Canon 5D announced !

"Ray Fischer" wrote in message
...
William E. Graham wrote:

"Noons" wrote in message
...

Now, do this experiment: slap a Sigma or Tamron
aftermarket lens on a 5D2 and try to go for resolution.
Then watch the corners on the result. Then compare to
what it can do with one of Canon's own "approved" lenses.
World of difference. I wonder why. Must be because
Canon is the only maker of "good lenses"? Well, then:
try with one of Canon's older lenses, not in the 5D2
database. There is only 16 or so there.

Hey, it's Canon's OWN statement, not mine!
Blame them!

I see no basic reason why the sensing plane manufacturers couldn't build a
sp that is shaped as a part of a conic section, and thereby eliminate the
soft corners that are inherant in most lenses.........

It's really hard to make such a sensor. The wafers from which sensors
are made are sawn off of a bole, and the circuitry is etched from a
pattern projected onto the surface.

A wafer fabrication plant runs hundreds of millions of dollars.
Curved chips would require an entirely new technology and, which
possible, would result in sensors many times the cost of existing
chips.


Well.......I didn't say it would be easy.....:^)
But, it would be possible. And, like most electronic processes, once
started, the cost would come down with time until, eventually, they would be
giving them away for virtually nothing. It is a solution that would
compensate for building many lenses with an optical correction built into
each lens......It's kind of like building hand-held shake compensation into
the camera body, rather than into each lens. In this case, you would be
building spherical focussing correction into the sensing plane, rather than
into every lens you want to mount onto the body of your camera.

"Archibald" wrote in message
...
On Sun, 21 Sep 2008 15:50:05 -0700, "William E. Graham"
wrote:


"Noons" wrote in message
...

Now, do this experiment: slap a Sigma or Tamron
aftermarket lens on a 5D2 and try to go for resolution.
Then watch the corners on the result. Then compare to
what it can do with one of Canon's own "approved" lenses.
World of difference. I wonder why. Must be because
Canon is the only maker of "good lenses"? Well, then:
try with one of Canon's older lenses, not in the 5D2
database. There is only 16 or so there.

Hey, it's Canon's OWN statement, not mine!
Blame them!

I see no basic reason why the sensing plane manufacturers couldn't build a
sp that is shaped as a part of a conic section, and thereby eliminate the
soft corners that are inherant in most lenses.........

Yeah, and automatically change the shape as required for all the
different lenses.

Archibald

On the contrary, if the distance from the optical center of the lens to the
sensing plane were the same for all points on the plane, one shape would be
ideal for any lens, and the cost of building the lenses would be drastically
reduced. The way it is now, a special element, ground to compensate for the
fact that the sensing plane is flat like a film plane, has to increase the
cost of building every lens.......If this could be incorporated into the
shape of the sensing plane, the cost of designing and building every lens
would be reduced. To me, it is a much more elegant solution, and one that
was virtually impossible with film, but much more practical with sensing
planes........

In rec.photo.digital William E. Graham wrote:

Well.......I didn't say it would be easy.....:^)
But, it would be possible. And, like most electronic processes, once
started, the cost would come down with time until, eventually, they would be
giving them away for virtually nothing. It is a solution that would
compensate for building many lenses with an optical correction built into
each lens......It's kind of like building hand-held shake compensation into
the camera body, rather than into each lens. In this case, you would be
building spherical focussing correction into the sensing plane, rather than
into every lens you want to mount onto the body of your camera.

The problem is that each lens in the series would now have to have
a specific target for field curvature. If a lens were as far off that
target as much as a current design lens is off flat-field, then it
wouldn't have any advantage.

I'm not at all convinced that designing a series of lenses of different
focal lengths for a very specific field curvature would be all that
much cheaper or easier than designing for a flat field.

Peter.
--

"Peter Irwin" wrote in message
...
In rec.photo.digital William E. Graham wrote:

Well.......I didn't say it would be easy.....:^)
But, it would be possible. And, like most electronic processes, once
started, the cost would come down with time until, eventually, they would
be
giving them away for virtually nothing. It is a solution that would
compensate for building many lenses with an optical correction built into
each lens......It's kind of like building hand-held shake compensation
into
the camera body, rather than into each lens. In this case, you would be
building spherical focussing correction into the sensing plane, rather
than
into every lens you want to mount onto the body of your camera.

The problem is that each lens in the series would now have to have
a specific target for field curvature. If a lens were as far off that
target as much as a current design lens is off flat-field, then it
wouldn't have any advantage.

I'm not at all convinced that designing a series of lenses of different
focal lengths for a very specific field curvature would be all that
much cheaper or easier than designing for a flat field.

Perhaps not, but it seems to be a fact that right now, most lenses, (even
very expensive lenses) are soft at the corners.....There must be some reason
for this, and I think the reason is that the sensing planes (including film)
are flat, which is an inherent departure from the obvious spherical nature
of optics. The sensing planes of digital cameras proviede a solution to this
problem that was not practically possible with film. I believe that
eventually the designers of camera systems will realize this and do the
obvious.

William E. Graham wrote:

"Ray Fischer" wrote in message
.. .
William E. Graham wrote:

"Noons" wrote in message
.. .

Now, do this experiment: slap a Sigma or Tamron
aftermarket lens on a 5D2 and try to go for resolution.
Then watch the corners on the result. Then compare to
what it can do with one of Canon's own "approved" lenses.
World of difference. I wonder why. Must be because
Canon is the only maker of "good lenses"? Well, then:
try with one of Canon's older lenses, not in the 5D2
database. There is only 16 or so there.

Hey, it's Canon's OWN statement, not mine!
Blame them!

I see no basic reason why the sensing plane manufacturers couldn't build a
sp that is shaped as a part of a conic section, and thereby eliminate the
soft corners that are inherant in most lenses.........

It's really hard to make such a sensor. The wafers from which sensors
are made are sawn off of a bole, and the circuitry is etched from a
pattern projected onto the surface.

A wafer fabrication plant runs hundreds of millions of dollars.
Curved chips would require an entirely new technology and, while
possible, would result in sensors many times the cost of existing
chips.

Well.......I didn't say it would be easy.....:^)
But, it would be possible. And, like most electronic processes, once
started, the cost would come down with time until, eventually, they would be
giving them away for virtually nothing.

And is the sky filled with rainbows in your world?

It is a solution that would
compensate for building many lenses with an optical correction built into
each lens......It's kind of like building hand-held shake compensation into
the camera body, rather than into each lens. In this case, you would be
building spherical focussing correction into the sensing plane, rather than
into every lens you want to mount onto the body of your camera.

There is no incentive. The cost savings in the lenses would be
exceeded by the sensor cost.

--
Ray Fischer

"William E. Graham" wrote:

I see no basic reason why the sensing plane manufacturers couldn't build a
sp that is shaped as a part of a conic section, and thereby eliminate the
soft corners that are inherent in most lenses.........

To the best of my understanding, the problem is not curvature of the plane
of focus. The Canon 17-40 and 20/2.8 simply don't project a sharp image at
the corners unless you stop down to f/16. At f/8, the corners are ugly mush
no matter where you set the focus. (The hyperfocal distance for a 20mm lens
on the 5D at f/8 is 2 meters, and with the lens focused to 2m, nothing is
sharp at the corners. But at f/16, the corners show reasonable detail.)

--
David J. Littleboy
Tokyo, Japan

On Sun, 21 Sep 2008 17:04:33 -0700, "William E. Graham"
wrote:


"Archibald" wrote in message
.. .
On Sun, 21 Sep 2008 15:50:05 -0700, "William E. Graham"
wrote:


"Noons" wrote in message
.. .

Now, do this experiment: slap a Sigma or Tamron
aftermarket lens on a 5D2 and try to go for resolution.
Then watch the corners on the result. Then compare to
what it can do with one of Canon's own "approved" lenses.
World of difference. I wonder why. Must be because
Canon is the only maker of "good lenses"? Well, then:
try with one of Canon's older lenses, not in the 5D2
database. There is only 16 or so there.

Hey, it's Canon's OWN statement, not mine!
Blame them!

I see no basic reason why the sensing plane manufacturers couldn't build a
sp that is shaped as a part of a conic section, and thereby eliminate the
soft corners that are inherant in most lenses.........

Yeah, and automatically change the shape as required for all the
different lenses.

Archibald

On the contrary, if the distance from the optical center of the lens to the
sensing plane were the same for all points on the plane, one shape would be
ideal for any lens, and the cost of building the lenses would be drastically
reduced.

I'm not so sure about this... I believe the field of focus also
changes shape with the subject distance.

Anyway, the whole problem could be solved with no cost at all if you
just shot subjects that were in a curved plane.

Archibald

On Sun, 21 Sep 2008 19:25:12 -0700 (PDT), Annika1980 wrote:

I've always lusted after the 1DsMKIII, but I knew I'd never be able to
afford the $8K price. Now Canon is giving me something even better
for $2700. And HD Video to boot! I predict it will be a Merry
Christmas at the Annika Estate.

It won't be a Happy New Year if you discover that better lenses
are needed, pushing the loot meter back up around $8k.


The only drawback I see is that you have to use the Live View screen
the shoot the video. I'm gonna look like one of those P&S dorks
holding the cammy out in front of me and looking at the LCD.
"Smille for grandma!"

I see your 5D II on a leg-shortened tripod with you sitting behind
it on a director's chair looking at the LCD, wearing a beret.
You'll still look like a dork, but a pro dork, not a P&S dork.

"ASAAR" wrote:
On Sun, 21 Sep 2008 19:25:12 -0700 (PDT), Annika1980 wrote:

I've always lusted after the 1DsMKIII, but I knew I'd never be able to
afford the $8K price. Now Canon is giving me something even better
for $2700. And HD Video to boot! I predict it will be a Merry
Christmas at the Annika Estate.

It won't be a Happy New Year if you discover that better lenses
are needed, pushing the loot meter back up around $8k.

The vast majority of Canon lens 24mm or longer will be fine on the 5DII.

So the only lens needed here is the Zeiss 18/3.5.

Brett may not need any new glass at all.

(Sheesh, I wish Canon had come out with a 20/2.8 "II" instead of the 24/1.4
II. The old 24/1.4 wasn't a lot better than the 24/2.8 stopped down a bit,
but it was pretty decent for that radical a lens. The 20/2.8 is no better
than the 17-40 (except at f/2.8).)

--
David J. Littleboy
Tokyo, Japan

On Sun, 21 Sep 2008 22:57:25 -0400, Rita Berkowitz wrote:

Really? So different that the 5d2 is just crap?

I wouldn't go that far as of yet! True, the original 5D was a totally
awesome sensor wrapped in a **** body, but I think (hope) Canon solved this
problem. And the link to the images I posted a few days ago show that it
has a very slight edge over the D700 in image quality. Hell it should, and
probably a lot more for a camera of twice the MPs. I think smart money is
going to be well spent on this body when using it with Nikkors and adapters.
This way you get the best of both worlds. From what I've read, the 5D Mk
II's AF module is extremely slow and cumbersome, which is no problem if the
user is using Nikkors since they only MF. And for serious studio and
landscape photographers that want to squeeze every drop of detail and
resolution from their kit's performance aren't concerned about AF.

From what I've read you'll have exposure metering problems when
using fast Nikkors on Canon bodies. Metering problems disappear
when f/4 and slower lenses are used.