One chip article by Steve Mullen

[Heath McKnight]

Check this out! An engineer I know at my TV station liked what he read about the one chip in JVC's HD10. Our own Steve Mullen wrote the article:

http://videosystems.com/ar/video_ccd...eded/index.htm

heath
www.904am.com

[Chris Hurd]

Yes, Steve is an excellent writer and that's one of the main reasons why I invited him to help moderate here at DVi.

[Frank Granovski]

An interesting techno-talk type article. I think this is the argument's bottom line: "Three chips and an optical-prism add bulk and cost."

[Heath McKnight]

<<<-- Originally posted by Frank Granovski : An interesting techno-talk type article. I think this is the argument's bottom line: "Three chips and an optical-prism add bulk and cost." -->>>

It was all way over my head, but the fact that my engineer said it was cool made me more confident.

When I get back from L.A. next week, I'm demoing it.

On another note, a lot of my friends who went to NAB didn't like this camera. Then again, they think I want to forego buying a $35 - 50,000 camera for this. Um, no, I don't feel like blowing that much money on a camera that will be out of date in a year. I'd rather buy this camera for 10 times less.

heath

[Steve Mullen]

<<<-- Originally posted by Frank Granovski : An interesting techno-talk type article. I think this is the argument's bottom line: "Three chips and an optical-prism add bulk and cost." -->>>

Not quite.

JVC's chip design if implemented with 2 million pixels could give full 1280x720 resolution and 3-chip color rez because it would have 2X over-sampling.

BUT, unless it was larger (1/2-inch, at least) it would be very insensitive and have very low light-latitude.

The camera manufactures will claim this will cause a increase in cost. This is true, but CCD cost may not be that big a part of delivered camcorder cost. I think we would all happily pay an extra $500 to eliminate the insensitivity and low light-latitude of the small chips from Sony, Panasonic, and JVC.

Note, the next gen single chip designs are all large, e.g., Foveon and Dalsa.

In short, we can stop counting chips WHEN chip designers give us chips that provide ultra resolution at low-cost and with no negative side-effects.

Until then, 3 CCDs are best, but cleverly designed 1 CCD chips can be OK.

[Frank Granovski]

Thanks, Steven, for the clarification.

On another note, one or 2 of Hitachi's first DVD 1 chip cams had 1/2" CCDs, or so I had read. Too bad they don't one of these into some of today's 1 chip cams.

[Heath McKnight]

Quick questions, but I'll also research them:

Does the HD10 capture 720i or 720p? And is that only at 30 fps or also at 60i? Or is 60i only HD at 480i or 480p? Or am I a moron here trying to sound smart?

heath


<<<-- Originally posted by Steve Mullen : <<<-- Originally posted by Frank Granovski : An interesting techno-talk type article. I think this is the argument's bottom line: "Three chips and an optical-prism add bulk and cost." -->>>

Not quite.

JVC's chip design if implemented with 2 million pixels could give full 1280x720 resolution and 3-chip color rez because it would have 2X over-sampling.

BUT, unless it was larger (1/2-inch, at least) it would be very insensitive and have very low light-latitude.

The camera manufactures will claim this will cause a increase in cost. This is true, but CCD cost may not be that big a part of delivered camcorder cost. I think we would all happily pay an extra $500 to eliminate the insensitivity and low light-latitude of the small chips from Sony, Panasonic, and JVC.

Note, the next gen single chip designs are all large, e.g., Foveon and Dalsa.

In short, we can stop counting chips WHEN chip designers give us chips that provide ultra resolution at low-cost and with no negative side-effects.

Until then, 3 CCDs are best, but cleverly designed 1 CCD chips can be OK. -->>>

[Peter Moore]

It captures 720, 30p. There is no 720 60i standard, at least it's not commonly used. 720 is always 60p (and the JVC cam just doubles each frame).

[Heath McKnight]

<<<-- Originally posted by Peter Moore : It captures 720, 30p. There is no 720 60i standard, at least it's not commonly used. 720 is always 60p (and the JVC cam just doubles each frame). -->>>

Can this be converted to different frame rates using Magic Bullet HD? Like 24p, 25p; etc.?

heath

[Peter Moore]

It's very difficult to convert 30p to 24p. There are other posts on this by people who know a lot more about the topic than I do, but it involves a lot of manual conversion and costs a lot to get it done right. Any automatic process is going to leave odd motion artifacts.

[Dennis Adams]

Steve:

Here's an interesting question that occured to me while reading your article: Wouldn't it be possible to make a 2 CCD camera? It would be a hybrid of a 1 CCD camera (using a pair of 2x2 filter patterns) and a 3 CCD camera (except with 1 prism instead of 2). As CCD resolutions increase (which, as you point out, obviate the need for 3 CCD cameras), this might be a moot point, but I'm intrigued about what the advantages could be.

I have felt for a while that large area, high resolution single CCD cameras will eventually win out over heavy, complex optical path 3 CCD cameras. The 5 megapixel still cameras prove this out; they just need to shoot 24 frames per second and the argument is over.

///d@

[Steve Mullen]

<<<-- Originally posted by Dennis Adams : Steve:

Here's an interesting question that occured to me while reading your article: Wouldn't it be possible to make a 2 CCD camera? -->>>

I know of two 2 CCD cameras: One by Sony for ED Beta and one by Minolta. A green filter was over one CCD and blue/red striped filter over the other. The G supplied the luma. All three supplied chroma.

[Chris Hurd]

There was another 2-CCD camcorder; the Panasonic AG-455 in S-VHS format. All color info to one chip; luma on the other.

[Dennis Adams]

Chris, Steve:

Thanks!

///d@