The inadvertant misinformation that resulted above from confusing pixels with TV lines of resolution is an excellent example of how important it is not to get hung up on numbers and technical specifications, and to focus instead on how the image actually looks.

If it is one thing that the shoot out has shown us it is that at the end of the day there are only small details between all of the HDV cameras that are different. In fact the resolution should be the least of your concerns. I would much rather have a clean less detailed image than a bad image with more detail. I do not even care anymore how much resolution a HD camera has. I just want the image to be clean and natural and offer a decent workflow. Take the Panasonic camera for example. The chips are only 960x540. They do use pixel shift to gain more detail however and the images look very clean. It one of the lowest detailed cameras of these HD cameras but yet it has very clean images. The 540 vertical chip from the HVX200 isn't all that much different than the single 540 field from the H1. Both use some level of pixel shifting interpolation to gain more detail. I do not know exactly what 24F is doing but it is something to do with pixel shifting of some type. That means some scenes may have more detail then others depending on the colors in that scene.

To me wanting a camera because it has 5% more detail is like wanting to pay $1,000 more for a cpu for a new editing system to render your projects 5% faster. It may be faster but who cares.

I used to question 24F at first as well but I for the life of me cannot find anything to complain about it and I hate interlaced video with a passion.

But I've heard at least one person say the HVX200 yields a lot of artifacts in dim lighting compared to the Canon XLH1, so there again it's not a clear-cut comparison. The HVX200 certainly seems to deliver a different "look" than HDV cameras.

By the way, the difference in detail between a camera with a 960x540 sensor and one with a 1440x1080 sensor can be non-trivial, and if the point of HD is to deliver clearer images then that's something worth thinking about. As Chris says numbers aren't everything, but they can serve as a useful reference point.

Ok, I know the HVX has great color, but I think it has more to do with its DSP than the mere codec. I get a little tired of "4:2:2" bandied about like it's the full-raster real deal. http://digitalcontentproducer.com/hd...ch_720p24_pt3/

A quote from a friend: "color space ratios are badly misused, and it is less ambiguous to quote absolute horizontal resolution numbers. 4:2:0 does not mean that there is no V or Cr information stored at all, it means that in each line, only one color difference channel is stored with half the horizontal resolution. The channel which is stored flips each line, so the ratio is 4:2:0 for one line, 4:0:2 in the next, then 4:2:0 again, and so on. DVCProHD is 4:2:2 but at a ratio of 1280 luma samples by 540 cb and 540 cr. Not really as much chroma bandwidth as one would think at first."

It's sub-sampled raster 4:2:2 and a fairly heavy pass of compression on the image too. DVCPRO HD has trouble with saturated reds and excessive detail. I know this from having captured raw SDI from the XL-H1 many times to the codec. (bypassing HDV compression)

The HVX is a great camera for how it handles images, especially for the lower res CCD, but let's not call it "true 4:2:2".

True HD 4:2:2 is sampling an image with half the chrominance of the luma in a given HD frame size, i.e. 1920x1080 or 1280x720, in which case the only camera that does this in the category currently is the XL-H1 (live camera head signal) to a full raster 4:2:2 codec (Sheer, CineForm, PhotoJPEG, Uncompressed) via SDI.

Barlow...great post and great link! Now I understand why the resolution differences have less affect in HDV and DVCProHD recording.

It also beings to answer some questions I've been asking. In your last paragraph you state:

Is that only from SDI, or is it also from HD component output? All of the camera's have HD component output. When doing this, would the different resolutions of the the Canon's interlaced CCD vs the JVC progressive CCD be more apparent (since the chroma information would be captured)?

Thomas and Kevin, it is my experience that a change is only noticible if it was a bottleneck. If it wasn't...you don't see the change. A 500hp M5 BMW is no faster than a 225hp 525 on Manhattan's 3rd Avenue in the evening rush hour.

Even then, in general use, a 33-50% change is generally needed for the change to be generally noticable. However, a 10-20% change may 'feel' better in some circumstances. The difference in power between a 525 and 530 BMW.

And what you are saying is that these cameras are all with in 10-20% of each other. And with different strengths and weaknesses, it all averages out.

Chris, is there a place people can post their configurations and post processing when posting footage? As Josh said, there is some outstanding footage posted. However, without knowing what setting acheived that affect it may be hard to produce. Maybe a standard way of posting settings.

Thanks to everyone,

David

Yes and no. The most impartial reviews of all these cameras conclude that they each have strengths and weaknesses which mean there isn't one clear "best" one for all users and all purposes. But if you want to talk numbers, the XL-H1 has resolution up to 50% better than the HVX200 and can deliver roughly 15 times the data bandwidth via HD-SDI, while the HVX200 can record up to 4X the bandwidth that the XL-H1 sends to HDV tape. So depending on what you want to do, that may all even out or it may not, and the only way to decide what works for you is to either test the cameras yourself or look carefully at an assortment of sample footage.

Which brings up the question, how does HD-SDI (on the H1 and yet to be released G1 and HD250) compare with HD component (most cameras support both 1080i and 720p) on the others?

Panasonic did a table that took the four sensors (960x540p, 1280x720p, 960x1080i, and 1440x1080i), and applied two factors: 1.5x for green shift (h and v on sensor 1, h on 3 and 4) and 70% vertical resolution on interlaced sensors. This resulted in 1440x810, 1280x720, 1440x756. and 2160x756 respectively. However, the recording formats are 1440x1080 Y for 3 and 4, 1280x720 Y for 2 and 960x720 Y for 1. And the chroma sampling was effectively 720x540 for 3 and 4, 640x360 for 2 and 480x720 for 1.

The Steve Mullen article that Barlow referenced indicated that in their native storage mechanisim (HDV and DVCProHD), the chroma samples recorded are below the resolution of the CCDs, and comparable between the formats. In that analysis, coupling the sensors with the recording format yields rather similar recorded resolutions.

If this information was valid, the recorded information would be between 1440x756 and 1280x720, about 15%. However the chroma sampling may favor the 1080i recording format. None of this deals with motion compression though. Or dynamic range net of compression.

Personally, I would love to see direct to Cineform recording in the camera. If that was a $3k option on the A1 (a C1?), would you buy it over a G1?

Is there a way to objectively measure the lens differences between the cameras?

None of this is relevant. None of this has any bearing whatsoever on "image quality." Image quality is determined solely by the person operating the camera. All that matters is where and how you point the camera. Forget this numbers nonsense and test the cameras yourself, as Kevin suggests above. Actual hands-on time is the only way you can make a logical determination about which one to buy. You're not accomplishing anything by spouting numbers and tech specs on an internet message board... in fact it's highly counterproductive.

I think this topic has outlived its original purpose. Let's please move forward with discussions that are more appropriate to this community, as in how to use this gear and what are we creating with it. Enough of this pointless "which one is better" crap. Thanks in advance,