Sony PMW-200 Brings HD 4:2:2 Workflow to XDCAM Camcorder Line

[Les Wilson]

Originally Posted by David Heath

...I'll be the first to agree that story telling, lighting and script may be more significant that absolute technical factors, but it's an irrelevant argument. Production factors and technical quality are separate matters. You seem to be implying they are mutually exclusive? Surely that's not true? High technical quality doesn't preclude high production values, and vice versa. What's wrong with best possible story telling etc together with best possible technical quality?

Technical alone isn't enough. Anyone knows ergonomics affects the quality of a shot.

In the case of the rotating handle, it enables shots not possible with standard handgrips... e.g. getting the shot or not. Ergonomics affect how fast you can get a shot versus miss it altogether... better ergonomics can make a shot better and bad ergonomics make a shot worse. The XF305 zoom is a stellar example. It's a technically superior zoom to the EX1 and longer to boot. Yet the ergonomics require flipping a switch to enable the servos versus manual.

Ergonomics in this class of camera are important. They'll make that technically beautiful image better or not or not even exist because you couldn't get the shot. Getting shots you couldn't get otherwise and making shots better are why some of us care about ergonomics and why some of us care about having to fuss with shoe extensions, LCDs you can't close to save battery, peri-scoping shotgun mounts that aren't removable, absent shot transition automation, switches on zoom rockers etc because they affect quality too.

I hear complaining about sore wrists yet nothing about the arm, shoulder and neck strains of day to day camera ops with steadi-cams and the other designs. The wrist issue is handled with exercise like any other profession requiring physical prowess. IMHO the baby has been thrown out with the bathwater.

I've said this before, the EX1R ergonomics had a powerful advantage over competitors that contributed to its competitiveness against HD422 cameras. I feel the PMW-200 changes deliver less advantage. Pity.

With it's fiscal woes, I don't know if Sony even has the wherewithal to build a follow-on to the EX3 but if they do, I hope any PMW-300 reflects a better appreciation of ergonomics.

[Galen Rath]

Today I noticed B&H has their price up, as opposed to the previous suggested retail price that they had listed. Their price is $6299, same as the previous normal price for the EX1R, so that is good news.

[David Heath]

Originally Posted by Eric Olson

While minimum camera requirements can increase quality, the more noticeable effect of such requirements is to create marketing opportunities for expensive gear. What the industry really needs is widely available software that can perform a detailed analysis of a particular video master to determine whether it has suitable characteristics for delivery over current satellite, cable, terrestrial broadcast and fibre optic networks.

That's very good thinking, but the problem I'd foresee is that of just too many variables, coupled with the problem of noone wanting to do a lot of filming and editing ...... only for the computer to say no! You need to know where you stand BEFORE production starts, even before cameras etc are purchased or hired.

Hence the idea of recommending minimum standards. From a codec point of view, it means that if you get a camera with XDCAM 422 or AVC-Intra 100, you'll meet the recommended criteria, end of story. It's POSSIBLE that something lower may be good enough, but this takes away the uncertainty.

Of course, it won't guarantee good story, lighting, editing etc - but at least you have taken away the technical variable! :-)

Originally Posted by Chris Lawes

PLEASE tell me that 4+ years later they aren't re-using the same old noisy low resolution sensor that is getting destroyed by GH2s etc?

That really isn't true. Except at the higher end, (F3, C300 and higher) none of the large sensor cameras has as good resolution as a 3 chip camera with 1920x1080 sensors, and that's true of the GH2 as well. I believe it measures in at about 700lpph, but also has considerable luminance aliasing. (Better than many DSLRs, better even than the AF100 which has the same chip, but not up to the standard in resolving terms of a 3 x 1920x1080 chip camera.)

Alister said "Cameras like the FS100 and GH2 have big sensors with big pixels, that's why they have low noise." I agree with all the basic sentiments in principle here, but it's wrong to include the GH2 here. It has a big sensor, but with a high photosite count - it is, after all, first and foremost a quality stills camera. That means firstly that the indvidual photosite size is still small, and secondly that not all the photosites can be read at video framerates. The evidence is that it only reads 1 in 4, whch is why it's noise/sensitivity is not what the sensor size might lead you to hope for. It roughly works out about the same as an EX.

The up side is a 3 stop differnce in depth of field characteristics.

[Alister Chapman]

Originally Posted by David Heath

Alister said "Cameras like the FS100 and GH2 have big sensors with big pixels, that's why they have low noise." I agree with all the basic sentiments in principle here, but it's wrong to include the GH2 here. It has a big sensor, but with a high photosite count - it is, after all, first and foremost a quality stills camera. That means firstly that the indvidual photosite size is still small, and secondly that not all the photosites can be read at video framerates.

Yes, my mistake. The EX1 and GH2 pixels are roughly the same size, about 3.6 microns. F3 and FS100 are around 10 microns, almost 3x the size.

[Koravik Rakpetchmanee]

PMW 200 has better EVF compare to EX1 or F3 ?

[Alister Chapman]

It's the same EVF. The LCD is new and improved.

[Mark Andersson]

Alistar, wondering if you checked the AWB on the PMW200?

The EX's were appalling, and at certain times a good AWB comes in handy.

[Les Wilson]

The discussion between Panasonic executives and the broadcast folks is hysterical. But how can the 1/3" chip based HPX370 outperform the EX3 in low light and noise? The EX3 has 1/2" chips.

[Bo Sundvall]

Hi

Perhaps a little of topic, but will it be possible to edit files from PMW200 in Premiere Pro CS5.5 directly or does it have to be some kind of plugin installed? Or must PPro be upgraded to CS6 to work together with the new codec? As I remember, the Canon XF300 MXF files needed one or two upgrades of Premiere before it worked OK.


Regards,

/Bo

[Koravik Rakpetchmanee]

Originally Posted by Alister Chapman

It's the same EVF. The LCD is new and improved.

Thank you Alister. By the way I read Mr Bloom's review he said EVF is pretty good so I am not sure what's that mean.

[Alister Chapman]

Frankly, I don't know why so many people moan about the EX1R EVF. It's the same resolution as the Canon C300 EVF and very similar. I find that if I use peaking or the expand function I can focus perfectly accurately with it and once calibrated expose with it. Is it perfect, hell no, It's a small EVF on a small camcorder, but it's not as bad as many make out in my opinion.

I believe it works fine with CS5.5 as it is the same codec as the PMW-500 and other optical disc cameras and these have been supported for some time via the built in media browser. It's certainly directly supported in CS6, that's what I used to edit the review.

[Kris Zimbelman]

"But how can the 1/3" chip based HPX370 outperform the EX3 in low light and noise? The EX3 has 1/2" chips."

This question has been raised and I would like the answer to it also.

[Alister Chapman]

Image processing and noise reduction can mask and mitigate many sensor issues. The HPX370 uses a lot of noise reduction to reduce image noise. Most of the time this is very effective and can clean up the image, but there are almost always trade offs when you start doing heavy noise reduction. The most commonly seen issues are smear and posterization. This can be seen during pans and fast motion as a blurriness to the image and the reproduction of skin tones and subtle textures can take on a plastic, smooth look. It can also manifest itself as a "glazed" look to the image or blockiness and floating pixels and blobs within the image.

Camera sensitivity on it's own is meaningless because a camera manufacturer can simply choose to increase or decrease the gain of the cameras circuits to achieve whatever level of sensitivity that they choose. What really counts is how much noise does the sensor produce with no amplification, because this affects the amount of noise reduction that will be applied.

The EX cameras and PMW-200 use noise reduction as well, as does the Canon XF series. The problem is that the more noise you wish/need to remove/reduce the harder the processing has to work and the greater the artefacts become. At 0db with most of these cameras the noise reduction process goes largely un-noticed, but as you start increasing gain you also multiply noise levels, so the noise reduction circuits have to work harder.

For example: lets say that at 0db Camera A has 2 units of noise, Camera B has 3 units of noise. Only one unit difference so barely noticeable. When you go to +6db gain (6db is double) then now camera A has 4 units of noise compared to B which now has 6. So to give an image that looks similar to camera A the noise reduction circuits in B must work much harder and there will be more artefacts. Go to +18db and the difference is even greater with A now having 16 units of noise while B has 24.

A lot of the standard static camera tests like resolution and dynamic range tell you little about noise reduction and the artefacts that it can produce. Skin tone reproduction and pans can be very revealing however. One of the things I noticed in the PMW-200 is an improvement in skin tone and subtle texture reproduction which I believe is a result of improved image processing.

I can't stress enough that when assessing the image quality of a camera you need to always look at the complete image. It is wrong to focus on any one single factor without considering how that may affect other aspects of the image. Just because you can point a camera at a chart and note that the aperture may be stopped down by half a stop more than another camera does not really tell you anything useful unless you also look at the noise levels and also look at the artefacts that may or may not be present due to noise reduction.

[Kris Zimbelman]

Thank you Mr. Chapman

[Eric Olson]

Originally Posted by Alister Chapman

It is wrong to focus on any one single factor without considering how that may affect other aspects of the image.

Melamine in the animal feed makes the protein content appear higher; electronic image processing and noise reduction makes the video sensor appear more sensitive. In the first case, the protein content is not higher and the food becomes poisonous; in the second case the sensor is not more sensitive and the video may contain unexpected artifacts. The problem in both cases is that optimizing a product to maximize an incomplete set of test results without other constraints always reduces the final product quality.