«

Haven't you figured out how to put guido in the killfile/bozobin? I confess
my kill file is growing daily with this group - now if everyone would stop
extending the threads, e.g. B&H, Christians etc. Wonder what the limit is
in MSOE? Oops, please don't let this initiate a M$ or OE rant

Cheers,

rma


"Mark M" <> wrote in message
news:Z1H0b.11165$nf3.9514@fed1read07...
>
> "Guido Vollbeding" <> wrote in message
> news:...
> > Tomash Bednarz wrote:
> > >
> > > http://www.canon.co.jp/Imaging/eosdigital/index.html
> >
> > Another obsolete product with a Bayer sensor which produces
> > only artificial digital images. Don't buy!
> >
> > Regards
> > Guido
>
> Guido=Idiot
>
> Film photos are artificial images rendered via chemical reaction.
> Music CDs are artificial representations of instrument vibrations.
> Etc etc...
>
>

Looks interesting! It might be my path into digital SLR photography. I'm
going wait for reviews for both this and Pentax's D-SLR. The Pentax uses a
CCD and not a CMOS, so it is worth looking at. The Pentax shots I saw looked
gorgeous color and exposure wise, but were sampled down and post processed
for the web.

bg


"Tomash Bednarz" <> wrote in message
news:bhv1u3$riq$...
> http://www.canon.co.jp/Imaging/eosdigital/index.html
>
>

Charles Robinson wrote:
>
> Guido Vollbeding <> wrote:
> :
> : But not all digital image sensors drop two thirds of the native
> : picture information like Canon's.
>
> It's funny that a guy with an email address that has the word "JPEG" in it
> is complaining about dropping information. Have you looked into the JPEG
> specs lately and figured out just how much "unnecessary" information is
> dropped when compressing an image into JPEG format?

Look into the JPEG spec and you will see that JPEG *requires*
a *true-color* picture as a source for optimal compression and
image quality!
If you drop two thirds of *native picture information* from the
*origin*, then JPEG can't operate with optimal quality/compression.
That's why I, as a JPEG specialist, complain so much about that
kinky Bayer principle which only generates artificial images
with inferior quality and compression.
From a data compression standpoint, the Bayer reduction is one
of the worst methods imaginable, while JPEG is one of the best.

Regards
Guido

On Wed, 20 Aug 2003 15:31:29 +0200, Guido Vollbeding wrote:

> MarkH wrote:
>>
>> Foveon = 3mpix x 3colour per pixel.
>> Canon = 6mpix x 1 colour per pixel.
>>
>> Most of us can see that the Canon sensor delivers two thirds the data
>> of the Foveon, not one third.
>
> That's one argument.
> But my argument was another:
> From the same *picture frame* (image scene), the Canon sensor uses only
> one third of the native light information for building its image, while
> discarding two thirds.

That's entirely untrue. It discards 2/3 of the chrominance information
from a 6 megapixel image, but luminance is recorded at every photosite.
The quality of the iamge before it goes through the ADC also has a lot to
do with the lens, and the lens lineup is one of the problems with the SD-9.

From the digital side, however you look at it the 6mp Canon sensor
delivers a processed image containing 6.3 million 12bit samples, and the
Foveon delivers 3 million 36 bit samples. The SD-9 has a noise level that
is significantly inferior to the D60 at all ISO speeds. The 10D is
superior to the D60, particularly at ISO 800 up. The SD-9 doesn't even do
anything over ISO 400.

>> But because the Foveon is mounted in a Sigma, I bought the Canon
>> instead (And I’m very happy with the images from my 10D).
>
> Some years ago I was happy with such images, too. But today they suck.

Most others disagree with you. Foveon is a great idea, but you can't judge
a camera purely on the sensor technology. The SD-9 is just not up to the
quality of the Canon offerings, with the restricted lens range, low ISO,
colour-clipping, chromatic aberration sensitivity, and softness problems
at higher ISO. Buying a DSLR is buying into a camera system, and Canon
offers a (IMO) clearly superior system to Sigma.

Foveon has the potential to be the best imaging sensor technology on the
market, and I'd love to see a Canon camera with a 6 Mp foveon sensor.
However, it's been over 18 months since the SD-9 came out, and Foveon has
gone very quiet, with the sensor used in the SD-9 remaining their top of
line product. 2 years ago, I was hoping that many cameras on the market
now would be using Foveon sensors, but at this stage there seems a very
real possibility that Foveon could just fade away, and never make it into
any other DSLR. It would be sad, but it seems all too likely.

Mike.

In message <>, Guido Vollbeding
<> writes
>Charles Robinson wrote:
>>
>> Guido Vollbeding <> wrote:
>> :
>> : But not all digital image sensors drop two thirds of the native
>> : picture information like Canon's.
>>
>> It's funny that a guy with an email address that has the word "JPEG" in it
>> is complaining about dropping information. Have you looked into the JPEG
>> specs lately and figured out just how much "unnecessary" information is
>> dropped when compressing an image into JPEG format?
>
>Look into the JPEG spec and you will see that JPEG *requires*
>a *true-color* picture as a source for optimal compression and
>image quality!
>If you drop two thirds of *native picture information* from the
>*origin*, then JPEG can't operate with optimal quality/compression.
>That's why I, as a JPEG specialist, complain so much about that
>kinky Bayer principle which only generates artificial images
>with inferior quality and compression.
>From a data compression standpoint, the Bayer reduction is one
>of the worst methods imaginable, while JPEG is one of the best.
>
>Regards
>Guido

Bayer reduction through a colour matrix filter is no worse that the
process (in reverse) by which you are viewing the image on your computer
screen. The effectiveness of the process is all down to resolution and
perception. And it works, perceptually, so stuff theory!
--
Alan F Cross

Mike Brodbelt wrote:
>
> That's entirely untrue. It discards 2/3 of the chrominance information
> from a 6 megapixel image, but luminance is recorded at every photosite.

Wrong. It records one of RGB and discards the other two at every photosite.
Luminance is *not* recorded at every photosite! That would be nice for the
B&W photographers, but unfortunately it is not the case.

> The SD-9 has a noise level that
> is significantly inferior to the D60 at all ISO speeds. The 10D is
> superior to the D60, particularly at ISO 800 up. The SD-9 doesn't even do
> anything over ISO 400.

The SD9 has a noise level that corresponds to *real* measured data,
*not* to artificial calculated data in the case of Canon.
Have you ever seen noise in a computer graphic? So you see less
noise in the Canon because it is a largely computer-generated image,
other than the real measured image with the Foveon.

> Buying a DSLR is buying into a camera system, and Canon
> offers a (IMO) clearly superior system to Sigma.

But the "clearly superior" system doesn't make superior images, so what?

> Foveon has the potential to be the best imaging sensor technology on the
> market, and I'd love to see a Canon camera with a 6 Mp foveon sensor.
> However, it's been over 18 months since the SD-9 came out, and Foveon has
> gone very quiet, with the sensor used in the SD-9 remaining their top of
> line product. 2 years ago, I was hoping that many cameras on the market
> now would be using Foveon sensors, but at this stage there seems a very
> real possibility that Foveon could just fade away, and never make it into
> any other DSLR. It would be sad, but it seems all too likely.

But why should Canon (and Sony, btw) go quality (Foveon like) if
they can sell their rubbish to the dumb masses which accept the
inferior quality?

Regards
Guido

Guido Vollbeding <> wrote in news:3F433230.4CBB094
@jpegclub.org:

> Another obsolete product with a Bayer sensor which produces
> only artificial digital images. Don't buy!
>

Hahahaha ... what would the world be without you Guido?
Much, much, much less interesting. Don't give up! Fight
for your cause - ridding the world of this evil Bayer
technology. We all sinners that use it art doomed!


Roland

Hi,

You can find all of the Canon EOS 300D news, press releases, product images
and an interview with the head of Canon Europe Marketing on my blog:

http://www.photographyblog.com

Mark

"BG250" <> wrote in message
news:...
> Looks interesting! It might be my path into digital SLR photography. I'm
> going wait for reviews for both this and Pentax's D-SLR. The Pentax uses a
> CCD and not a CMOS, so it is worth looking at. The Pentax shots I saw
looked
> gorgeous color and exposure wise, but were sampled down and post processed
> for the web.
>
> bg
>
>
> "Tomash Bednarz" <> wrote in message
> news:bhv1u3$riq$...
> > http://www.canon.co.jp/Imaging/eosdigital/index.html
> >
> >
>
>

wrote:
>
> Hi,
>
> You can find all of the Canon EOS 300D news, press releases, product images
> and an interview with the head of Canon Europe Marketing on my blog:
>

Isn't it amazing what copy-and-paste can do for producing repetitive
unoriginal posts - luckily it works quite well in adding to my kill
file.
PLONK!

--
regards,
dslr

> That's why I, as a JPEG specialist, complain so much about that
> kinky Bayer principle which only generates artificial images
> with inferior quality and compression.
> Regards
> Guido

Tell me, Guidolt...
....How much Sigma Foveon stock do you hold?

You're an idiot, and...the newest member of my round-file club.
Bye bye.