On 25 Feb 2003 15:33:15 GMT, mitchgross@aol.com (MitchGross) wrote:

>>>The DVC200 has 800 lines as oppossed to 530 for the Sony.

>>This is a misconception. They both have the same absolute

>>resolution limits of 720x480, which translates into about

>>540 TV-lines horizontal resolution, which neither can

>>theoretically achieve, but the better will approach more

>>closely (though the 530 lines of the Sony is purdy durn

>>close enough! ;-). The resolution cannot be significantly

>>higher, given the limits of the D25 format itself...

>Wrong. The higher the resolution of the camera head, the higher the resolution

>of the image recorded to tape. Even though the codec allows a smaller maximum

>resolution, you don't get to choose which pixels stay and which go. Think of

>it as a grid on a grid, and those to grids only kinda line up. The finer the

>first grid the cleaner the transference to the second grid. Otherwise no one

>would bother to make a chip with a higher resolution. This can be easily and

>yes, dramatically illustrated simply by pointing both cameras at a resolution

>chart, recording the image and playing it back. The higher resolution camera

>will yield a higher resolution final image.

This is true for analogue, but not digital once the

"lesser" imager closely approaches the format limit,

though high-pixel-count CCDs can help improve

resolution, as you point out, as can higher-resolution

lenses (as we were talking about), but *not above* the

digital format limit...

>>>DVC200 has a Fujinon Bayonet lens which is higher quality than the Sony and

>>>can be changed.

>>Again, if a lens is diffraction-limited by f4 (the Sony),

>>given the resolution-limitations of the medium, NO OTHER

>>LENS will render a higher-resolution final image at f4 or

>>smaller with the same medium (though it can at a wider

>>stop...).

>Wrong again. These are easy tests and comparisons. Just point the two cameras

>at the resolution charts and test. A higher resolving lens will ALWAYS yield a

>higher resolving image, even on a low resolution format. And while the lens on

>the Sony isn't total crap, it certainly is not the finest optical glass ever

>made. One can buy a number of lenses for the DVC200 that certainly outclass

>it. Anf BTW, who limits themselves to a stop of f4 or tighter?

That was my point. And I have often made the point before

with analogue imaging systems that improving one part of

"the chain" *will* improve the end result (short the

format's maximum resolution - and your example

illustrates this), BUT, in a digital system in which the

lens is diffraction-limited, and the format is resolution

limited, and all else is equal but lens quality, there

cannot be a resolution difference at stops smaller than

the diffraction-limited aperture of the "lesser" lens...

>For those interested in a printed illustration of these differences, American

>Cinematographer ran an article last Spring comparing the XL-1 with stock lens

>to the XL-1 with the best possible lens made for it and the XL-1 with the P+S

>Technik adapter with a $15,000 cine lens mounted on it. The lens resolution

>charts DRAMATICALLY illustrated the differences. Even though the 35mm adapter

>introduced various abberation from rephotographing of a glass element surface,

>shooting through that very high quality lens produced significantly higher

>resolution final images. And these differences were apparent even on the

>fairly low-resolution sensor of the XL-1.

If these tests were done at a stop wider than the first

diffraction-limited stop of the Canon lens, I could

easily believe that the above would be true - but without

specifying the aperture used, this information may only

indicate that at wide stops (that may or may not often be

used, as you point out), the better lenses are better...

>If the recording format was the only barrier, then the image from a $500

>camcorder would be equal to an $18,000 DVCam camera. It isn't.

> Mitch

No, it isn't - there are other factors, like CCD size

and type, associated electronics, gamma adjustability,

etc...