Strobing / flicker effect when panning in 24p

[Alister Chapman]

But of course there will be a mixture of strobing and blurring in my test clips. I was using a 1/50th second (180 degree) shutter.

During the pan, while the shutter is open the camera moves a short distance so it captures an image that includes a lot of background that is blurred due to the motion while the car which is reasonably static in the frame is less blurred. Then the shutter is closed for the same period, meanwhile the camera is still moving. When the shutter re opens the camera is pointing in a slightly different direction so the background jumps while the car stays still. The effect is that the background appears to jump while the foreground car does not.

This is completely normal and totally expected from any sensor or medium exposed in this manner. If you have ever been to an IMAX show you'll know just how bad this can look if the pan speed is too fast.

There is nothing wrong with the way the EX1 captures motion. There are thousands of EX1's in daily use, yet there are less than a dozen (very vocal) people claiming that there is some kind of fault. There are even fewer examples of this "problem" being posted as clips. I think what we are seeing is people getting a camera that can shoot true progressive with a level of detail and resolution that really shows up the issues you get with shooting progressive at low frame rates. Then in many cases the material is being viewed on computers or monitors that add further issues and inconsistencies. This is then being viewed by people not used to seeing such images that don't understand the very nature of progressive and whole frame capture by video and film cameras.

Rolling shutter won't add strobing, it can add skew but only with very,very fast pans or very fast motion and even then it's only slight. Take a look at the lamp posts in my test clips, they don't suddenly lean as I pan.

My test footage was shot at 25p to ensure that there were no issues with pull up from 24p which is another stumbling block for many users not used to working with progressive. The difference between 24P and 25P is only 4% and just about impossible to see visually.

If you are going to do any tests I urge you to do side by side, like for like tests. I can make the output from my EX1 look ugly, I can also make it look beautiful. The difference between ugly and beautiful may just be a small difference in the speed of a pan or the focal length of the lens. Unless you do a like for like test how can you be sure that what you are seeing is down to the camera and not the user.

[Paul Curtis]

Steve,

Must also figure in that your 750 is 2/3rds so unless you match the aperture then you'll have less DOF which is a major way of hiding the effect in backgrounds. (even then DOF is a factor of lots of things not just sensor size).

Shame you don't have any of the offending footage to view, that would help enormously!

cheers
paul

[Steven Thomas]

Originally Posted by Steve Phillipps

because of the fair volume of people raising the issue I don't think it can just be disregarded.

Fair volume? Hmm.. with only a couple (who don't even own the camera) and out of the thousands who do....

It's not magic, it's only frame rate and shutter speed here.

[Dennis Joseph]

Just incase anyone thinks my settings are wrong, I am shooting 1080 24p Angle 180 , detail OFF and still getting judder. I put on a soft FX 3 filter yesterday to make it even softer and it still had judder which was shocking.

[Steven Thomas]

Post the footage...
It's hard to buy something like this. especially when there's no proof. Not to mention the majority of us use this camera (along with other cameras) professionally.

[Steve Phillipps]

Alister, Not all people who've seen this effect are not used to "a camera that can shoot true progressive with a level of detail and resolution". I've shot most of 'em, extensively, Varicam, Sony 750, Phantom HD etc. So it's not as if I or some of the others have just come up from a VX1000!
And I don't see it as helpful to say "less than a dozen (very vocal) people" either, very few people have an agenda here, we're just making comments and observations. Very often I've seen threads like this degenerate into insults and patronising posts (usually by those defending a piece of kit they own and have paid good money for).

Steve

[Alister Chapman]

But it is just a handful of people saying there is a problem out of the thousands of EX1 users and owners and they are being very vocal about it, not just here but in other forums as well, refusing to believe that there is no problem. Not one single person that claims there is an issue has been able to post a clip showing it, yet EX1 material must be some of the simplest material in the world to get off the camera and convert.

I have posted a clip that as far as I can tell clearly demonstrates that the EX1 handles motion no differently than any other progressive camera. Now it is being suggested that my test is flawed as both cameras exhibit stutter.
The F350 and EX1 are very different cameras and generate progressive images in very different ways yet they both look the same. What does that tell you? It tells me that the images are as they should be. I could repeat the test with my HD100 but why bother when there are some that clearly don't want to believe that the EX1 works as it should.

I don't understand why anyone expects there not to be judder or stutter when you are doing pans while shooting at 24 or 25P with a 1/50th shutter. That judder is why we have been using 50i or 60i for TV broadcasts for the past 40 years.

Yes I like my EX1 and as a user of one since January I am pretty sure I know what it can and can't do. I've shot almost exclusively using 25P and I've sold footage to Discovery HD, Nat Geo, NHK and others and no one has complained about judder or stutter.

I give up on this thread. No point in saying any more.

[Dominik Seibold]

Nobody of those claiming that there's a judder-problem were able so far to express that in objective measurable technical terms.
The simple question is:
What exactly does the ex1 wrong? How can I measure it?

[Paul Curtis]

I think we just need to await some example footage and then start the discussion again. The ideal would be a comparison of a camera that is said to perform correctly and the EX1 in the same situation.

You see i don't really understand what those that have problems are expecting to see. You open the shutter for X amount of time X times a second and the image is according to that.

The only thing i can think of is that stuttering sounds like a playback issue whereas strobing sounds like a recording one.

cheers
paul

[Dennis Joseph]

I will post something soon. In the mean time, what benefit does it do that I post a clip of what I am seeing if you guys can't see it from your own monitor. If you don't see it on your monitor with your clips, what makes you think that mines will be any different, unless mine is defective.

Long story short, it does not have the same motion blur when panning as a film camera. The judder looks nasty and I have several people that I showed my footage to that pointed it out. People who have never touched a camera with comments like "is that normal"?

I love the image quality of this camera, workflow, the feel, but the rolling shuuter and the flicker effect has to go.

[Steve Phillipps]

Well Dennis, that pretty much sums up what I think too.
Thing is Alister, when you say "I don't understand why anyone expects there not to be judder or stutter when you are doing pans while shooting at 24 or 25P with a 1/50th shutter", the answer is because you don't see it with other cameras. You see blurring yes (kinda gives you a nostalgic feeling for interlaced really!), but not stuttering. That's the difference that I was pointing out.
Also I think it's a bit odd to say that "EX1 material must be some of the simplest material in the world to get off the camera and convert", after all the reason I can't post it is that I can't get the bloody shots into mpeg or some form that I can post! Unlike many here I don't do my own editing, call me old fashioned but that's not my job, I just shoot the pictures.
I can put images onto my Macbook and convert them to mpeg4 from Panasonic HPX2100, HPX500, Canon XL, even Phantom HD, but couldn't do it from EX1!
Steve

[Craig Seeman]

Steve, I can at least make a suggestion to give you the ability to make your files generally viewable.

http://ffmpegX.com/download.html

The above shareware utility can take the EX1 MP4 and encode them to a variety of formats (H264 for example) that you can upload or otherwise view and share.

[Steve Phillipps]

Thanks Craig, unfortunately I've deleted my test files, didn't think I had any further use for them.
Steve

[Steve Connor]

Originally Posted by Dennis Joseph

Allister, thank you for that footage.
I refused to believe that 35mm film 24p is the same as video.

This I will agree with, IMHO ANY video 24 or 25P is not the same as film - it's a close approximation but the motion still looks different but that's true of all video cameras.

However compared other video cameras I see no difference in the EX1 progressive footage.

[Dave Elston]

Hi All,

Not wishing to add more noise to this already weighty thread, I simply want to suggest the following 'acid' test...

Dominik Seibold posted a few grabs and merged frames a few days ago which were taken from his EX1 on a spinning chair.

My idea is for a similar repeateable test, but instead place a mirror on the chair (or even better a turntable) with a distant point source light - now you can point two or more different cameras at the mirror and hit record to see how motion is rendered by the different hardware as the reflected light source passes across the frame.

I was a little confused why Dominik chose a shutter angle of 11.25° in his test (=1/768 shutter speed !?) but accept that that was his choice for whatever reason.
I have faked a 'normal' 180° motion blur (added crudely in photoshop) on his merged frames and would expect a very similar result to this from any progressive camera...

Anyone have the inclination (and access to a variety of cameras) to do this would help to settle this issue/thread pretty quickly IMO.

Just my 2p ;)

Cheers,
Dave.