Got a call from my dealer in Sydney asking me to check our cameras for this problem as he'd already had one unhappy client ring to complain.

The camera bought for work shows the problem, at the same focal lengths as mentioned, mostly top RH side. The vignette can be made to move by panning the camera hard back and forth sideways.

Got home and checked my personal camera. Problem looked exactly the same but then I looked more carefully. The problem seems to persist right to the full focal length of the lens but it becomes less obvious. It seems that the vignette goes out of focus rather than out of frame. It's affecting a signficantly larger part of the frame but to a lesser extent, in fact it seems to expand to the point of affecting the whole RH side of the frame and the operation of the OIS significantly brings it more into the frame.

My dealer did confirm that the units in Australia are from the second batch. Sony Au believed the problem was fixed in manufacturing and therefore didn't test the cameras for it locally as part of their inspection process.

Perhaps the alignment has been changed from the first batch to the second batch and Sony believes that's fixed the problem but they shifted the alignment too far the other way, that's just my speculation of course but so far out of 2 cameras in my sample group both exhibit the problem. Based on comments from others who have units from the second batch the results are very similar.

I'm not too certain I could live with this problem. Indoor shots might be OK but outdoors with blue sky covering the top of the frame I think it could be fairly obvious.

Thought I'd add that this is happening to mine as well. After further testing I am convinced that from the top to the bottom along the right hand side is displying the vignetting - not just the corners.[/QUOTE]

Check that out with your waveform monitor. Visual inspection can be in error.

I'm all for making measurements but how to do this?
As I see it firstly we'd need an evenly lit large uniform surface. That I suspect is no trivial task without a spot light meter to go over the whole surface.
Having obtained that there's then the question of a suitable waveform monitor. I have a hardware monitor that I can use to select any one line but no SDI or HD SDI inputs. I could capture some footage into Vegas however I'm not certain just what the waveform monitor in Vegas actually shows us. I did a quick test using gen media and the results have me confused, it looks like it's the average of the whole frame in some strange fashion. With only black and white I'm getting part of the waveform at an intermediate value.

Perhaps capturing a frame and analysing that with PS would be better?

With all due respect, I don't think any measurements are necessary to confirm it, if it's there. Evenly lit surface? Just aim the camera at the sky and shoot; if you can see drastic brightness fall-off to any corner of the image, the camera is vignetting.

Finding the actual reason, or determining how severe it is - these are other matters, and indeed require more scientific methods. But I guess it's not us the end users who is supposed to do that.

I did use the Vegas waveform monitor which gives a fairly good indication of variation across the frame. However it is an accumulation and I intend to run a frame in PS to look more specifically at parts of the frame.

Bob,

If you don't have a spot meter, you can use your EX1 as one. Just set up the numerical brightness reading in the center of frame and pan the center marker around your surface to see if it's evenly lit. Zoom in all the way for the most accuracy.

And yes, Photoshop is arguably a better tool than a waveform monitor for looking at brightness across a surface because it allows direct readings in two dimensions, whereas the waveform aggregates each vertical column of pixels.

Of course, as Piotr said, I am not sure of the value of measuring the phenomenon - it's clearly visible once you shoot any reasonably even surface.

There has only been two reported in the whole EU channel, the discussions on here have made it seem alot worse than it is. No more will ship from sony with the slight defect, they're holding everything while each unit is checked.

Is it me, or is the above self-contradictory?
Please share with us the source of this news - my UK supplier "promised" me last week that I'd be getting mine this week! FWIW, my regular Polish supplier "knew" already back in October they would not get their first shipment before January, or even February next year, which made me order elsewhere :(

It's probably best that we don't start speculating about the size of the problem again, we won't achieve anything by doing so other than to have the thread locked again!

It really doesn't matter how big the problem is for Sony, we know it exists on some of the cameras purchased by members on this forum and we also know that Sony are doing something about it.

I have just received an email from my supplier with a statement from Sony informing him that all future deliveries to the EU from the factory will not be affected, and that any customers who do receive cameras in current circulation that are affected will have their cameras picked up and repaired free of charge.

That's good enough for me, I would be happy to have another EX1 now safe in the knowledge that the problem will definitely be fixed, and for free!

Thanks Paul. I'm sure Sony will do the same here in the states.
That's good news.

Paul, did you get your camera repaired by Sony?

No, I returned it for a refund. Hopefully getting another one early Jan.

I talked with Sony's Juan Martinez at DV Expo in Los Angeles, and we had a chat about the PMW-EX1 vignetting issue. We've also discussed it since then, and here's what I know so far.

First, some background: The prototype EX1 I had for review in October showed excessive "port-holing" (a very soft vignetting, with picture edges up to a stop darker than the center; the clear center was a circle perhaps 90% of picture height, with progressive darkening outside that area), and the optical stabilization caused motion-specific vignetting in the form of sharp-edged corner cropping, both problems occurring only at certain focal lengths, and only with the lens iris wide open.

My review unit was reported to have a misprogrammed lens, which caused the severe portholing and back-focus issues. I think the service menu back-focus fix described elsewhere on this forum would fix these sorts of things, but should not affect the asymmetrical corner-cropping so widely discussed, which is an alignment / centering issue.

I now have another EX1 for review, this time a nearly-final pre-production model (only the analog component board differs from full production models). This EX1 shows no abnormal port-holing: there is some slightly visible at full telephoto, full wide iris, and if I shake the camera with Steadyshot on, it steers around in response. However, at rest the port-holing is barely noticeable, and it's pretty much evenly centered, with no corner favored over another, and if I shoot a real scene instead of a blank wall, it's impossible to see, as the scene's own detail and variation in tone overwhelms the visibility of the port-holing.

Note that this is *normal* for industrial-grade wide-range zooms; it's certainly there on my Z1, and on the 15x Canon zoom on the DSR-500 I sometimes use, and on the stock Fujinon zoom on the JVC HD100--but, as on my current EX1, unless you shoot a featureless, blank wall, you'll never notice it.

More worrying is the asymmetrical vignetting or corner-cropping many people are seeing. The EX1 that Samy's DV & Edit had at DV Expo showed a noticeable corner crop in the lower right at full telephoto; my current review unit, like many of the ones discussed here, shows a tiny crop in the lower left corner, strongest between 10mm and 15mm, and most visible at full wide aperture. At worst, it only affects the outermost 5% of the image--If I shoot a DSC CDM chart, the darkening only creeps in halfway to the 90% safe area marker--but it is obvious.

And, as others have noted, when you shake the camera, the corner-cropping moves in response to Steadyshot.

Juan Martinez, a "Senior Manager, Technology" for Sony, confirms that some of the cameras shipped have this problem. This is the first one-piece, fixed-lens camera that Sony has done where the lens comes pre-built and pre-tested as a unit from Fujinon, and is then integrated with the camcorder body at Sony. While the lenses arrive bench-tested and approved, it appears not all of them are being mounted on the camera bodies in perfect alignment (having worked on several engineering projects involving third-party components, I can attest that it's easy for something to go seriously askew at the final integration stage without anyone noticing until it's too late). Juan said, "we go over these things with a fine-toothed comb, rejecting cameras even for almost invisible cosmetic defects. So, yes, we're taking this very seriously." Sony has reworked its internal processes to ensure that these problems won't happen in the future. The good news is that the camera is heavily backordered, so there isn't a lot of faulty stock sitting in warehouses waiting to be sold. The stuff currently being airshipped (as of 14 December 2007) should all be correctly aligned.

As to the cameras in the field, It looks at this time like Sony will have a procedure to set them right. Juan spoke off the record regarding the details, so I can't be more specific, but I'll spread the news (as will every other reporter concerned with the EX1) as soon as it's announced. It may take a while--remember the audio fix for the PD150?--but if you have an EX1 with obvious, asymmetrical corner-crop vignetting (not normal port-holing), there should be a remedy coming.

Now, will it make the lens perfectly flat, corner-to-corner, even with Steadyshot on? Probably not: aside from the slight edge darkening caused by port-holing, severe camera agitation may cause even a perfectly centered lens to corner-crop (or at least edge-darken) momentarily as Steadyshot steers the image around. On my current EX1, violent shaking in the 10mm-15mm range causes the upper left and lower right corners to corner-crop very slightly, perhaps 3% at worst, although at rest, or with Steadyshot off, both corners are clear.

If we assume that this EX1's lens is centered along that diagonal axis, we might expect a fully-centered lens to corner crop equally on all four corners this same way: perfectly clear most of the time, but showing a momentary corner crop at certain focal lengths when Steadyshot is working hard. My Z1 does exactly that, and over a wider zoom range, too.

But to keep things in perspective, I've never noticed this in my real-world Z1 footage over the past couple of years; it's only now, when I'm specifically looking for it, that I can see it when shooting a blank wall. When I shoot real-world stuff, scene complexity has (so far) always overwhelmed the occasional corner crop; if I were to shoot a flat sky or other expanse of smooth, featureless color, I might see it, but in all my footage of people and birds and water and cityscapes and foliage and the like, it may be there, but it simply doesn't draw attention to itself. If you spend your time shooting blank walls, it's worth worrying about. If you shoot the real world? Not so much.

So, to sum up: yes, a bunch (unknown percentage) of EX1 have alignment issues causing asymmetrical corner cropping. Sony is aware of this; Sony agrees that it's a problem. Cameras being delivered now are supposed to have the problem fixed, and Sony is planning to fix the problem on existing cameras: details to be arranged.

Adam, thank you for this detailed information!

I look forward to hearing how Sony will handle the problem cameras in the field.