Got a chance to put EX1 up against HVX

Barry_Green

Moderator
I talked someone into letting me come over and work with their EX1 for a day and a half. Brought some charts, some lights, etc. My main goal was *not* to do some exhaustive camera comparison shootout. My main goal was to find out what new things the EX1 brings to the table, and find out what it's like to work with. I've had a tapeless HD camcorder for two years now -- how does the EX1 compare? I also had an HVX so I had 'em side by side.

I'm not going to post a bunch of stills because every time I do that, people just look at the pictures, and then they start calling me names and flamefests and stupidity ensue. So forget that. Plus, still frames will only show you still frames, and the EX1 *excels* at still frames, so what would be the point in showing us what we already know? We already know the EX1 is noticeably sharper than the HVX. What I wanted to know is: what does it DO? How does it do it? And what differences will it make to an actual shooter? And is it time to trade in the HVX?

So the first thing I looked at was sharpness. And oh yeah, the EX1 delivers on that. 800 observable lines of horizontal res on a Chroma Du Monde. Fantastic. Er, just don't move the camera. Move it, and you get a noticeable resolution drop, more than can be accounted for through simple motion blur. And it's not due to the codec either, you can see the effect on the live output and you can see it on a scope. Heck, you don't even need that -- you can see it right on the (as already acknowledged, excellent) LCD! Just turn on peaking (especially in yellow -- excellent feature, that peaking-in-yellow!) and point at a high-detail scene; take it outside, put it on wide-angle and point at some trees. You should see a whole lotta yellow, because the peaking is showing all the sharp high-frequency detail, right? Now pan, and -- whoops? Where'd all the yellow go? You'll notice the peaking disappears whenever you move, and comes back when you stop. Well, so does your hyper-sharp resolution -- it all disappears whenever you move. The faster you move, the more it drops, and that may be attributable to motion blur, but even very minor moves cost you a couple hundred lines of res. So in motion, the sharpness advantage over an HVX is very little. I have theories as to why (I think it's doing some manner of aggressive noise reduction, perhaps through frame accumulation) but I don't know for sure and I couldn't find any other way to verify my suspicion.

Sensitivity -- yep, the EX1 is more sensitive. How much? Depends on your tolerance for gain on the Sony, but overall, not a tremendous amount. It's about 1/2 stop more sensitive. I think I know why Adam Wilt said the camera is 320 ASA -- because it pretty much is. I didn't have my light meter with me, but based on what I know about the HVX I'd rate the EX1 more around 500, which is within about 1/2 stop of 320. Adam rated 320 off a pre-production model, so maybe the final model is 500? Anyway, what I tested is about 1/2 stop more sensitive than the HVX. Here's how you know -- set up a chart and set both cameras to render the image very comparably -- same gamma (HD NORM on HVX, I think I used Standard 3 on the EX1), same color matrix (normal on EX1, NORM on HVX), same recording format (1080/24p on both) and same shutter speed (180 degrees, or 1/48). Now set the iris so that you're getting exactly 50 IRE on both cameras, and then look at your f-stop. The HVX was at precisely f/2.8, the EX1 was at precisely f/3.4 That's 1/2 stop, folks. Not the outrageous claims I've heard elsewhere of "three stops faster" and such! (and to anybody who wants to dispute this, just test it! I've told you the exact procedure. Set 'em so they render the image comparably, set the gamma similarly, and adjust the iris to get equal brightness. It's a half stop different.)

Now, the thing is, the Sony is quite a bit cleaner in noise, at least at low gain levels. You could go to 6dB of gain and the noise is about the same as the HVX, maybe a tad cleaner. Maybe you could go to 9dB of gain, although that might be pushing it. At 6dB of gain, the "effective" sensitivity improvement over the HVX is 1.5 stops. That's getting somewhere, that's noticeable. But you need to use gain to do it, so -- decide for yourself.

Okay, noise -- at 0dB, no doubt the EX1 is cleaner. Combined with the sharpness, the clean image really gives you that "looking through a window" feeling. That's very cool, and may be exactly what some people are looking for -- if that's what you want, the EX1 delivers. Doesn't look anything like the film look, of course, but it does look like that super-sharp looking-through-a-window HD look that you see on DiscoveryHD, etc, and that's a very valid look. If you add gain it gets noisy, so don't do that... I mean, if you were not in your right mind and decided to go to 18dB of gain, it's incredibly noisy (looks about the same as an HVX at 18dB) but it also robs the image of a lot of detail. Resolution goes out the window as the noise comes in, sort of like the "oil paint" effect. So don't use a lot of gain. Didn't see any bad effects at 6dB, so I think that's probably just fine to use, but stay away from much higher.

There were many things about the EX1 that were positives. The LCD is very sharp and the colored peaking is wonderful. The sharpness is great. The small size is nice. The delete-last-clip function is nice (although it's also MANDATORY on this SxS system, more on that later.)

But there were lots of things that just flat-out suck too. How the **** are you supposed to mount a mattebox to this thing? Look at the onboard mic, it sticks out past the front of the lens by about two inches! I tried to put my Vocas on it, I couldn't even get it near the lens, and forget about trying to use 4x4 filters. The only way you could do it is to load from the side maybe? If anyone can use a conventional top-loading mattebox I'd like to know which brand. I've had Chrosziel, CAVision, and now Vocas, and I don't see how any of 'em could work with this EX1.

Secondly, it's impossible to hand-hold one-handed. You just can't do it. The fatboy HVX feels as light as an HV20, compared to the EX1. The ergonomics are astoundingly bad for handheld. I can't describe it to do it justice, you just have to experience it for yourself. You can manage okay with two hands (and I use the HVX with two hands), but just get used to the idea that you'll have to use two hands with the EX1.

Okay, now onto my main gripe: SxS SuXS! It was infuriating to use; for anyone used to P2 I don't know how you could stand this. It's like they totally missed the point on what solid state is supposed to be. It's so slow to actually work with the system, it's like using a tape camera, only worse. Okay, look, I've gone around the globe teaching people about the HVX and P2, so I'm used to talking to a lot of shooters and showing them how to use the system. One thing I do is I show how quickly you can go check your footage -- just pop into playback mode and play your clips. Takes about a second, maybe two. On the EX1 the same process takes 14 SECONDS. Count that out, folks. Think about it. "Let's play back that clip", the director says. Okay, count it out: one one thousand, two one thousand, three one thousand... all the way to 14 one thousand. Now, the camera's rebooted into playback mode and you can play a clip. 14 seconds? On a tape camera like the DVX I could toggle over, REWIND THE TAPE, and be viewing footage in less than 14 seconds!

Okay, another thing I point out in training seminars is that even in the middle of playing back HVX footage, if something started happening that you need to record, you can just punch the record button and the camera will automatically jump into camera mode and start recording. Takes maybe 2 seconds. This is part of the revolution of tapeless, this is part of what makes going solid-state so darn cool, right? I mean, if you're fooling around showing off some footage outside the courtroom and suddenly the doors fly open and the defendant is marched out, you can punch "record" and by the time you point the camera at the scene, it's recording, right? On the EX1? Not possible. It takes about 12 seconds to go from playback mode back to record mode. Again, that's crazy. 12 seconds? The defendant would already be in the cop car and driven away in 12 seconds! How are people supposed to use this for ENG? You're setting yourself up to get screwed.

Okay, here's where it gets worse -- let's say that you decided to show your friend a clip so you push to "playback" mode, and three seconds later the door bursts open. The camera's barely started booting up into playback mode, so you push the switch to camera mode -- guess what? You're out of the game for a full 25 seconds. It has to *finish* booting up in playback mode before it'll even register that you swapped to camera mode! So, you've waited three seconds, now you have to wait 11 more seconds for it to finish booting up, and then you have to wait another 11 seconds for it to re-boot in camera mode. 25 seconds, round trip. Ridiculous. 25 seconds is a long, long time to not be able to shoot. Makes that 7-second tape spool-up time seem instantaneous, doesn't it?

Granted I was swapping back and forth a lot in my testing, so I experienced these delays proportionally more than an average shooter might, so it irritated me more than it may irritate you. But I wanted to drop-kick the thing after a while, and I seriously don't know how an ENG/news guy would put up with it. When you need to go, you need to GO, not sit around waiting for the camera to dawdle around for up to 25 seconds...

Okay, more on SxS: it took 4 minutes and 20 seconds to offload an 8GB card. An 8GB P2 card takes 4:25 on this same laptop (Sony laptop with both ExpressCard and PCMCIA slots). So any claims of how the ExpressCard SxS system is "so much faster" are just marketing nonsense.

Okay, what else... oh, the lens. Yes it has true manual zoom. Felt exactly like the HVX's. Yes it has manual focus. Other than hard stops it didn't do anything the HVX doesn't. Feels a little different, but gives the exact same result (but with less rotation). Iris ring was nice though, I do like that, everyone should have a proper iris ring. Saw plenty of chromatic aberrations, a tad more than the HVX (yes, more), but way less than the Canon or JVC. I'd rank C.A. performance as average. Interesting to note, practically the whole camera is lens; they have a film plane marker on it that's almost all the way back to the cooling fan slot. What looks like the camera body is actually all lens inside (or so I think). The fat butt on the end is probably where all the recording unit is, the rest (in front of that slot) seems to be basically all lens.

Rolling shutter? Oh yes. Absolutely it has rolling shutter issues. People who say it doesn't have it either a) don't know what they're talking about, or b) haven't encountered it, but it's there in full force. The EX1 has a full complement of rolling shutter issues; it does skew and it does wobble and it does partial exposure.

The menu system is an interesting mix of give 'n' take. Very extensive menus with wonderful controls, seemed about comparable to the Canon XHA1/XLH1 in that respect. But navigating the menus is annoying because the little spinny wheel sometimes goes up when you tell it to go down, and the menus don't loop through -- you can't roll off the bottom and back around to the top like you can on the HVX. DVX was that way and it was annoying; they fixed it in the HVX and I really missed it on the EX1. Level meters didn't have any witness marks at -12 and -20, that was annoying. The LCD is sharp and clear, but it also gets very cluttered with all the stuff that gets displayed; I appreciated the HVX's method of pushing that stuff outside the visible frame. Wish I could have both -- a 4:3 LCD so the "clutter" goes outside the 16:9 frame, but with the sharpness of the EX1's LCD. About that "clutter": with the EX1 they have a "direct" menu where instead of going into the menus through the "menu" button, you use a miniature joystick to navigate around the on-screen displays -- when you click 'em it becomes a menu. I thought that was kind of clever, but some functions that should be on switches, aren't (like shutter speed? Hello?) -- you have to change those things through the "direct" menu. So it's got some give and take, some things are definitely better, some things are a step backwards.
 
(PART II)

As for the recording format: I didn't try to "break" the codec. It looks good on the still shots, but I was only really doing lab tests and the recorded images looked great in the lab. Someday I'll put it through some real-world shooting scenarios and see how it holds up. I suspect it won't hold up as well as XDCAM HD, because it has to handle 33% more data in the same bandwidth, but I'd like to find out for sure.

Overall: if you want that looking-through-a-window grainless HD look, the EX1 delivers that, and it does so with a 1/2 stop more sensitivity too. There's no free lunch though, and it comes with a lot of penalties -- rolling shutter bullcrap, intolerable handhold ergonomics, and the goofy SuXS workflow (man I hated that... it's like amateur hour for tapeless... there's NO METADATA! No information about the clips at all, no text memos, no serial numbers or data tracking... the entirety of the metadata is a shot marker or two! SHOT MARKERS. Hello, we had shot marks on TAPE. This is supposed to be tapeless solid state, get with the program folks!)

I can't explain why they limited their tapeless system so much. Only thing I can think of is they're trying to replicate the XDCAM disc workflow as closely as possible? If so, it's stupid -- they walked away from so much of the potential of what solid state recording can do. Is SxS better than tape? Yes. Is it better than disc? Well, not really -- it's more of an either-or choice; you get the same workflow as disc but in a smaller package -- is that meant to be appealing? I couldn't even edit directly! I had to "export" the MP4 files to "MXF" files before I could even bring 'em into Sony Vegas (granted, I'm using version 7, maybe that's been fixed in version 8?) What's the point of having a super-fast medium if you can't even edit from it? Had to do the same kind of thing with FCP. Goofy. Marcus should make a "Raylight" for SxS.

I dunno, folks, I mean, I can see some definite advantages (sharpness & noise mainly) but the tapeless SxS just comes off looking half-assed compared to the P2 workflow (forgive the crass term). I just don't know what clients I would recommend this product to. Interviews? Yes, definitely. ENG? No way. Weddings? Absolutely not. Film production? Maybe. People who actually LIKE XDCAM-HD discs? Sure, they'll probably love it.

I'm hoping to get some more time with it later in the month where I can put it through some more tests. Right now I'd say it's six of one, half dozen of the other. I think the HVX is a lot more workable and usable, and it's a lot less expensive. The EX1 is sharper, cleaner and more sensitive, but it's got some issues and working with it was a real pain in the patoot (comparatively; I guess if you've been using XDCAM discs then maybe this EX1 will seem sent straight from heaven but if you're used to P2 I think it'd be tough to swallow the limitations).
 
Hey Barry! I don't know if you have an answer for this after the type of testing you did... but you specify that camera motion causes a drop in sharpness when the camera moves; how about when there's moderate motion in the frame, but the camera's still? And how bad are the rolling-shutter artifacts in high-motion scenes with a still camera?
 
Thanks Barry, very interesting.

I don't understand how it can't have metadata though. Wtf? by the way I laughed when I saw you swear, albeit it was in asterixs. lol Could metadata stuff be added with a firmware update? (maybe the rushed the release, just guessing)

The stuff about losing res when moving the cam was interesting too, would like to see some footage.

The thing that first astounded me about the cam was how the hell you would mount a matte box? Kind of like having a house with 3 walls.

Looking forward to a more detailed write up from you.... no pressure.
 
Last edited:
Nice write up as always.

One question - what's the focusing like on the camera?
You mentioned how much crisper the LCD is than the HVX, but I'm curious as to any 'focus assist' type of features.
Or is the LCD so crisp that focus isn't an issue?
 
Nice write up as always.

One question - what's the focusing like on the camera?
You mentioned how much crisper the LCD is than the HVX, but I'm curious as to any 'focus assist' type of features.
Or is the LCD so crisp that focus isn't an issue?

I believe the "perfectly sharp" areas show up yellow. I've only heard it explained, never seen.
 
Hey Barry! I don't know if you have an answer for this after the type of testing you did... but you specify that camera motion causes a drop in sharpness when the camera moves; how about when there's moderate motion in the frame, but the camera's still?
Didn't test for that. I imagine it would be the same; I doubt the camera knows why a pixel moved, the sensor just sees pixels in a different position.

This res-lowering thing puzzles me and I want to spend some more time to get to the bottom of it. I mean, I have a working theory and it all makes sense except for when I test it in a way that should reveal it -- and it doesn't pan out.

It seems like it's doing aggressive noise reduction, which averages out the fine detail across multiple frames (noise goes, as does fine detail, especially in motion). But I can't detect motion trails, like I'd expect there to be, so I'm not sure.

Overall, as a practical matter, I'm not so sure it's a bad design decision. Lowering fine detail in motion sounds bad, but it actually would help the codec out, so it's probably clever engineering. Feed the codec too much detail and it'll choke, but smooth it out a bit when the codec needs it most and you get OVERALL better results. Note: I'm not saying the res-lowering is a disaster or anything, I'm just saying that it happens. I'd have to do some real-world stuff to see if it makes a difference to the point of where I'd care; initial impression is that it's probably not much of an issue because it works in concert with the codec and that's what you'd end up with anyway. But it'd be nice to test the HD-SDI output with that Convergent XDR at 160mbps i-frame.

Even so, that'd be a largely academic test. So I don't really know the answer to this one yet, and I haven't had a chance to put it through any sort of real-world tests. And that's all I really care about; lab testing, to me, is really only a springboard to identify what type of issues might arise in a real-world circumstance.

And how bad are the rolling-shutter artifacts in high-motion scenes with a still camera?
For flashes, pretty bad. For skew? Nothing to worry about. Something'd have to be moving very, very fast to even exhibit lean. A race car, maybe. A big semi truck barrelling down the freeway, maybe.

Where the rolling shutter is going to bite you, on the EX1, is in partial-exposure flashes, and in vibration. If you reverse the motion of the camera quickly, that's when it gets bad. I want to try it with a car mount, strapped to a harley, stuff like that. For mounted on a stable tripod you shouldn't have much issues unless you encounter something like putting your tripod on a riser and someone walks across the riser, causing vibration.
 
Last edited:
One question - what's the focusing like on the camera?
Well, there's two modes, the AF/MF mode and then the "full manual" mode. In AF/MF it's exactly like any other Sony cam. In "full manual" mode it appears to engage some sort of clutch that physically controls *something* in the lens. I'm not 100% sure how it does it, because it's a little funky. It might be true physical manual control, but if so it's based on a different type of lens design.

Did you ever use the zoom ring on a Z1/FX1? That was some fake hack-job emulation where they tried to make you think it was full manual but it wasn't at all. It was the same bogus rubbery servo stuff, but done as a ring instead of a rocker switch. Basically they tried to me-too the DVX's zoom ring, but the DVX was true physical manual. The Z1 was fake; if you drove the servo zoom motor the zoom ring wouldn't turn (no physical connection to the lens). So you could manually zoom to full wide angle, and then swap to power zoom and zoom to full telephoto but the manual zoom would still be showing full wide angle. See, no connection. But then when you swapped to "manual" mode, the lens elements would rapidly zoom back to full wide angle, so it would match what the bogus zoom ring showed, right? Only, it zoomed back at the maximum shot-transition speed, so it still took about 1.5 seconds or so to get there, so you knew it was fake. And if you tried to do a "snap zoom" you could see that there was a huge lag between how far you moved the fake manual zoom ring and how fast the zoom actually happened.

Now, with the EX1, it seems like with the focus ring they *might* have done the same thing. The zoom is absolutely pure manual, but the focus... well, it's a little odd. I'm not 100% sure on this, it's probably actually manual but it might be an improved version of what the Z1's zoom ring did. When you focus at minimum, you can then snap over to MF/AF mode and manually focus to infinity, right? And the manual focus ring marks don't move at all, so you're focused on infinity but the "manual" zoom ring still says minimum. And then when you swap into "true manual" mode, you can see the lens elements quickly re-focus back to what the manual ring says. So it's *possible* that it's the same fake stuff. However, there's a hefty "clunk" when you toggle between MF/AF and "true manual", so maybe there's some sort of clutch being engaged which is then giving a true mechanical linkage? I'm not sure.

In any case, it doesn't really matter, because it feels like it's responding truly manually, so what difference does it make? As far as feel, it's really not a big deal over the HVX's way; both provide precise repeatable manual focus with distance markings. The EX1's method is more familiar to broadcast users, but for precision and controllability they're about the same; the EX1 adds hard stops. As to whether the "true manual" is "true manual" or not, it's a difference that makes no difference. It's like Canon's 24F vs. 24P -- 24F works well enough that if there's a difference, I don't care. If the EX1's "true manual" is "true" or "fake" doesn't really matter, it performs fine.

You mentioned how much crisper the LCD is than the HVX, but I'm curious as to any 'focus assist' type of features.
Or is the LCD so crisp that focus isn't an issue?
Oh, the LCD's nowhere near enough to be reliable enough without some manner of focus assist! It's only 640x480. Blows away most camcorders' LCDs, but it's less than 1/4 of full res. So you need focus assist, definitely. The EX1 includes a full-screen "expanded focus" like the Z1/FX1, but you can actually use it while recording too (unlike Z1/FX1). It's tricky because if you forget you're in focus assist, you might be recording a very different frame than you think you are. But it's nice that they enabled full-frame. The peaking is great, you can set it in red/yellow/blue/white and you can set different levels of peaking. But, you can't use peaking and magnified focus assist at the same time -- WHY!??!?! Stupid. They definitely should let you; using "focus in red" along with "pixel-to-pixel" is heavenly on a BT-LH80. If they offered both on the EX1, I might dare say you wouldn't necessarily need any manner of external monitor. Without both together, um... maybe you can get away without a monitor, maybe you can't.

(EX1 users, if I got this wrong and you actually *can* use peaking + expanded focus together, let me know).
 
Last edited:
Hi Barry, Nice review

I heard RED has a rolling shutter does it have these issues ?

Red has a rolling shutter, as does Infinity. I've seen skew and partial exposure on a Red, and I've seen skew on the Infinity. Claudio Miranda, ASC did a comparison of the Red and F23 and he noted that when he accidentally bumped or kicked the tripod he noticed wobble on the Red.

So it exists. But Red has also said that they've aggressively worked to reduce the issue as much as possible. And, a Red is a heavy, heavy rig once you outfit it with a cinema lens and whatnot, so it should be a lot harder to induce vibration. And if anyone's managed to tame the rolling shutter, my bet would be on Red.

But no, Red is not immune, nor is Infinity. It's a pretty safe bet that if something has a rolling shutter CMOS sensor, it's going to exhibit rolling shutter issues. Maybe they can optimize it some, but even Jim Jannard has said that it's still going to exhibit some skew. The question is: are you likely to use the product under scenarios where rolling shutter artifacts will *be* an issue? For ENG, definitely. For cinema? Maybe not. Cinema's a lot more controlled circumstance. And the Red's a cinema camera.
 
Last edited:
Nice.
Similar to the JVC 'everything in blue' focus.
Similar but not exact; the JVC turns the screen black and white so that the only color in the frame at all is the colored peaking. The Sony doesn't do that, it leaves the image in full color and the peaking is also colored. So it's a lot easier to see the colored peaking on the JVC, in fact it's almost impossible to *miss* the peaking. On the Sony it depends on how much is in sharp focus, the peaking could be very minimal on a shallow-DOF shot or it could be all over the screen on a wide-angle outdoors deep-f-stop shot. But when there's not much in hyper-sharp focus, the peaking can get lost if it's just working on a few pixels (same with any other camera though).

But, keep in mind the JVC had a very low res viewfinder and very low res LCD and no expanded focus, so they had to be aggressive with their approach. Having peaking get lost in the color of the scene was not an option for them. I think Sony's got the peaking exactly right. I just wish they'd allow it to also work with expanded focus, then they'd have it ideal.

I really would like to see Panasonic adopt something like that.
Definitely. That's the way the BT-LH80 works, and that's why I'm gonna buy one. It'd be fantastic for all manufacturers to adopt this tech in-camera.
 
Last edited:
Is the severity of rolling shutter the same for all cameras? I don't see how Red could possibly "reduce" it (based on what the article said about what rolling shutter artifacts are).

Also, are rolling shutter artifacts actually noticeable when viewing footage, or only on frame-by-frame type stuff?
 
Wow Barry.
I was expecting you to just write back with "Yeah, focus in red."
I should have known better.
Thanks for the details!

there's a hefty "clunk" when you toggle between MF/AF and "true manual"

So what you're saying is that we're going to start seeing 'What's that Clunk?' threads in the Sony forum?
:)


Yeah, I know, you go to all the trouble to write that and I pick out the one lame thing I can make a joke about....
 
Thanks for the super detailed review, Barry. You touched on a lot of specific things I hadn't heard from anyone else. Unfortunately a lot of them are deal breakers for me...

I wasn't going to be able to get my hands on the camera any time soon to test it out myself so I really appreciate your hands on report.
 
Is the severity of rolling shutter the same for all cameras? I don't see how Red could possibly "reduce" it (based on what the article said about what rolling shutter artifacts are).
Well, see, you're entering some tricky territory. I don't know what Red did to optimize it. I do know how Thomson tried to reduce the rolling shutter effects on the Infinity; they actually run the chip at double speed to speed up the roll rate through the frame. So when you shoot 24p, the Infinity is actually running at 48fps and discarding every other frame, but they're running at that faster rate so it'll scan down the chip faster, thus cutting the noticeability of the artifacts.

The faster they can roll down the frame, the more it starts to act like a full-frame sensor.

But I don't know what Red's doing. I do know that Jim Jannard and Jarred have both specifically said that they were aware of the issues and that they have optimized the sensor to minimize any impact from the rolling shutter.

The faster your frame rate, the faster the rolling shutter scans down the frame, and so the less skew or wobble you'll see. That's why rolling shutter stuff is worse in 24p and less noticeable in 60p/60i. I also shot some rolling shutter tests at 4fps just to see how bad it might get, but I haven't even reviewed that footage yet (and it's 5:00 a.m., so... I'm signing off here pretty soon!) :)

Also, are rolling shutter artifacts actually noticeable when viewing footage, or only on frame-by-frame type stuff?
Way noticeable on footage. Look at Matt Jepsen's FresHDV post where he showed the cop car/strobelight thing. It's all over the footage and it's also visible live; rolling shutter isn't like MPEG artifacts because you can actually see the rolling shutter artifacts happening live. As for lean/skew, the easiest way to see it is to pan one way and then reverse direction and go the other way. The faster you do this, the more lean you'll see (the same lean is there, but the transition from the vertical elements leaning to the right, to the vertical elements leaning to the left makes it a lot easier to see. When would you do such a thing in a real-world scenario? Well, I'd imagine a basketball game would be the type of thing that might bring it out. Panning fast one way, then quickly reversing direction.

The worst is vibration, because then the camera is constantly reversing direction, and that's when you get the bad "wobble." Running with the camera COPS-style, or mounting it to a Harley, or mounting it to a hard car mount or a helicopter... those are the kinds of things that'll lead to the vibration wobble.
 
Back
Top