Nikon did what I hoped Canon would do...

Eventually, all of Canon's DSLR's will be full-frame. Which means that eventually, the smaller FOV lenses will have no use anymore, and a lens far out-lasts a body. Once people move forward to full-frame cameras (and inevitably Canon will have a line, ranging from inexpensive POS to high-end pro SLR) that Canon produces, these lensees will become garbage.

Consider that the camera advances far faster than the lens, and before you know it, the specially designed lenses for smaller sensor cameras that you have purchased are obsolete literally, like computer hardware, and its resale value down the drain because full-frame is what's going to happen down the line.

I would rather have a normal EOS lens today, and use it on something with a FOV crop, then upgrade to a full-frame down the line when prices have dropped, instead of buying these special lenses and then one day moving to full-frame and finding no more use for the lenses.

--
Tuan Nguyen.
Editor-in-Chief, CarAndModel.com
Technology Editor, Firingsquad.com
Contributing Editor, Maximum PC Magazine

Canon EOS D60
Canon 28-70mm f/2.8 L
Canon 70-200mm f/2.8 L IS
Sigma 50mm f/2.8 EX Macro
Canon 550EX and 420EX
Canon TC-80N3 Remote
 
Most if not all of the inner lenses areas contribute to a smaller imaging circle as well. You cannot "shave" a F/2.8 lens for 1.6 crop usage. Incoming light rays have to have a free passage through the whole (full) aperture hole, no matter what is the needed imaging circle on the focus plane.

If you want to make a seriously slimmed down version of true 300 mm focal length lens, it has to be F/4 or smaller aperture lens.

Matti J.
If you can design a 300/2.8 (true 300mm, not "effective") that has
a front diameter of less than 107mm, you'll probably win a Nobel
Prize in physics for it.

A smaller image circle would allow some parts of the lens to be
made cheaper/smaller. But you've still got to have that front
element be BIG.
I agree on the front element. Basic physics. I also wonder if the
front element on a 300/2.8 is bigger than 107mm? I don't know but
would be curious.

There undoubtably has to be some size weight savings for a smaller
image circle. I go back to the 300/2.8 for MF. It was a beast.
If I remember right it weighed over 10lbs and I will bet the front
element was much larger than 107mm. Its simply more difficult to
cover a larger area with a similar line/mm resolution. Look at LF
lenses. On a line/mm basis they are terrible. The quality comes
from the image size and the fact it has to be enlarged a lot less
than 35mm or MF.

Remember at one time people scoffed at getting quality images from
an image the size of 35mm. History has a way of repeating itself.

--
Jonathan
--
Aphorism - a shining substitute for genuine intellect
 
Anyone, with even the very basic understanding of optics, knows that making a smaller lense is pure stupidity in every possible way. Period.

Sam
 
Ahh, my first digital camera (and camera technically), the Kodak 2MP DC3400. Actually, the built in lens was near flawless, but this was Kodak's idea of a wide angle adapter, which gave about 19mm. Problem was, after you cropped out the crisp part, you were left with the original FOV.

Jason
Sheesh - that looks like it was taken with a really bad copy of the
14mm Sigma prime on a full frame body!

--
Olympus C2100UZI +B300 +A28, Canon D60, EOS7

My Ugly mug and submitted Photos at -------->
http://www.photosig.com/viewuser.php?id=27855

 
You need to post an example. As I mentioned, the blue type CA that you see with digitals doesn't happen with film, but that does NOT mean it's the fault of the cameras. If you put a good lens on the camera you will not see this problem. However, you still can get blooming, and in general when people are refering to this type of CA, they're talking about where blooming is mixing with this blue CA. Phil covers this in his reviews of consumer cams pretty well.

Jason
Actually, there's a free program / plug-in (mentioned above) that
can be used to get rid of the CA. Much better than using film
grain to hide the CA that occurs.

Yes, CA occurs even on 20mm primes. It's hard to completely get
rid of on really wide-angle lenses.
Just curious if anybody has really tested the same exact lens on
digital and film to look at CA. I have noticed that even some of
my tele lenses like my 200/2.8L and 70-200/4L produce CA on scenes
such as tree branches against a cloudy sky. Never seen anything
like this with film.
This isn't the CA we're talking about. This is a mixture of CA
caused with blooming. This doesn't really occur with film.
However, we're talking about real yellow/blue green/red etc. type
CA that you do see with film. Here's an example of the CA, look at
the sidewalk:



Jason
My suspician is that CA is largely caused by the algorithms used to
convert a color mask to full color information. I think wide
angles make this worse for some reason, but I have seen CA in
places I never would have seen with film with lenses other than
wide angle.
--
Jonathan
 
I think this is a bad move for Nikon. They're segmenting their lens line even further. Nikon use to emphysize their backward compatible compatible capability so much. Now it's very clear they're going to abandon it. Buying a Nikon lens is a lot of hassle now, to look at all the different type of lens and compatibility with different camera. AFS only work on the latest compatible bodies, G won't work with any manual camera, the D series for all the new metering stuff, etc. I once sold off all my Canon SLR equipment to get a prosumer digicam to use for a few years. I'm now starting with all new purchase, Nikon's lens line is really too confusing and iteroperability is so limited now. With the Canon EF line, a 15 years old lens work the same on the D60 except may be slower. If you want high quality just get an L. It's that simple. I'm pretty I'll be purchasing a Canon DSLR in the future because of this.

I doubt that Nikon're going to save much money by going to the smaller image size with the lens. I've read from pop photo that most of the expense from making a super wide on a 35mm slr comes from making the image focusing so far from the back of the lens in order to clear the moving mirror. That effectively make a wide angle lens a two part deal, with wide angle front elements to capture the image and a telephoto rear elements to focus on to the sensor. They're not going to avoid this with this new line. Cost saving will be every limited expecially they're only going from 1 to 1.5. If they're going to 1/2 inch like the comsumer digicam, that's entirely different.

just my 2 cents,
Tom
 
This has happened before, although since many of the members of
this forum have come into SLRs only in digital, they will not
remember.

APS. Very similar image size to Nikon's digital sensor. Nikon
released several IX-nikkor lenses, which were F-mount but smaller
image coverage. Canon, however, produced the 24-85 and 22-55
lenses, with full 35mm frame coverage. Whilst the 24-85 has proven
so successful that Nikon have had to copy it (put a 24-85 F3.5-4.5
AF-S next to a 24-85 F3.5-4.5 USM and you'll see what I mean),
where are the IX-nikkors now?
One could put an 'F' mount 35mm camera lens on a Nikon APS SLR but it was not possible to go the other way by putting an IX Nikkor on a Nikon 35mm SLR. The lens mount went into the body too far and would damage the mirror if tried. IX Nikkor are in the same place as APS SLRs - camera heaven.
The other thing to consider is that presently, ANY EF lens will
work COMPLETELY on ANY EOS body. Put the latest 70-200 F2.8L IS USM
onto an EOS 600, and you can use USM and IS. The same CAN NOT BE
SAID of nikon, where only a mere handful of bodies can use VR (not
even the current F55 can), and the much-vaunted "backwards
compatibility" is sorely lacking in present bodies (mainly F80 and
derivatives, ie D100, S2, DCS14), and it comes, as seen with G
series, AF-S and VR lenses, at the expense of forward compatibility.
I doubt that D1x owners, AKA 'pros' will have much trouble mounting the new 70-200mm 2.8 VR lens also introduced recently to their D1x/h cameras. Yep another copy as you would put it but I don't really want to fight about it.
NO such issues exist in the Canon EOS system, and I, and many
others, will be glad to see Canon keep it that way.
Except that one can mount an old manual focus Nikkor to a new D1x.

I dare you try putting a Canon FD mount on your D60.

My point is does it really matter?
 
Seems smart. Don't irritate your current customers (yes I am mad I
have to spend $1500 for a half decent wide angle on my D60) and set
yourself up for both amateurs and pros in the future.
Which lens are you talking about?
If Nikon had this lens available when I bought my D60 I may have
made a different decision.
I wouldn't. I could have bought a 14 mm prime, or Sigma's 15-30 -- either would have suited me better in the short term ... but the 16-35L is a much better long-term investment for my needs. Especially considering I have a good f/2.8 zoom for the long haul, while Nikon has a f/4 zoom that works for now, and with untested quality.
The lens can only be viewed as a plus my anyone contemplating
anything other than the very expensive full frame DSLRs now
available.
Well ... I'm expecting a not-so-expensive full-frame DSLR in the not-so-distant future...
Fuji should send Nikon a big thanks also.
Agreed!
 
Yup. Your point is well taken. I want lenses that'll also work on my film bodies and on the next digital SLR, etc. It is too limiting to have a series only for today's features when the morrow will bring a different sort of level playing field.

Gorham
--
Shooting Digital in Maine!
Please respond only through the forum.
Eventually, all of Canon's DSLR's will be full-frame. Which means
that eventually, the smaller FOV lenses will have no use anymore,
and a lens far out-lasts a body. Once people move forward to
full-frame cameras (and inevitably Canon will have a line, ranging
from inexpensive POS to high-end pro SLR) that Canon produces,
these lensees will become garbage.

Consider that the camera advances far faster than the lens, and
before you know it, the specially designed lenses for smaller
sensor cameras that you have purchased are obsolete literally, like
computer hardware, and its resale value down the drain because
full-frame is what's going to happen down the line.

I would rather have a normal EOS lens today, and use it on
something with a FOV crop, then upgrade to a full-frame down the
line when prices have dropped, instead of buying these special
lenses and then one day moving to full-frame and finding no more
use for the lenses.
 
Seems smart. Don't irritate your current customers (yes I am mad I
have to spend $1500 for a half decent wide angle on my D60) and set
yourself up for both amateurs and pros in the future.
Which lens are you talking about?
16-35L of course. I want to buy a decent lens from Canon at a decent price. That should not be so difficult, but right now it is impossible.
If Nikon had this lens available when I bought my D60 I may have
made a different decision.
I wouldn't. I could have bought a 14 mm prime, or Sigma's 15-30 --
either would have suited me better in the short term ... but the
16-35L is a much better long-term investment for my needs.
Especially considering I have a good f/2.8 zoom for the long haul,
while Nikon has a f/4 zoom that works for now, and with untested
quality.
The lens can only be viewed as a plus my anyone contemplating
anything other than the very expensive full frame DSLRs now
available.
Well ... I'm expecting a not-so-expensive full-frame DSLR in the
not-so-distant future...
Don't hold your breath. It will happen eventually, but its probaly several years away.

I think you do not understand the cost implication of a full frame sensor. Unlike other semiconductor products a full frame sensor cannot be reduced in size to get cost advantages. The reason the processor in your PC keeps getting cheaper is they continue to make it from a smaller portion of the wafer, hence more chips from one wafer. This does not apply to sensors. As time goes on there will be savings in producing them, but full frame will always be at least twice as expensive as the sensor in a D60. Also if it has more MP than the smaller sensor you need faster processors and memory. Someday that difference will not be too much, but it will take several years easily.

In some way it is like other electyronics and costs will continue to decline, but there is one major difference in the cost between the two types of chips and that is that the physical size of the chip matters.
Fuji should send Nikon a big thanks also.
Agreed!
--
Jonathan
 
Eventually, all of Canon's DSLR's will be full-frame
Only if all Canon DSLRs are going to cost > $2k. Cost of a sensor is geometrically proportional to size. And Moore's "law" assumes that devices can get smaller which mean it does not apply to sensors. When the Pentium CPUs were the die size of APS imaging sensors, they still sold for $500. And that's when you are producing millions!

If Canon is ever going to produce a true "Digital Rebel", then they will need a lens like this. Because if they don't, then Sigma will. (Heck, Sigma will anyway because the can sell to it for Nikon, Fuji, Canon, and Sigma SLRs.)

--
Erik
 
16-35L of course. I want to buy a decent lens from Canon at a
decent price. That should not be so difficult, but right now it is
impossible.
For an ultrawide zoom SLR lens to be damn-near as sharp ( at f/11 ) as a 50 mm prime, I would think it deserves to be called more than "half-decent." I think you're not giving Canon enough for what they pulled off, especially at f/2.8!
Don't hold your breath. It will happen eventually, but its probaly
several years away. [...] I think you do not understand the cost
implication of a full frame sensor. [...]
I understand all of this. Still, Kodak pulled it off for $4K, or double the price of a new D60. And I think it's as sure that we'll find cheaper methods of building big chips as it is that we'll have horseless buggies, put a man on the moon, and one day take photographs without film.
 
Still, Kodak pulled it off for $4K,
They've not pulled it off yet. (And not for $4k, closer to 5k). I'm sure that leaving off the AA filter helps.
double the price of a new D60.
But if they can get the cost down to D60 cost, then the D60-type camera should cost closer to $1k, right?
And I think it's as sure that we'll
find cheaper methods of building big chips as it is that we'll have
horseless buggies, put a man on the moon, and one day take
photographs without film.
Where's the flying cars! I want my flying car!

--
Erik
 
Eventually, all of Canon's DSLR's will be full-frame
Only if all Canon DSLRs are going to cost > $2k. Cost of a sensor
is geometrically proportional to size. And Moore's "law" assumes
that devices can get smaller which mean it does not apply to
sensors. When the Pentium CPUs were the die size of APS imaging
sensors, they still sold for $500. And that's when you are
producing millions!

If Canon is ever going to produce a true "Digital Rebel", then they
will need a lens like this. Because if they don't, then Sigma will.
(Heck, Sigma will anyway because the can sell to it for Nikon,
Fuji, Canon, and Sigma SLRs.)

--
Erik
Someone who understands the cost structure difference between sensors and other elctronics. Thats refreshing.

--
Jonathan
 
16-35L of course. I want to buy a decent lens from Canon at a
decent price. That should not be so difficult, but right now it is
impossible.
For an ultrawide zoom SLR lens to be damn-near as sharp ( at f/11 )
as a 50 mm prime, I would think it deserves to be called more than
"half-decent." I think you're not giving Canon enough for what
they pulled off, especially at f/2.8!
I was not calling the 16-35 half decent. My comment was that there is no 'half decent' lens in that range from Canon that is affordable. Only one that costs $1500. Thats not affordable in most peoples book.
Don't hold your breath. It will happen eventually, but its probaly
several years away. [...] I think you do not understand the cost
implication of a full frame sensor. [...]
I understand all of this. Still, Kodak pulled it off for $4K, or
double the price of a new D60.
Actually the Kodak is $5K and I have not seen a pic from it yet.
And I think it's as sure that we'll
find cheaper methods of building big chips as it is that we'll have
horseless buggies, put a man on the moon, and one day take
photographs without film.
Some day. Does Canon expect all the people on a budget to wait for years (it will not happen in a few months) while they figure it out.

Canon cold easily make a lens like this. It is no great engineering feat. It is pathetic that they do not.

Jonathan
 
However, the key is that eventuall all high-end technology trickles down to affordable consumer based products. One day (and it won't be too far away), a PowerShot S-something or a PowerShot G-something will have a full-frame sensor. And when that day comes, these specialized lenses will become useless.

"A cost of a sensor is geometrically proportioned to its size" - this isn't a law, it is the current state of economy in the processes of producing and manufacturing sensors. As full-frame sensors become more widely available and as the technologies that produce them become more economical, then the price of using them ultimately decreases.

You cannot compare a CPU die shrink to that of a sensor. Because a sensor can be full-frame whether or not there are more pixels in it. Contax's N Digital has a 6MP full-frame sensor, but yet the D60 is not. The process of manufacturing processors is that they need to squeeze more transistors into smaller and smaller sizes because of many factors such as signal strength, signal continuity, materials, power requirements, and such.

The trend for physical sensor size is to proceed to 35mm film area. Whether or not more megapixels are crammed into the sensor as it increases in dimension is unrelated. They could make 2MP full-frame if they wanted to, being that the pixel desnsity decreases and pixel pitch increases.

The trend is not to produce smaller sensors, it is to approach full-frame. Therefore one day, the minimal sesnor size of a digital camera worthy of being one is having a full-frame sensor. This process of course takes time, but this length of time will be shorter than the life of a typical EOS lens. Buying a lens designed to fit current sensor sizes, how long do you think that type of lens can last if Canon's goal is to make larger sensors with more mega pixels? Not too long.

Take a look at 4 top current Camera models in different price range from top to bottom:

1Ds, 1D, D60, and the G3

All have different sensor sizes. Would it be economical and wise for Canon to produce a lens line for a body that will eventually be gone? The D30 has already been discontinued, and its life span is miniscule compared to an EOS lens of any calibur.

--
Tuan Nguyen.
Editor-in-Chief, CarAndModel.com
Technology Editor, Firingsquad.com
Contributing Editor, Maximum PC Magazine

Canon EOS D60
Canon 28-70mm f/2.8 L
Canon 70-200mm f/2.8 L IS
Sigma 50mm f/2.8 EX Macro
Canon 550EX and 420EX
Canon TC-80N3 Remote
 
However, the key is that eventuall all high-end technology trickles
down to affordable consumer based products.
No, not all. I don't have (and cannot afford) a personal jet aircraft. And not because people haven't tried making them cheaper.
One day (and it won't
be too far away), a PowerShot S-something or a PowerShot
G-something will have a full-frame sensor. And when that day comes,
these specialized lenses will become useless.
Why on earth would a PowerShot need a full-frame sensor? I'll assume you mean a "digital Rebel". But whatever price point you pick for a full frame SLR, I'll ask for one at 0.5x that price. So, you will have to get the full-frame price down to $500 or so (the price of a film rebel body + some for electronics and inflation) before it will not make any more sense.
"A cost of a sensor is geometrically proportioned to its size" -
this isn't a law, it is the current state of economy in the
processes of producing and manufacturing sensors. As full-frame
sensors become more widely available and as the technologies that
produce them become more economical, then the price of using them
ultimately decreases.
Um, why do you think that this is "the current state of economy"? If cheap, large, semi-conducter devices could be made, then the person or company who figured out how to do it would be disgustingly rich. Digital imaging sensors would be a tiny part of this market.
You cannot compare a CPU die shrink to that of a sensor. Because a
sensor can be full-frame whether or not there are more pixels in
it. Contax's N Digital has a 6MP full-frame sensor, but yet the D60
is not.
And the Contax costs $5k to $6k and the D60 $2k. Hmm, wonder why THAT is ;-)
The process of manufacturing processors is that they need
to squeeze more transistors into smaller and smaller sizes because
of many factors such as signal strength, signal continuity,
materials, power requirements, and such.
Still, the price (costs are secret) of a processor is still more strongly correlated to size than any other factor. That is why the price can go DOWN even when the performance goes up.
The trend for physical sensor size is to proceed to 35mm film area.
Whether or not more megapixels are crammed into the sensor as it
increases in dimension is unrelated. They could make 2MP full-frame
if they wanted to, being that the pixel desnsity decreases and
pixel pitch increases.
And this is precisely the point. A full-frame 2MP sensor will cost pretty much the same as a 6MP full-frame sensor or even a 14MP full-frame sensor. So there is no (marketing) reason to make such a beast. (Actually, there do appear to be some physical reasons why very large pixels do not work that well, but that's beside the point.)
The trend is not to produce smaller sensors,
The trend is to produce smaller sensors -- they just are not useful in SLRS. One reason digital cameras have gotten cheaper is that the sensors have gotten smaller. Once, the 1/2" sensor was common. Now look at the P&S market: the inexpensive cameras have 1/2.7" or even 1/3.6" sensors. Two guesses why and the first does not count.
Would it be economical and wise
for Canon to produce a lens line for a body that will eventually be
gone? The D30 has already been discontinued, and its life span is
miniscule compared to an EOS lens of any calibur.
I am assuming that Canon will always have a Dxx camera in their lineup which will be about one half the cost of the full-frame sensor. Under this scenario, the body does not "go away", so it does remain feasible to produce the lens. Just rephrasing your assertion another way does not prove your point.

--
Erik
 
So taken from what you wrote, your point is:

Trend: To make smaller sensors with more megapixels.

Canon should make a line of lenses for non full-frame sensors.

---

Okay, so if the sensors are on a trend to getting smaller and smaller to make cameras smaller and more convenient, then if Canon comes out with a line of lenses for today's small sensors, what will happen when even smaller sensors come out? Then Canon's lenses designed for today's small sensors will be in the same situation as today's EOS lenses used on bodies with small sensors.

New trend: Canon keeps making more lenses for ever decreasing sensor sizes?

I'm sorry but if I were Canon, I wouldn't be doing that. Camera and sensor sizes change far quicker than lenses can improve. And to continually put R&D money into making lenses that are efficient to smaller and smaller sensors is not an option. Nikon has stuck to 1.5X FOV crop sensors. That's why they can do this.

Canon has a whole range of sensor sizes, they're not going to make ever more ranges of lenses for cameras with sensors that will ultimately continue to change in size. That is the key.

--
Tuan Nguyen.
Editor-in-Chief, CarAndModel.com
Technology Editor, Firingsquad.com
Contributing Editor, Maximum PC Magazine

Canon EOS D60
Canon 28-70mm f/2.8 L
Canon 70-200mm f/2.8 L IS
Sigma 50mm f/2.8 EX Macro
Canon 550EX and 420EX
Canon TC-80N3 Remote
 
I understand what you're saying, but if it's as simple as that, then a Buick must cost more to produce than a BMW Z5 convertible, since it's bigger. While sensor die size is of course significant, there are many, many more variables in DSLR production cost than that. The 1Ds sells for $8k because people are queing up to buy it at that price. Production cost for these things are completely unknown to anyone except Canon. Since they are the only ones who know the real numbers, and they seem to be following a full-frame strategy, it would seem that this is not a show-stopper. The D60 has the same die size as the D30, and yet costs less.

Why hasn't Sigma produced a single non-full-frame lens yet?
Eventually, all of Canon's DSLR's will be full-frame
Only if all Canon DSLRs are going to cost > $2k. Cost of a sensor
is geometrically proportional to size. And Moore's "law" assumes
that devices can get smaller which mean it does not apply to
sensors. When the Pentium CPUs were the die size of APS imaging
sensors, they still sold for $500. And that's when you are
producing millions!

If Canon is ever going to produce a true "Digital Rebel", then they
will need a lens like this. Because if they don't, then Sigma will.
(Heck, Sigma will anyway because the can sell to it for Nikon,
Fuji, Canon, and Sigma SLRs.)

--
Erik
 

Keyboard shortcuts

Back
Top