Yes, we obviously live on very different planets.
The Sigma with the Foveon chip shows much more balanced noise between
the channels, and considerably less in the red and blue channels from
what I see from the very page you sent me to. Further, Sigma/Foveon
released some new firmware which further reduces the red noise, while
improving the "sharpness" of the image. The write up is below:
Since you just want to be "right" I'm not going to continue this
discussion. If you read my comments fully, you see that I concede that
the Foveon chip in its current format, is probably best suited to
mid-level consumer cameras, which, by the way, are exactly the ones
which this thread started about, (The Canon G3). Frank Paris was
expressing his displeasure with the G3 and it's lack of full color
gradients in small things like tree branches which led to a flat look
rather than the rounded nature of the branch which was depicted on film.
I stated that this may well be the result of the limited color
resolution provided by that camera in spite of its 4 MP image sensor,
and the fact that the G3 is considered one of the better consumer
cameras out there. I believe one reason for this problem is the Bayer
filter pattern.
Yes, decent Bayer interpolation cameras are appearing in the lower price
ranges now, but it took years to get it there. WHo knows what the next
generation of Foveon color sensors will look like or cost.
On the consumer digicam end, I still believe the Foveon chip will prove
to have advantages not easily corrected in lower priced Bayer systems.
Art
David J. Littleboy wrote:
>
> From: "Arthur Entlich" <artistic-1@shaw.ca>
>
>
> Side by side reviews I have seen do not indicate noise is a major
> problem with the Sigma.
> <<<<<<<<<<<
>
>
>
> Slightly worse than the D60, but the 10D improves significantly on the D60.
> 10D luminance channel noise is one whole f stop lower. But yes, the SD9
> problems at higher ISO are artifacting, not noise, and may be
> implementation, not sensor, problems.
>
>
> There did appear to be some software/firmware
> code problems, however. Considering this is the first real
> implementation of an otherwise new technology challenging a fairly
> mature technology, I still believe it holds tremendous potential and
> should other companies take on the chip design, some very interesting
> product may come of it. I still really like the concept, and hope it
> flourishes, at some future date.
> <<<<<<<<<<<
>
> Again, it doesn't solve a real problem. When I was deciding whether or not
> to blow serious money on a film scanner, the downloaded D60 shots I looked
> at made 645 scanned with the Epson 2450 look sick. That's _seriously_
> impressive for a lousy 6MP.
>
>
> I am not impressed with most Bayer interpolation based image capture
> systems, although once it gets into the 6 or more megapixel rate, the
> results are pretty decent.
> <<<<<<<<<<
>
> Are we on different planets? The Bayer cameras (especially the 10D, but even
> the consumer ones at their lowest ISO setting) seem flipping amazing to me.
> Much better images than film can even dream of on either a per-scanned-pixel
> or a per-area basis, e.g., the 10D is 1/2.56 the area of a 35mm frame, but
> only slightly worse image quality of a full-frame film Provia 100F image,
> but vastly better at ISO 800 than anything 35mm film in a color negative
> material ISO 200 or over.
>
> What more could one possibly want than a full-frame sensor with 16 million
> 10D pixels???
>
> David J. Littleboy
> davidjl@gol.com
> Tokyo, Japan
>
----------------------------------------------------------------------------------------
Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe
filmscanners'
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or
body
