A suggestion to Olympus

I stand corrected, I should have said Olympus is the only
manufacturer producing only telecentric lens (or perhaps I should
say near-telecentric) for it's DSLR. However, Nikon's range is
widening greatly.
That'a a pretty good way of putting it. Big question is, where's
Canon?
Could you please elaborate a little on this - where is Canon? You
mention that many existing Nikon lenses are 'near-telecentric', but
is this also true of existing Canon lenses?
Canon is fine for telephotos, but they are very limited in wide angle offerings.

The 14mm f2.8 has a good exit pupil location, as does the 24mm TS-E.

The zooms are another story. Neither the 16-35mm f2.8 nor 17-40mm f4 have the near telecentric exit pupils of the Nikon. They're both around 60mm. Good for the 1.6x or 1.3x crop, but not good for the full frame.
I read some of the extremely well researched articles on your
website and was particularly interested in your comments entitled
'Will "full frame" replace "APS sized" digital?' at:
http://www.swissarmyfork.com/digital_photography_ifaq.htm

I have been trying to answer this question myself for several
months now but my feelings about this are as cloudy as ever!
Personally I hope that the answer to this question is 'no', since
the quality of 35mm film can be equalled by an APS sensor based
system that is lighter, more compact and cheaper than its full
frame counterpart. The obsession that some people seem to have
about the 'desirability' of full frame sensors is, in my opinion,
totally misplaced (especially since most people rarely print at
larger than, say, 12x8). But although I hope that the answer to
the question is 'no', I fear that it will turn out to be 'yes' due
to the commercial pressures generated by large numbers of existing
customers with legacy 35mm lenses and the marketing pressures on a
particular brand of not being full frame when its main competitors
are (in particular, Nikon V Canon).
I think that the market pressure will cause Nikon to bring out a full frame camera in the next couple of years, but I don't see either Nikon or Canon trying to make it the mainstream.

The "legacy lenses" are, for the msot part, still quite usable on APS sized DSLRs. I think one overestimates the "pressure" that the legacy lens owners can bring to bear. Old lens owners didn't stop canon switching to the EOS mount, and they didn't stop Nikon from releasing cameras that couldn't meter with the older Nikon manual focus lenses.
I am currently a Canon DSLR user (10D), but in the past have been a
Nikon DSLR user (D100) and, going even further back, at different
times have owned both Canon & Nikon 35mm film SLR systems. Due to
the general uncertainty in the DSLR market and, most especially,
the lack of clarity of Canon's strategy, once more I find that I am
having a crisis of confidence in the long-term desirability of my
current system and find my eye straying towards other brands!

As I said above, I am a 'believer' in APS or 4/3 sized sensors and
ideally would like a well-matched body-lens system based on one of
these formats. But which brand is most likely to deliver such a
system in the long-term? Canon seems to be heading towards full
frame, and even if they continue to produce APS bodies (the 10DmkII
may be) I don't expect them to produce quality matching lenses
because Canon's top quality bodies have larger sensors (with crops
of 1.3x or 1.0x).
Well, even a couple of good 1.3x lenses would do wonders for Canon's position. The Nikon 12-24 has a bigger coverage circle than 1.5x, it almost hits 1.3, and is fine at the 1.37 that Leica is adopting for the Modul-R. If Canon put out a 12.5-25mm, just scaling up the Nikon design a bit, they would have an effective 16-32 for the 1.3x (and that would make a lot of photojournalists happy) and a 20-40 for the 1.6x users.

(to be continued)

--
A cyberstalker told me not to post anymore...
So I'm posting even more!

Ciao!

Joe

http://www.swissarmyfork.com
 
The main questions I have are:

1. Will Canon continue with APS sensors and produce some quality
EF-S lenses to match?
I think Canon will produce something like Nikon's DX lenses. They will probably do something like I suggested above, make the lenses cover a 1.3x crop. I'd bet at least 3 lenses for Photokina. That's just a marketing psychology thing. One or two is a PR stunt, 3 or more is "commitment". But they could do what Nikon did, show one lens and talk about 3 more coming.

(It won't be EF-S. EF-S is not Canon's DX. In Nikon terms, it's like the old IX Nikkors for the APS film SLRs. By altering the mirror design of the 300D dRebel, the lens rear element can protrude farther into the sensor chamber, so a really low cost kit lens can zoom a little farther into the wide angle range).
2. Has Nikon decided to settle on APS sensors and DX lenses or will
it jump to full frame some time in the future? (At present Nikon
appears to be more committed than Canon to APS sensors.)
I think they will do what Canon did, a ful frame that sells in the 10's of thousands of units, while the APS sized cameras sell in the millions.

I'm going to go out on a limb, based on digging through the recent patents, and say that Nikon is going to try to "scoop" Canon by launching a full frame with offset microlenses to counter the problems Canon and Kodak are having with "legacy" wide angle lenses.
3. Is the 4/3 system here for the long-term or will it collapse due
to lack of commercial success? (Has the Olympus E-1 been a
commercial success?)
I think 4/3 will be a long term success. The small sensor is "good enough" for an awful lot of purposes, and the ergonomics and build are great on the camera and some of the lenses. I don't see anyone other than Oly building bodies, though.
I know that no one can answer these questions with any certainty,
however it is getting increasingly expensive to switch systems and
this will only increase as my lens collection grows, so the sooner
I can identify the 'best' long-term system the better!
I don't think there is a "best" system, near or long term. It's probably going to stay much like it is today, Nikon satisfies some people's needs best, Canon satisfies others.
I am very attracted to the Olympus E-1 system, but how I would miss
Canon's IS USM telephotos! (How does Nikon's VR AF-S compare?)
Quite well. I've traded bags with Canon using friends before, the overall concensus is that the Nikon 70-200 is an extrodinary lens. The other AF-S VR offerings are also quite respectable.

But Canon still has IS on the big guns. If I were shooting birds on the wing with a Wimberly, I'd much rather shoot the Canon 500mm, and have the gimbals set near floating, than have to lock up the gimbals to use the Nikon.

Then again, if I had to guess what Nikon R&D has cooking, I'd bet they have more VR lenses as a higher priority than full frame.
All suggestions welcome Joe (and from anyone else, epecially those
with 'inside knowledge'!).
I'm running short on that, these days. :(

--
A cyberstalker told me not to post anymore...
So I'm posting even more!

Ciao!

Joe

http://www.swissarmyfork.com
 
I like those. Expecially the second one.

The thing to keep in mind about those articles is that it's "image side" telecentric lenses that matter for high quality digital photograph.

"Object side" telecentric is only important to certain specialized macro photography uses.

--
A cyberstalker told me not to post anymore...
So I'm posting even more!

Ciao!

Joe

http://www.swissarmyfork.com
 
That Foveon's color problem is fundamental and impossible to fix?
In other words, is it fundamentally possible to achieve perfect
color reproduction with Foveon?
It is not possible to have accurate color with any three filter system unless the filters reproduce the spectral responses of the himan eye (or at least have responses that can be linearly combined to reproduce those of the human eye).

The Foveon spectral responses do not resemble those of the human eye, and cannot produce a metamerism free image. The example I generally use is something painted a good "sky blue" color, such that it matches a particular blue sky in real life. Photograph this with a camera that matches the eye well, and the colors appear identical. Photograph it with one that has mismatched responses and the colors look different from each other, despite appearing identical to the human eye.

With more filters, the filters can have very "non biological" responses and produce quite good results. 6 regularly spaced bandpass filters (or lowpass filters, if you don't mind doing some subtraction) do a pretty good job.

Using more layers means thinner layers, and the sensor gets noisier. It's already noisier than a comparable Bayer, so I don't think this will be a viable solution.

http://forums.dpreview.com/forums/read.asp?forum=1027&message=7330479

Another problem is that the red layer doesn't appear, to me, to be deep enough to accuratly discriminate deep reds and orange reds.

I think this is why the last 50nm (the deep reds) was omitted from Foven's technical presentations.

There are tradeoffs there. The deeper the red, the more diffusion you get between cells (Sigma software already has "red sharpening" to counteract this). And deeper fabrication reduces accuray of the chip artwork, as well as decreasing yield (and increasing cost).

--
A cyberstalker told me not to post anymore...
So I'm posting even more!

Ciao!

Joe

http://www.swissarmyfork.com
 
Great information, Joe.

One question, how do you see in-body image stabilization performing relative to in-lens, and do you think it will be a viable competitor in the Minolta and possibly other (Pentax, Olympus, etc.) DSLRs.

Edwin
 
Great information, Joe.
One question, how do you see in-body image stabilization performing
relative to in-lens, and do you think it will be a viable
competitor in the Minolta and possibly other (Pentax, Olympus,
etc.) DSLRs.
I'm not that optimistic. The stabilizer in the smaller Minoltas doesn't work as well as the ones in Nikon and Canon lenses. It doesn't have the power or the speed to do as good a job.

Even moving from the A-1 "2/3 inch" sensor to the Sony APS sized sensor in the Pentax/Minolta/Nikon cameras means the moving sensor system has to cope with moving something that's at least 8.8x more massive (2.7x wider, 2.7x longer, 1.2x thicker). Since you have to move the sensor 2.7x farther to have the same degree of optical stabilization as an A-1, you've got 20x more force (23.8x more powerful actuators and control system) just to match A-1 performance with an APS sized camera. If you actually want to match a Nikon or IS lens, you have to at least double that.

Stabilizers in the lens mean you're moving things in a point in the optical system where the motions don't have to be as dramatic, and using rotary actuators instead of linear ones.
--
A cyberstalker told me not to post anymore...
So I'm posting even more!

Ciao!

Joe

http://www.swissarmyfork.com
 
Joe good morning and thank you very much for your comments.

I have a query concerning retrofocus/telecentric lenses. In the swissarmy site we can read "a digital lens could be described as a lens with one or more of these characteristics : .....strong retrofocus or telecentric design... etc"

As far as I know, the retrofocus design was introduced to wide angle lenses only, in order to increase the "physical distance between the rear lens element and the film", to create space for the SLR's mirror swing. The original intention was not to decrease the ray's incidence angle (this was not an issue in the film days). So, I believe the two terms (retrofocus - telecentric designs) are not the same, or not?

Besides, I suspect that the term "telecentric design" was used by Olympus. I do not remember if Nikon, Canon, Sigma etc, have used this term.

Additionally, I have read that the Sony sensor used in the Nikon D70 is sensitive to light coming even at an angle of 19 degrees from the perpendicular. If this is true, then what is the point in designing long throw/telecentric lenses? I think, you mention (in the site) that most wide angle designs have been improved during the last decades, and the incidence angle does not exceed 19 degrees.
regards,
George
 
This thread makes me believe all the more that the ONLY way Oly can compete is to move the 4/3 system upscale and create the first true dSLR system.

By this I mean:

1. interchangeable viewfinders ... INCLUDING EVFs as well as optical finders.

2. interchangeable and upgradeable backs. Why do new sensors have to mean new bodies?

3. a slot for a camera programming card. A dSLR is a computer and it should be possible for third parties to develop dSLR software! Imagine plug ins a la Adobe!
4. add-in modules for use on s-ecialized optical systems such as microscopes.

Such a camera would command a premium price and allow Oly to make $$$ off accessories as well as the initial body.
 
Many thanks for your replies Joe - as usual full of interesting facts backing up intelligent opinion.

I am still very uncertain about which system/brand is my best bet for the future, but, of course, that is the nature of the beast - rapidly advancing technology mixed with commercial pressures and secrecy - and, moreover, this problem is more down to my psychology than your excellent explanations! I realise that, in practise, the only sensible strategy is to get the system that most closely meets my requirements and is available today and accept that in the future some other system may become better for my needs and re-appraise the situation at that time. Planning ahead at the moment is virtually impossible.

The source of my difficulty is, to a large part, down to the fact that even today there is no system of any brand that fully meets my requirements and the uncertain future only compounds the problem. Anyhow, enough of this navel gazing ...
I think Canon will produce something like Nikon's DX lenses.

(It won't be EF-S. EF-S is not Canon's DX. In Nikon terms, it's
like the old IX Nikkors for the APS film SLRs. By altering the
mirror design of the 300D dRebel, the lens rear element can
protrude farther into the sensor chamber, so a really low cost kit
lens can zoom a little farther into the wide angle range).
So Nikon's DX lenses and your suggested Canon lenses are/will be only partially digital in as much as they cater for a smaller image circle but still assume a full sized 35mm refel mirror. So even these lenses (and the bodies with which thet are designed to be used) are unnecessarily large.

For example, I have just ordered a Canon 17-40mm f4 L lens for use with my 10D, and then I look at the Olympus 14-54mm f2.8-3.5 lens. The Olympus lens has a wider (and more useful for my needs) zoom range - exceeding the Canon at both ends and a superior maximum aperture, and yet it is also smaller (in both length and diameter), lighter and takes smaller filters (which has significant cost implications if several filters are used). The lenses have similar build quality (both weathersealed) and from what I have seen the image quality is also similar.

The one advantage of the Canon lens is that it has a larger image circle and so, in the future, can be used with larger sensors, but I have no interest in a larger sensor per se. I would only buy a body with a larger sensor if Canon made it impossible for me to buy the same specced body with a smaller APS or 4/3 sensor, since the image quality of the latter are now, and will be in future, perfectly acceptable (in comparison with a larger sensor constructed with the same technology).
2. Has Nikon decided to settle on APS sensors and DX lenses or will
it jump to full frame some time in the future? (At present Nikon
appears to be more committed than Canon to APS sensors.)
I think they will do what Canon did, a ful frame that sells in the
10's of thousands of units, while the APS sized cameras sell in the
millions.
So, even if I jumped ship (back) to Nikon now I would be no better off in the long-term, apart from the fact that Nikon does have a better range of DX-type lenses than Canon.
I think 4/3 will be a long term success. The small sensor is "good
enough" for an awful lot of purposes, and the ergonomics and build
are great on the camera and some of the lenses. I don't see anyone
other than Oly building bodies, though.
I'm interested that you think 4/3 will be a long term success - I wish I had your confidence. If I did then I might well pick up the phone now, cancel my order for the Canon 17-40mm lens and replace it with an order for the Oly E-1! (The only other thing that stops me doing this is the lack of long telephotos (with IS) for 4/3 - the Oly 300mm f2.8 does not count as it is ridiculously expensive!)

My doubts about the long term success of 4/3 are not technical related (I have no doubt that that the 4/3 sensor is indeed "good enough" and in the future will be able to support higher pixel counts and lower noise). No, my doubts are purely commercial - are enough people buying the E-1 to sustain Oly's research and sustain the production of new improved bodies and lenses (both from Oly and third parties, such as Sigma)?
I don't think there is a "best" system, near or long term.
I should say that I meant "best" purely in terms of my personal requirements - requirements differ from person to person and so there is not, and never will be, an overall "best" system.

Again, thanks for you help Joe.

Terry.
 
I don't know where Olympus is heading. The E-1 doesn't seem to offer much that isn't available elsewhere. And Olympus doesn't offer anything that can produce the quality of a 1Ds. They certainly haven't made a convincing argument to me that they are on to anything special.

All of this talk about telecentric lenses and alternative camera designs fails to note that digital cameras are already pretty phenomenal picture making machines. There are lots of offerings that suit almost all needs and new ideas are coming out all the time.

I have a Canon 1Ds and shoot wide angle with it all the time. The results are great. It's images are comparable to what I used to do on 6x6 and 6x9 but much easier. (You should try a 35mm Apo Grandagon on a 6x12 format - even with a center filter and f16, illumination is pretty uneven.) So I'm a happy camper. A lot of pros love the new 1D MarkII. It offers speed and quality. I'm sure the next model of the 1Ds will be great. I also have a Dimage A2 and that camera offers great features and quality in a cheap and affordable package.

It is becoming very hard for Olympus or anyone (even Nikon) to compete with Canon for the pro market. That's not to say that other companies can't sell some generic cameras or contribute some novel features or unique designs.

So the way I see it playing out is that demanding pros will use full frame DSLRs. (These will become less expensive and more common.) A number of people will stick with the APS sensors until the full frame ones come down in price at which point the APS cameras become irrelevant. Eventually, some full frame DSLR's will be just as compact and light as some fullframe film SLRs are today. And some will stay heavy and robust like the largest Canon and Nikon pro models.

There is a reason why 35mm SLRs evolved in the first place and other designs became less popular. (The Olympus Pen F had a limitted following.) This will hold true with digital unless some groundbreaking design changes occur. But I don't think the current cameras are much of limiting factor for the creative photographer. They already offer so much.

Along with this, the prosumer cameras (like the A2 and Pro 1) will improve enough that they'll be the camera of choice when size, weight and convenience is important.

Alan Goldstein

http://www.goldsteinphoto.com
 
..., I have read that the Sony sensor used in the Nikon
D70 is sensitive to light coming even at an angle of 19 degrees
from the perpendicular ...
That sensor probably has SOME sensitivity at 19º; for example the Kodak KAF-5101 sensor in the E-1 certainly does.

However, the degree of sensitivity is important: for the E-1's sensor, sensitivity at 19º is about 50% of sensitivity to light coming in perpendicular. So if light reached the corner of the frame at 19º off perpendicular, ther would be about one stop of "vignetting".

As far as I know, all current sensors likewise haveenough reduction of sensitivity at 19º for the fall-off in the corners to be noticable.
 
Joe good morning and thank you very much for your comments.
You're quite welcome.
I have a query concerning retrofocus/telecentric lenses. In the
swissarmy site we can read "a digital lens could be described as a
lens with one or more of these characteristics : .....strong
retrofocus or telecentric design... etc"
As far as I know, the retrofocus design was introduced to wide
angle lenses only, in order to increase the "physical distance
between the rear lens element and the film", to create space for
the SLR's mirror swing. The original intention was not to decrease
the ray's incidence angle (this was not an issue in the film days).
So, I believe the two terms (retrofocus - telecentric designs) are
not the same, or not?
Technically, they're not, and I should clarify that section of the FAQ.

But they sort of go hand in hand. Retrofocus lenses have their rear elements far forward of their rear nodal points.

The exit pupil if a lens is almost always forward of the rear element, which means it's also forward of the rear node. Since the rear elements are typically 40mm or farther from the film plane, you don't typically see exit pupils closer than 50mm to the film plane. For a 20mm wide, that means you have an exit pupil 2.5x farther from the film plane than the rear node, so you're already "near telecentric" in optical character (but not enough for the Oly definition).
Besides, I suspect that the term "telecentric design" was used by
Olympus. I do not remember if Nikon, Canon, Sigma etc, have used
this term.
Actually, Nikon uses it in the literature for their point and shoot cameras. And that's a good example of a telecentric design that isn't retrofocus. The back elements of the lenses come very close to the sensor, but the exit pupil is very far forward of the sensor.
Additionally, I have read that the Sony sensor used in the Nikon
D70 is sensitive to light coming even at an angle of 19 degrees
from the perpendicular. If this is true, then what is the point in
designing long throw/telecentric lenses?
There's "sensitive" and "optimal". Consider the 50mm f1.4. The exit pupil is about 50mm from the film plane (convenient... That's why I'm picking on it) and 35.7mm in diameter, or 17.9mm in radius. So marginal rays (the "perimeter" of a cone of light from the lens) from the right side of the lens end up hitting the left side of the sensor (12mm from center) at an angle of 31 degrees from perpendiicular. arctan( (12+17.9) 50).

Now, most of the rays from the lens are not in the far right margin of the cone, so it really doesn't impact light falloff (vignetting) appreciably. But it does contribute just enough really off axis light to cause strange problems.

My new pet theory is that the off axis marginar rays are the cause of the color shift that some people have observed in the D70. People blame the high shutter speeds, but I'm thinking it's the huge apertures you have to use to get the high shutter speeds.

http://forums.dpreview.com/forums/read.asp?forum=1034&message=8683851
I think, you mention (in
the site) that most wide angle designs have been improved during
the last decades, and the incidence angle does not exceed 19
degrees.
Quite true. I don't think there's a lens on the market for 35mm SLRs that has an exit pupil closer than 50mm to the film plane. That means that DX 1.5x crop sensor s never see light more than 16 degrees from perpendicular in the chief rays. And they're therefore very well behaved under most conditions. There may be some effects due to extreme angle marginal rays, but those appear to be very rare and mild in most cameras, as in the D70 effect that you really have to work at to see.

--
A cyberstalker told me not to post anymore...
So I'm posting even more!

Ciao!

Joe

http://www.swissarmyfork.com
 
..., I have read that the Sony sensor used in the Nikon
D70 is sensitive to light coming even at an angle of 19 degrees
from the perpendicular ...
That sensor probably has SOME sensitivity at 19º; for example the
Kodak KAF-5101 sensor in the E-1 certainly does.

However, the degree of sensitivity is important: for the E-1's
sensor, sensitivity at 19º is about 50% of sensitivity to light
coming in perpendicular. So if light reached the corner of the
frame at 19º off perpendicular, ther would be about one stop of
"vignetting".

As far as I know, all current sensors likewise haveenough reduction
of sensitivity at 19º for the fall-off in the corners to be
noticable.
Fortunatly, several things work in our favor.

The first is that the Sony sensor used by Nikon et. al. seems to perform better off axis than the Kodak. I've measured it in one degree increments.

The next is that the most extreme angle you encounter with conventional (50mm exit pupil) lenses on an DX sized sensor is 16 degrees from perpendicular in the corners, not 19 degrees.

Third, your, full stop los at 19 degrees on the Kodak sensor is reading the worst curve, the horizontal. You have to do these as vector sums, horizongal and vertical, and read the curve twice.

Or, build the vector sums into a new sensor curve, which is what I normally do.

If you look at a 50mm exit pupil and a 16x24mm sensor (DX rounded up to a perfect 1.5x, just to make it easy) you end up with 13.5 degrees horizontal, 9.1 degrees vertical. That's, as best as I can read the Kodak chart, 0.76 horizontal x 0.96 vertical, for 0.72 total, or 0.45 stops.

And, since we don't see a 0.45 stop vignet (I've checked) on a real D100, I'd say the actual pictures support my angular measurements, and I'll maintain the position that exit pupil is not a "real world" problem on 1.5x and 1.6x crop Nikon, Canon, Pentax, or Minolta cameras.

If it's a problem, there's automatic vignetting correction capability in Nikon Capture.

--
A cyberstalker told me not to post anymore...
So I'm posting even more!

Ciao!

Joe

http://www.swissarmyfork.com
 
Many thanks for your replies Joe - as usual full of interesting
facts backing up intelligent opinion.
You're quite welcome.
I am still very uncertain about which system/brand is my best bet
for the future, but, of course, that is the nature of the beast -
rapidly advancing technology mixed with commercial pressures and
secrecy - and, moreover, this problem is more down to my psychology
than your excellent explanations! I realise that, in practise, the
only sensible strategy is to get the system that most closely meets
my requirements and is available today and accept that in the
future some other system may become better for my needs and
re-appraise the situation at that time. Planning ahead at the
moment is virtually impossible.
Yes, makes life interesting, doesn't it. I also try to weigh in the cost of a switch. Let's try a hypothetical. Canon offers the valuable (to me) ability to meter stopped down on a bellows, and IS on the 300mm f2.8, features that I don't have in my current Nikon gear.

The cost of the switch is the effort to sell several dozen major pieces of gear (lenses, bodies, battery grips) and about 100 small bits (focusing screens, custom tripod mounting plates, etc) and accumulate a similar assortment of Canon stuff. And the financial loss associated with selling my gear and buying new gear.

By the time I figure in time and money, the swirch is going to cost me at least $2000. For that, I'll work around the lack of a stabilized 300mm f2.8 and metering on the macro stand.

Then there's the "learning curve" issue that is very difficult to calculate.

I've been shooting Nikons for decades. Canons focus in the opposite direction. The control layouts (the entire control philosophies) are different.
The source of my difficulty is, to a large part, down to the fact
that even today there is no system of any brand that fully meets
my requirements and the uncertain future only compounds the
problem. Anyhow, enough of this navel gazing ...
Yup.
I think Canon will produce something like Nikon's DX lenses.

(It won't be EF-S. EF-S is not Canon's DX. In Nikon terms, it's
like the old IX Nikkors for the APS film SLRs.
So Nikon's DX lenses and your suggested Canon lenses are/will be
only partially digital in as much as they cater for a smaller image
circle but still assume a full sized 35mm refel mirror. So even
these lenses (and the bodies with which thet are designed to be
used) are unnecessarily large.
Not particularly. About the only time you bring the rear element any closet to the sensor in a lens design that's "near telecentric" is when you're "stretching" a low cost zoom design to get a bit more wide angle. This doesn't help at the telephoto end, so the lens is still full length and full weight.
For example, I have just ordered a Canon 17-40mm f4 L lens for use
with my 10D, and then I look at the Olympus 14-54mm f2.8-3.5 lens.
The Olympus lens has a wider (and more useful for my needs) zoom
range - exceeding the Canon at both ends and a superior maximum
aperture, and yet it is also smaller (in both length and diameter),
lighter and takes smaller filters (which has significant cost
implications if several filters are used).
The one advantage of the Canon lens is that it has a larger image
circle and so, in the future, can be used with larger sensors, but
I have no interest in a larger sensor per se.
But it's the large image circle, not the sensor to mirror distance, that makes the Canon lens so much larger and heavier. A Canon DX would be as small and light as the Oly, because you'd reduce the coverage circle.
I would only buy a
body with a larger sensor if Canon made it impossible for me to buy
the same specced body with a smaller APS or 4/3 sensor, since the
image quality of the latter are now, and will be in future,
perfectly acceptable (in comparison with a larger sensor
constructed with the same technology).
Well, Sigma just launched a line of 3 reduced image circle lenses (OK, they're low cost "kit lenses") and Tokina has been showing a 12-24 reduced circle (supposedly high quality) that's a lot like the Nikon 12-24.

Canon has to get with the program sooner or later. They use 1.6x for cameras like 10D and 300D with sales in the millions, they know how big the market is.
I think they will do what Canon did, a ful frame that sells in the
10's of thousands of units, while the APS sized cameras sell in the
millions.
So, even if I jumped ship (back) to Nikon now I would be no better
off in the long-term, apart from the fact that Nikon does have a
better range of DX-type lenses than Canon.
Yup. There are other reasons, but it's all back and forth. Canon has the nice IS long glass. Nikon has the coolest portrait lenses.

TBC...

--
A cyberstalker told me not to post anymore...
So I'm posting even more!

Ciao!

Joe

http://www.swissarmyfork.com
 
I think 4/3 will be a long term success. The small sensor is "good
enough" for an awful lot of purposes, and the ergonomics and build
are great on the camera and some of the lenses. I don't see anyone
other than Oly building bodies, though.
I'm interested that you think 4/3 will be a long term success - I
wish I had your confidence. If I did then I might well pick up the
phone now, cancel my order for the Canon 17-40mm lens and replace
it with an order for the Oly E-1! (The only other thing that stops
me doing this is the lack of long telephotos (with IS) for 4/3 -
the Oly 300mm f2.8 does not count as it is ridiculously expensive!)
And big and heavy. I can't figure that one out, at all.
My doubts about the long term success of 4/3 are not technical
related (I have no doubt that that the 4/3 sensor is indeed "good
enough" and in the future will be able to support higher pixel
counts and lower noise). No, my doubts are purely commercial - are
enough people buying the E-1 to sustain Oly's research and sustain
the production of new improved bodies and lenses (both from Oly and
third parties, such as Sigma)?
I think it's doing well enough for Oly to continue R&D. Even if they're moving at the 70,000 unit/year numbers that Oly was talking about, that's still 100 million dollars/year, assuming each user takes about a $3000 "package" with body, a couple of lenses, grip, teleconverter, and Oly ends up with 1/2 the money. Even if only 3% makes it back into R&D, you're talking enough to launch a couple of lenses a year and a body every 2 years.
I don't think there is a "best" system, near or long term.
I should say that I meant "best" purely in terms of my personal
requirements - requirements differ from person to person and so
there is not, and never will be, an overall "best" system.
And, of course, personal requirements change with phaese of the moon (literally, for photographers. It's all about light).
Again, thanks for you help Joe.
Any time, Terry.

--
A cyberstalker told me not to post anymore...
So I'm posting even more!

Ciao!

Joe

http://www.swissarmyfork.com
 
Hello Joseph,

First, I am not sure why you believe it takes less force to move a Glass LENS in lens based IS versus moving a much lighter sensor (CCD) ?

Secondly:

While it's true that it takes more FORCE to move a heavier CCD than a lighter one, the FORCE does not increase with the distance moved, only the ENERGY required: It's the acceleration of a mass that takes determines the forcerequired, the distance only adds to Energy: AND once the required speed is reached, maintaining the movement speed takes much less energy (only overcoming air resistance and some friction)

Geir Ove
Great information, Joe.
One question, how do you see in-body image stabilization performing
relative to in-lens, and do you think it will be a viable
competitor in the Minolta and possibly other (Pentax, Olympus,
etc.) DSLRs.
I'm not that optimistic. The stabilizer in the smaller Minoltas
doesn't work as well as the ones in Nikon and Canon lenses. It
doesn't have the power or the speed to do as good a job.

Even moving from the A-1 "2/3 inch" sensor to the Sony APS sized
sensor in the Pentax/Minolta/Nikon cameras means the moving sensor
system has to cope with moving something that's at least 8.8x more
massive (2.7x wider, 2.7x longer, 1.2x thicker). Since you have to
move the sensor 2.7x farther to have the same degree of optical
stabilization as an A-1, you've got 20x more force (23.8x more
powerful actuators and control system) just to match A-1
performance with an APS sized camera. If you actually want to match
a Nikon or IS lens, you have to at least double that.

Stabilizers in the lens mean you're moving things in a point in the
optical system where the motions don't have to be as dramatic, and
using rotary actuators instead of linear ones.
--
A cyberstalker told me not to post anymore...
So I'm posting even more!

Ciao!

Joe

http://www.swissarmyfork.com
 
Hello Joseph,

First, I am not sure why you believe it takes less force to move a
Glass LENS in lens based IS versus moving a much lighter sensor
(CCD) ?
OK, look at it this way. The image shifting prism on a Nikon 70-200mm AF-S VR is a virtual 120mm away from the sensor. It's about 30mm in diameter. A rotation of just 0.48 degrees will shift the image 1mm on the sensor. That's a movement of 0.125mm at the edge of the disc, and (if I'm getting the moments of inertia right, in my head) about same force as moving the sensor 0.06mm, assuming the mass of sensor and prism are equal.
Secondly:
While it's true that it takes more FORCE to move a heavier CCD than
a lighter one, the FORCE does not increase with the distance moved,
only the ENERGY required:
If you increase the size of the sensor, and need to control vibration in fractions of a sensor's width (or pixels) you've proportionately increased the distance you need to move. If you have to move an increased distance in the same amount of time, you've increased force.
It's the acceleration of a mass that
takes determines the forcerequired, the distance only adds to
Energy: AND once the required speed is reached, maintaining the
movement speed takes much less energy (only overcoming air
resistance and some friction)
If you're reaching some "cruising velocity", you've obviously not reached the design limit of the vibration reduction system.

--
A cyberstalker told me not to post anymore...
So I'm posting even more!

Ciao!

Joe

http://www.swissarmyfork.com
 
Joe, good morning,
you are an invaluable source of information.
Thank you very much, especially for the time spent to answer!
regards,
George
 

Keyboard shortcuts

Back
Top