX100 V 1080p ----- 13.9 ms (14.4-14.0---14.2-13.1) (first 2 with stabilization, last 2 without)
D750 -------------- 14.5 ms (14.6-14.1-14.6-14.8)
1DX II 4k 24&60 --- 14.7 ms (14.6-14.9-14.9-14.6)
RX10 -------------- 14.8 ms (14.5-14.8-15.0)
GH5 --------------- 15.0 ms (official, preproduction)
BM 4.6K ----------- 15.2 ms (official) my measurement was 16.3 ms (16.3-15.9-16.8)
a6300 1080p24 ----- 15.2 ms (15.8-14.6)
GX85 1080p -------- 15.4 ms (13.8-16.7-15.7)
GH3 --------------- 15.5 ms (15.4-15.7-15.4)
G7 1080p ---------- 16.9 ms (17.5-16.3)
XT2 1080p --------- 16.9 ms (16.9-16.0-17.4-17.3)
RX100 IV 1080p ---- 16.9 ms (16.5-17.3) (without stabilization it's slightly faster: 16.1)
RX100 V 4k -------- 17.5 ms (18.1-16.3-17.7---17.7-16.9-18.5) (first 3 with stabilization, last 3 without)
BMPCC ------------- 17.7 ms (official) my measurement was 17.8 ms (17.7-17.7-20.0-20.0)
a7s APS-C 1080p --- 19.5 ms (20.3-18.4-19.5-19.5)
a7R II 4K FF ------ 19.9 ms (19.3-19.6-19.4-21.2)
5D3 --------------- 20.5 ms (20.7-20.5-20.4)
D5200 ------------- 22.4 ms (22.5-22.1-22.6)
GH4 4K/UHD -------- 22.0 ms (official) my measurement was 22.5 ms(23.2-22.3-22.7-22.7-22.4-21.8-22.4-22.8-22.6)
a99 II 4k --------- 23.4 ms (23.3-23.5)
BMC --------------- 23.6 ms (official) my measurement was 25.0 ms (26.7-24.8-23.5)
5D2 --------------- 25.9 ms (25.5-26.4-25.8)
5Dsr -------------- 27.7 ms (27.5-27.9-27.6)
G7 4k ------------- 28.1 ms (27.8-28.5)
GX85 4k ----------- 28.9 ms (30.2-29.0-28.8-27.5)
NEX-5N ------------ 29.4 ms (28.8-29.6-28.9-29.8-29.8-29.1-29.7)
a7s II FF 1080p --- 30.3 ms (29.5-31.2)
a7s II FF 4K ------ 30.4 ms (30.2-30.6)
a7s FF 1080p ------ 30.5 ms (30.1-32.0-30.5-30.3-29.2-30.9)
XT2 4k ------------ 30.7 ms (30.5-30.7-31.2-31.2-30.8-29.9)
NX1 4K ------------ 30.9 ms (30.6-31.6-31.4-30.7-30.2)
NX1 UHD ----------- 32.6 ms (32.9-32.0-32.9-32.5)
a7R II 4K S35 ----- 33.3 ms (35.6-32.2-32.8-35.3-32.6-31.1)
RX100 IV 4K ------- 36.6 ms (36.4-36.7) (without stabilization it's slightly faster: 35.7)
a6300 4K 24fps ---- 39.0 ms (39.2-38.9)

OTHER NUMBERS, MOST OF THEM FOUND BY SQUIG SOMEWHERE ON THE WEB:

Film --------------- 5 ms? (here)
Scarlet ----------- 14 ms
F65 --------------- 14 ms
AF100 ------------- 14.85 ms
FS100 ------------- 15 ms
FS7 --------------- 15 ms (here)
C300 -------------- 16 ms
Red One MX -------- 16.6 ms
7D ---------------- 21 ms
5D MKII ----------- 25 ms
GH1 --------------- 25 ms
D90 --------------- 33 ms
 

 

I am curious to see how the sony a9, a99, and canon 1dx ii stack up against each other in rolling shutter.

I get asked a lot to do crash cameras in cars - and in the past, I have used the a7s ii in 1080p mode.  But these guys now seem like a better option to prevent jello.

Same Ed...I have that forum post bookmarked and check it regularly! Thanks for sharing here..

[condra]

Anyone know where would one find rolling shutter info on the latest full frame models such as Nikon Z7 etc?
I'm really curious to see how Canon and Nikon fare against Sony in this particular area.

Perhaps it's a matter of waiting for user generated data?