Canon 1DXmkII 4K Video jello.
- Thread starter lay85
- Start date
But the problem we're experiencing is no panning, just basically stationary but bumpy. There is no scene moving across the frame or no object moving across the frame.
Anyway, I've sent Canon some sample footage today so should get an answer from them soon.
Anyway, I've sent Canon some sample footage today so should get an answer from them soon.
Upvote
0
So I spoke to Canon quite a bit yesterday, they have watched the sample video and are aware of the issue but dont have a definate answer yet but I have heard rumours of a software upgrade.
As far as the 200 dust spots on the sensor, they tell me the warranty is void if I get the sensor cleaned at a non Canon service centre which is a pain as I live 5,000km from the nearest one so I'd be 1 month without the camera for each time it's cleaned.
As far as the 200 dust spots on the sensor, they tell me the warranty is void if I get the sensor cleaned at a non Canon service centre which is a pain as I live 5,000km from the nearest one so I'd be 1 month without the camera for each time it's cleaned.
Upvote
0
Josh Denver said:
Would a solution to their RS problems be to change the firmware to readout for 24/30p at the same speed as 60p?
What is wrong with 60p look?
Upvote
0
From some other threads on CR unless they are really problematic just ignore them and remove them when occasionally they show up in blue sky etc. Ones that I fussed about seeing when I really looked for them generally don't show up in my shots so now I just don't fret.
Jack
Jack
Upvote
0
lay85 said:
In which case, unless you are using some form of camera stabilisation like a steadicam or gimbal, the bumpy up and down motion of being in the car is equivalent to quickly panning left and right which is what would cause the jello effect, just on a different axis.
Upvote
0
serenaur said:
Please excuse my lack of experience in this but does a camera that has IS built in function similarly to having a steadicam type device?
Jack
Upvote
0
It seems clear that lay85 has a problem in his camera, and it has nothing to do with operator error.
I tested my 1DX Mark II rolling shutter at 23.976FPS against my 5D Mark III, based on Josh Denver's post that the sensor speed could be 30ms. I recorded whip pans with the two cameras mounted on the same tripod. The 1DX Mark II appeared to be noticeably faster than the 5D Mark III shooting at 24P, giving vertical lines less of an angle of skew. For me, this supports the measurements by Cinema5d.com that the 1DX Mark II has a speed of 14ms to the 5D Mark III's 18ms. Changing from 60p to 24p does not appear to change the sensor speed.
As for Jack's question, the camera does not have any stabilization in the body, but is limited to stabilization offered in various Canon lenses.
I tested my 1DX Mark II rolling shutter at 23.976FPS against my 5D Mark III, based on Josh Denver's post that the sensor speed could be 30ms. I recorded whip pans with the two cameras mounted on the same tripod. The 1DX Mark II appeared to be noticeably faster than the 5D Mark III shooting at 24P, giving vertical lines less of an angle of skew. For me, this supports the measurements by Cinema5d.com that the 1DX Mark II has a speed of 14ms to the 5D Mark III's 18ms. Changing from 60p to 24p does not appear to change the sensor speed.
As for Jack's question, the camera does not have any stabilization in the body, but is limited to stabilization offered in various Canon lenses.
Upvote
0
lic4 said:
Small misunderstanding here; I'm referring to cameras that do have IS. That is, is it significant or is a gimbal type external stabilization required for all cameras, regardless?
Jack
Upvote
0
That actually contradicts with my findings. Mine is close to what lay85 has, unless you are saying that I have a faulty camera as well.lic4 said:
Let me be frank and be sorry if I don't trust much your opinion, with only 2 posts (really no offense intended) but I judge your opinion based on what I have. Unless you are Nero.
Please provide the link from cinema5d.com saying that 1DX II at 24p is 14ms. I must have missed it. My opinion, 14ms must be for the 60p. My camera says 24p 4K is worse than 5D III 1080P, which lay85's issue is.
Upvote
0
No offense taken. I signed up recently after reading the post about the rolling shutter, because I had just invested in the camera and was alarmed by what I read on this board.
After the camera arrived today, I performed the tests with both cameras tethered together, shooting 4K 24p on the 1DX Mark II and 1080 24p on the 5d Mark III. I'll have to edit the footage together and grab the frames again, because I stupidly forgot to save the project after I saw the evidence.
In my camera, the 1DXii exhibited a little less skew to vertical lines than the 5D3.
Here is the link showing Cinema5D.com's word on the sensor speed:
https://www.cinema5d.com/canon-1d-x-mark-ii-vs-canon-1d-c-which-one-shoots-better-video/
(It's under the heading "rolling shutter")
And this link shows a chart with the measurement for the 5D Mark III:
https://www.cinema5d.com/rolling-shutter-sony-a7s-vs-arri-amira-canon-c300-5d-mark-iii-1dc-panasonic-gh4/
In the comments of this article, a DPReview administrator says that he sees the same sensor speed in 60P and 30P:
https://www.dpreview.com/news/7057004492/don-t-get-ahead-of-yourself-canon-eos-5d-mark-iv-rolling-shutter-test
It's difficult to test the cameras 60P vs 24P because I don't have something moving at a consistent speed to capture on my one camera, but I was satisfied with the comparison to the 5D Mark III. When I get some time, I'll run a comparison of the 1DX Mark II at 60P to the 5D and post those here as well.
After the camera arrived today, I performed the tests with both cameras tethered together, shooting 4K 24p on the 1DX Mark II and 1080 24p on the 5d Mark III. I'll have to edit the footage together and grab the frames again, because I stupidly forgot to save the project after I saw the evidence.
In my camera, the 1DXii exhibited a little less skew to vertical lines than the 5D3.
Here is the link showing Cinema5D.com's word on the sensor speed:
https://www.cinema5d.com/canon-1d-x-mark-ii-vs-canon-1d-c-which-one-shoots-better-video/
(It's under the heading "rolling shutter")
And this link shows a chart with the measurement for the 5D Mark III:
https://www.cinema5d.com/rolling-shutter-sony-a7s-vs-arri-amira-canon-c300-5d-mark-iii-1dc-panasonic-gh4/
In the comments of this article, a DPReview administrator says that he sees the same sensor speed in 60P and 30P:
https://www.dpreview.com/news/7057004492/don-t-get-ahead-of-yourself-canon-eos-5d-mark-iv-rolling-shutter-test
It's difficult to test the cameras 60P vs 24P because I don't have something moving at a consistent speed to capture on my one camera, but I was satisfied with the comparison to the 5D Mark III. When I get some time, I'll run a comparison of the 1DX Mark II at 60P to the 5D and post those here as well.
Upvote
0
Still, I can't find in any of these links anything saying that 4K 24P 1DX II is at 14ms. And that is what I am challenging, because my camera can't do that. 60P is a totally different matter.
The first week I got mine (4-5 months ago) I shot a rehearsal at 1080P because my concern was if I could get the singer in focus at 85mm 1.4 using the DPAF. I was very pleased with the result.
I bought bigger cards and a week later I shot the whole concert at 24p 4K and 3 hour of footage was useless. Luckily, my B camera was a 5D III at 1080P 24P and I used the footage from it.
I learned it the hard way.
Ever since I shoot at 30P, from time to time at 60P in between shots, and if by mistake I rolled the dial too much and I am at 24p it takes 10sec and I will recognize it right away.
Just shoot someone's close up on the street and you will see it.
I assume that you know by heart the look of your 5D III, so judge it by yourself.
The first week I got mine (4-5 months ago) I shot a rehearsal at 1080P because my concern was if I could get the singer in focus at 85mm 1.4 using the DPAF. I was very pleased with the result.
I bought bigger cards and a week later I shot the whole concert at 24p 4K and 3 hour of footage was useless. Luckily, my B camera was a 5D III at 1080P 24P and I used the footage from it.
I learned it the hard way.
Ever since I shoot at 30P, from time to time at 60P in between shots, and if by mistake I rolled the dial too much and I am at 24p it takes 10sec and I will recognize it right away.
Just shoot someone's close up on the street and you will see it.
I assume that you know by heart the look of your 5D III, so judge it by yourself.
Upvote
0
Jack Douglas said:
No, it's more like shooting video with an IS lens on a body without in-body stabilisation. I use a Sony A7SII which has IBIS with a Metabones adapter and my Canon glass for work and from my limited experience of hand held shooting (I shoot video for work almost exclusively on a tripod or slider), I've found it to basically be like adding IS to my non-IS lenses. I also have a mechanical glidecam clone for personal use, though I have barely had a chance to use it since I got it as I have been so busy shooting for work. The times I have managed to briefly get out with it have shown that IBIS and gimbal stabilisation are not even comparable. Steadicam/gimbal stuff is just that much better *as long as the operator is skilled,* which I am not!
Saying that, the recent leaps in combining IBIS and lens stabilisation seems to be yielding good results in certain situations. I can't find it now (I just spent about 30 mins looking through my internet history for it!), but a fairly recent video I watched of what I believe was an Olympus E-M1 mkII matched with an Olympus stabilised lens showed impressive results for casual use. If I'm able to find the video again I will update my post, though it may have been a different camera which might be why I can't find it.
@OP - sorry for the thread hijack! I am weighing up a 1DXII purchase with some video use in mind so am watching related threads quite closely at the moment.
Upvote
0
serenaur said:
Thanks very much for this. I'm learning more each day. I'm guessing the OP is stuck with trying to get answers from Canon and I'll be watching for more on that.
Jack
Upvote
0
**edited with an additional explanation of results.
This is a test using a opening and closing door with cameras positioned one over the other at the same vertical angle. Please forgive me for doing this so crudely; it isn't pretty, but it's good enough for me.
The six images show the results I found. The first four are images at 23.976fps, 4K MJPG on the 1DX Mark II next to 23.976 1080 All-I on the Canon 5D Mark III. I included a red line from the 5D III next to the 1DX to show the difference.
The final two images are comparisons with the 1DX Mark II at 60p. The relative difference and improvement in rolling shutter to the 5D Mark III appears to be the same either at 24p or 60p. I was surprised to see that the difference to the 5D is not that much.
**the difference in slope is more pronounced at times because I was not opening and closing the door each time at the same speed, and I tended to swing it faster in one direction due to my positioning. --thanks for understanding!
This is a test using a opening and closing door with cameras positioned one over the other at the same vertical angle. Please forgive me for doing this so crudely; it isn't pretty, but it's good enough for me.
The six images show the results I found. The first four are images at 23.976fps, 4K MJPG on the 1DX Mark II next to 23.976 1080 All-I on the Canon 5D Mark III. I included a red line from the 5D III next to the 1DX to show the difference.
The final two images are comparisons with the 1DX Mark II at 60p. The relative difference and improvement in rolling shutter to the 5D Mark III appears to be the same either at 24p or 60p. I was surprised to see that the difference to the 5D is not that much.
**the difference in slope is more pronounced at times because I was not opening and closing the door each time at the same speed, and I tended to swing it faster in one direction due to my positioning. --thanks for understanding!
Attachments
Upvote
0
I'm confused why the sensor readout time would change at all when switching from 24p/30p/60p.
I would assume the sensor readout time would be constant at the given resolution (1080p or 4k) and the change would be the interval of how often a readout is taken (to determine 24p/30p/60p). Is this not the case?
I would assume the sensor readout time would be constant at the given resolution (1080p or 4k) and the change would be the interval of how often a readout is taken (to determine 24p/30p/60p). Is this not the case?
Upvote
0
It's now been over a month since I contacted Canon about this and still have no explanation or reply since 5th December.
I did however get the sensor cleaned and wow, what a difference. I had been a bit unhappy with the still image quality but now they look quite a bit better and sharper but the camera did arrive with so much dust on the sensor, worse I've ever seen in 10 years.
RE: The video issue again, we are using one of the best stabilisation systems ( 2x kenyon gyro) and have been for 6 years with no issue and the 1920 x 1080 footage is super smooth so just the 4k which is crap.
I did however get the sensor cleaned and wow, what a difference. I had been a bit unhappy with the still image quality but now they look quite a bit better and sharper but the camera did arrive with so much dust on the sensor, worse I've ever seen in 10 years.
RE: The video issue again, we are using one of the best stabilisation systems ( 2x kenyon gyro) and have been for 6 years with no issue and the 1920 x 1080 footage is super smooth so just the 4k which is crap.
Upvote
0
