I have an EX1 for a weekend. Anyone want me to run tests?

actually the skew and wobble are both manifestations of rolling shutter distortion. if you get skew you'll also get what people are calling wobble, which is just rapid, short changes in acceleration causing the rolling shutter to distort the image alternately in one direction then the other. it looks more jello-like if you do it up and down, since that is the same direction of the rolling shutter so the picture squishes and unsquishes.

you can easily cause this distortion in a camera with a rolling shutter: just zoom to the longest focal length and jiggle the camera up and down. i have ex1 footage i shot yesterday that looks comically bad when i did that *intentionally*.

the thing is, under what real-world conditions do you need to do that? i see rolling shutter distortion in real world ex1 footage, but you really won't notice it unless you've got a shot that needs to be shakey.


Alex, I think you have a couple of your terms mixed up. The partial exposure is right and is easy to see in the pictures that you provided. Thanks for that. The skew problem(which you are calling wobble) is when you pan side to side and it causes things to look skewed or like they are leaning. If that footage you showed on Youtube is from the EX, then yes, it definitely still exhibits skew. Wobble has to do with vibrations, mainly in the tilt directions up and down. With a CMOS that is not corrected like the HV20, this can cause the footage to "wobble" literally like a bowl full of jello and is quite disturbing and a definite deal-breaker.

The amount of skew in your footage does not bother me because I will never be doing whip pans that fast. Wobble due to vibrations(walking, on vehicles, helicopters, etc) is where my main concern lies and so far nobody has mentioned it yet. Thanks for posting these clips and footage Alex, very much appreciated.
 
Haha, wow, I go to work for a few hours and this is what happens. Hmmm, intriguing.

Alright, so everyone is saying that this is an HV20? Well, I do have the raw MP4s from the camera sitting right here. I'll try to outline the conditions as best as possible for both the skew and the partial exposure.

When I tested for skew I was shooting 1080 24p with the shutter angle @ 180 degrees. I had it mounted in studio on a pretty beat up boegen 501 head and the lightest pair of sticks around. What I did is aim the EX at something thin and bright, then I zoomed in to about 40mm and panned the camera from left to right. It really wasn't that hard, and you can see the results now. I don't have any webspace to host the raw clip, but if anyone out there can I'll give them the MP4. I don't know if that'll make a difference though, because I'm sure people will just claim that I converted my "HV20" footage or some bullshit like that.

The flash exposure is a really weird case, and I'll explain it here. We decided that, as a crew, we'd try to fit the EX into our live to tape workflow so we brought it to a basketball game on friday night. The photographers from the campus newspaper must have been there too, because some guy was mounting four flash bulbs around the court: one on the backs of either set of bleachers (one was just left of the camera) and there were also bulbs on either end of the court. They were all synced and went off at the same time. I just happened to be lucky enough to get one on camera. I also have the raw MP4 for that too. For that shoot we were using 1080 60i with the shutter off, although in retrospect I should've shot 720 60p. My bad I guess.

So yeah, I guess I'm a liar or something? Seems kinda weird. I dunno, I guess maybe I thought people would think to take this at face value, but the internet is always full of surprises.

P.S. Everything besides the recording format was set to factory default by the way. Only thing I changed was the video out and I added the 4:3 markers in the LCD.
 
Last edited:
Probably because you want to. No offense.


Nope... just heard the rattle.


What is rattling on your EX1.

Maybe a sling on the Tri Pod or maybe a cable?

It sounded heavier but I don't know.

That's why I asked.


You said when you got the wobble you shot anything slim or bright.

Does the wobble show up under any other conditions?

Do you think this is a show stopper or can you live with it?
 
Its not that we doubt you, Alex. We just need confirmation is all. If someone could copy his results then it would help alot. From what i saw in that vid, though, those pans were really really fast. I doubt even the red could handle those pan speeds lol.
 
Nope... just heard the rattle.


What is rattling on your EX1.

Maybe a sling on the Tri Pod or maybe a cable?

It sounded heavier but I don't know.

That's why I asked.


You said when you got the wobble you shot anything slim or bright.

Does the wobble show up under any other conditions?

Do you think this is a show stopper or can you live with it?

Man, I dunno about the rattle. Could be the crappy old 'pod, could be a loose battery on the camera. Truth be told I wasn't paying too much attention to the audio.

I saw skew on just about every distinct line in the frame if I panned fast enough, even when I was shooing as wide as I could. That said, it wasn't really something I saw all the time because I had to be panning pretty fast for it to happen. For instance, when we were shooting the game from the stands I didn't notice it on the lines painted on the court floor, but I did notice it when I was whipping it around in the studio. See, I was pushing things to the limit. This video was one of those stress tests designed to coax some skew from the rolling shutter, and it did. Coincidentally, the partial exposure just happened when the camera was put right next to some unusually strong flashes of light.

Is it a show stopper? Well, I doubt I'd shoot with it in a moving car, but other than that I really didn't think that it was an unworkable machine. I liked the dof, the display had all the information I could have ever wanted, I love the SxS workflow, and the manual lens is exactly what I've been waiting for. I'm not in a rush to get an HD camera though, and the EX doesn't change that. I'm going to wait for a while to see what is going to be released next, because this camera definitely ups the ante on a lot of stuff out there.
 
Last edited:
Alex I assumed that that was NOT your youtube video. Just something you found on the 'tube. If that was yours then i take back the bogus remark. I'm still amazed no one else has seen that wobble. Maybe the extreme back and forth is not a practical test yet Adam Wilt earthquaked the cam and never mentioned this. Most have said the wobble was minor. That youtube was not minor.

Was this shot at the 35mbs mode or 25? Could that be the factor?

Is it possible for you to do the same test with another cam to see if its just yours?

Not shooting in a car? Not shooting with strobe or flashes? I mean whats the point then if you're limited this way. Someone did post at another site with some fast jerky handheld and it did not show any visible wobble. The camera mounted in the plane on Sonys site shows no issues yet in a plane and looking at the sky there are not many vertical objects but still no sky skew.

Someone needs to take this cam out and really make it work.

For me car shots, mounted on bikes with quick pans will be the determining factor. Hope someone does those shots soon.
 
Last edited:
Alex, If that's what you captured with the EX1, then so be it.

I guess there's trade-offs with everything. I have no intentions on fast pans, especially at 24P When used as an effect, it would be just that.

Now if this was happening under normal image movement, or normal dolly moves, I'd be REAL concerened. Based on my testing, it's not an issue for me.

If this is an issue, I'd suggest staying away from the EX1 !
 
Not shooting in a car? Not shooting with strobe or flashes? I mean whats the point then if you're limited this way.
That's the question, and that was the point of my article -- to point out that new technologies are changing the rules of what kind of shots we can get, and people need to be aware of this. The other issue is computer graphics and match-moving.

I'm not saying an EX1 is unsuitable for these purposes (yet, as I haven't tested it); what I am saying is that the presence of a rolling shutter *raises the possibility* and a prudent shopper should test the particular camera they're interested in to see if it is indeed fit for their intended purpose. You can no longer take for granted that they all perform basically the same (with respect to sensor characteristics or recording format) because they're not made the same anymore. It used to be that everyone used CCD and intraframe compression at 60i. Now we have rolling-shutter CMOS and long-GOP at varying bitrates and GOP sizes and frame rates. And some of these technologies will render certain types of shots either difficult or impossible to get. So buyer beware, and educate yourself.

The camera mounted in the plane on Sonys site shows no issues yet in a plane
Hmmm, imagine that. I bet the footage on JVC's site doesn't show any split-screen either... :D

Someone needs to take this cam out and really make it work.
That's what will happen over the next few months. And as that happens, the bloom will fade off the rose and we'll start to see what it's really capable of. It always happens, every product goes through this phase. The Red's starting to get its share; there's apparently quite a bit of noise in the blue channel and it's causing some issues and users are starting to talk about it. You've only ever heard heaping praise about the Red, but now that users are getting it and using it, you'll start hearing about issues. The same thing will happen with the EX1 as more and more users get it and start putting it in different circumstances.

What you have to determine is how well it will work (or even whether it will work) for YOUR circumstances.

For me car shots, mounted on bikes with quick pans will be the determining factor. Hope someone does those shots soon.
Kaku Ito usually puts every new camera to the test, and he does mountain biking, and he's already planning on doing some mounted-to-a-bike shots.

But seriously, the results of those shots will likely only be of interest to someone who does mounted-to-a-bike shots, right? Even if it fails disastrously on shots like that, if someone else will be using it for sit-down interviews it could still be a perfect cam for them, y'know?
 
Good points. I can appreciate the truth of what Barry's saying. Each new "breakthrough" product eventually gets put through the wringer, and flaws and gotchas float to the surface.

CMOS in professional video cameras is in its infancy, and it's the pioneers who'll have to be dodging arrows. EX1 users are the testbed for what will probably be assessed and improved upon in future product offerings.

It's all about prudent shopping and assessing the tradeoffs. As of this day, there's still no perfect budget HD camera...but we're getting closer.
 
Last edited:
We're going to be hosting some raw mp4's from Alex (we meaning myself and a few other folks) if Alex wants to do that. Once we have a link I'll pass it along to him and he can feel free to post it everywhere.

Laer,
Jason
 
This video was one of those stress tests designed to coax some skew from the rolling shutter, and it did. Coincidentally, the partial exposure just happened when the camera was put right next to some unusually strong flashes of light.

you'll get partial exposure on EVERY flash shot with a rolling shutter unless the flash starts and ends exactly at the top and bottom of a shutter scan, which it rarely does. i have footage with maybe a dozen flashes, and every one is exposed across two frames. i don't see what the problem is. it's just a flash!

people really need to rent this camera and try it out in some real world situations. unless you're shooting intentionally shakey footage you won't notice the skew. this is a fantastic camera; the rolling shutter is just a feature :) that you learn to shoot around if you need to.
 
I would still like to know if that youtube was shot at the lesser 25 Mbs or the 35Mbs VBR, since the camera does both, and if that is a factor in wobble and skew.
 
Bitrate of the recording codec would have no effect on the amount of wobbling or skewing that goes on. You can see the wobbling/skewing live, when the codec hasn't even touched the footage.

Wobble/skew/partial exposure, etc., are all characteristics of the sensor, not the compression format.
 
I think there will be two videos uploaded. there is one right now... Unfortunately, I can't view it.

But, here is the link to download it.
www.dvxfest.com/samples

too bad I can't see what all the fuss is about since I can't view the video. I guess I need some other codec :)

Later,
Jason
 
Barry,

Are you planning on a hands on review of the EX1?

If so, that would be awesome as I highly value your opinion on these things.

As Always, Thanks sharing your expertise!
 
CMOS in professional video cameras is in its infancy, and it's the pioneers who'll have to be dodging arrows. EX1 users are the testbed for what will probably be assessed and improved upon in future product offerings.

Actually, CMOS has been used in broadcast cameras longer than most people realize. Ikegami started using them back in 2004 with the HDK-79EC multi-format camera. This camera has full raster 2.5 megapixel sensors and it's still used today by mobile production trucks to shoot sporting events like the US Open, NASCAR and from the Goodyear Blimp. They also make the HDL-40HS, which is a high-speed 720/120p super-slow mo camera and it's based on CMOS.

Thomson Grass Valley developed their own CMOS and Vision Research already uses them in their Phantom HD, Digital Cinema and super high speed (1000 fps) cameras. Even ARRI uses CMOS in the D-20. You've got to wonder if all these companies are on to something with this CMOS thingie. The only holdouts have been Panasonic and JVC, although JVC doesn't make very many cameras. But someone on another forum did say a Sony engineer at a trade show said they would keep using CCDs for their high-end cameras.

By the way, it is possible to have a CMOS sensor with a global shutter. Apparently this is common in industrial cameras for machine vision. I'm sure there must be trade-offs though...probably one of them being higher cost because the pixel structure is more complicated.
 
Last edited:
Back
Top