Kodak out of FF DSLRs...is Canon next?

1) film accepts light from a wider angle than a digital sensor does for correct reproduction

2) Last Photokina C mentioned the problem on photosite size (there was an article in Let's go digital when visiting Canon booth but cannot find it right now) please bear in mind that CMOS photosite is surrounded by circuity

3) Canon Cmos has small sensors but microlenses cover a larger area, this is an obvoius highway to CA

4) there is no higher cost to the manufacturer, only to the final user

Prints can be made in any size. It is just a question of using the right size of paper, or cutting the paper if necessary.

The beauty of the A-standard for paper sizes is not that you don't have to think about cropping, it's that if you have, for example, too much A4 paper in your factory, but not enough A5 paper to meet demand, all you have to do is cut the excess A4 in half. It's a trick to make manufacturing easier, and less expensive.

Naturally, less expensive paper is more popular, so it has stuck around.

While a number of people see FF as the holy grail of (D)SLR photography here are the numbers:

Canon is only making 2,000 FF DSLRs per month. This hasn't changed from the move from Mk I to II. Whether the limiting factor is supply or demand or a combination, I don't know.

Kodak sold 80,000 FF DSLRs in 2004, roughly three times the number of FF DSLRs Canon produced in 2004.
http://forums.dpreview.com/forums/read.asp?forum=1000&message=13175982

In contrast, smaller (cheaper) DSLR output has sky-rocketed. Take for example the D70. It started in the 70K/mo output, then went to 90-100k. Now the next iteration (D50 and D70s) are a combined 220k/month.
http://forums.dpreview.com/forums/read.asp?forum=1034&message=13730876

So asking the question is not ridiculous. I am not saying it will happen, I am just saying that asking the question is not ridiculous at all.

Peter said:

There is absolutely no link between kodak abandoning DSLR and canon
abandoning FF.

This whole thread is ridiculous.

Click to expand...

--
My brand new photography blog: http://www.livejournal.com/users/photographyetc/

Roger Krueger said:

photogaustin said:

I think many of you are just forcing your wishfull thinking on
technology, instead of looking where it is going.

Canon's 350 has less noise as well as smaller pixels than the 1DS
MK II.

Click to expand...

Lower noise? Says who? Accordinng to dpreview it doesn't even have
lower noise than the 20d!

Click to expand...

Your right. I had looked at pictures from the IDS, not the Mark II. I don't really trust the noise charts, but if you look at the one on the Nikon D2X you'll notice the D2X, 1DS Mk II and 20D are all very close. These are great results. With film I normally shot 400 or slower. Though I did sell some concert pictures with 800 pushed to 3200. All these DSLRs blow away where film was. Look at the charts, but better look at the picutures and decide for yourself if the smaller pixels make noise much worse.

Roger Krueger said:

Roger Krueger said:

Techonology will first be used in cameras that sell. Kodak
was the leader in full frame sales, and decided it was a bad
business.

Click to expand...

Because they had a crummy camera they had to price dirt cheap to sell.

In camera history the 14mp Kodak resembles very much Kodak's
previous attempt at a world-beater camera, the Ektra of 50 years
ago. They took direct aim at Leica, built a camera that was
superior in a number of aspects, but had several dreadful flaws as
well. When the market decided it didn't want their flawed gem they
shrugged and gave up on it rather than fixing it. Were rangefinders
a bad business? Not in the 50's. Lots of people made money--just
not Kodak.

Roger Krueger said:

Canon may very well keep these large sensors to spread
marketing FUD, which may help pull sales of the 1D, 20D, 350, etc,
but I doubt it is a winning value proposition on its own.

Click to expand...

At $8000 I've gotta guess their margins are a lot better than
Kodak's were at $2700

Roger Krueger said:

The oly 4/3 lens line up does prove that smaller high quality
lenses can be made for full frame on a smaller format. Nikon is
quickly moving there lens lineup to its new full frame (which is
1.5 times smaller than canon's and 35 mm outdated definition).

Click to expand...

Calling everybody's format full frame is just childish. By your
definition every camera is full frame and the term becomes useless.

Click to expand...

I think legacy 35 mm film size does not make much of a standard in the digital age. That was my point.

Roger Krueger said:

Roger Krueger said:

The
D2X shows that high densities can be produced.

Click to expand...

Like that's a surprise.

Roger Krueger said:

So the only likely advantage of 35 mm full frame in the future is
shallower depth of field,

Click to expand...

And noise. And resolution (any noise/pixel density you can achieve
on a crop sensor can be scaled to a higher-resolution FF sensor).

Roger Krueger said:

and the exclusivity of owning a more
expensive cammera that doesn't perform as well.

Click to expand...

By what criteria? That FF shows what's been true all along, that
cheap wide-angles are soft in the corners? If you get an equivalent
angle of view in an equivalent cost lens on a crop format your
problems are at least as bad. It's just that there are a lot of
inexpensive, mediocre ways to get to 20mm on FF. There are a lot
fewer ways, especially light, cheap ways, to get to 13mm on 1.5x.

Click to expand...

Getting to these lenses is just a matter of the camera companies re-engineering there lenses to the new size. To get high quality in the next generation of sensors, I think they will have to do that any way. Your right there isn't an easy or cheap way to get a 20mm 35mm equivilant. The nikor 12-24, and oly 7-14 have been re-engineered for the sensors and represent a good compromise. Your benefit comes out on cheaper, lighter telephotos. But if you absolutely have 20mm or smaller as your highest priority then I'm out of my level of expertise.

Roger Krueger said:

Roger Krueger said:

New lenses will
probably be needed for the higher res lower noise sensors of the
future

Click to expand...

Lower noise affects lens design how?

Click to expand...

Your right that thoughts here were not clear. Higher res show more of the faults in your lenses.

Roger Krueger said:

It's a pretty safe assumption that top-of-the-line lenses are being
made as good as the market will bear now. Canon and Nikon have
never seen the need for an incredibly complicated and expensive
extreme quality 20ish lens. Contax and Leica did. Guess who's
making money? Now that Contax actually has a market for their
15-element monster, they've stopped making it. How funny is that?

Longer lens designs will probably out resolve sensors for some
time. It's just the corners of wide designs that are trouble.

Click to expand...

I mostly agree. The top of the line canon and nikons are very good, and the long primes and zooms should be fine.

But technology is already giving us 12 megapixel at a 1.5 x camera (D2X). Zooms that go to the normal or wide end that were not designed for these sensors will be the limiting factor in quality.

I totally disagree about Contax and Leica though. Bad execution, poor manufacturing practices, bad marketing, and lack of inovation are what killed these companies. Quality of lenses had nothing to do with there businesses goning down hill.

I expect canon will continue to make full frame even if they lose money on the line. From a purely marketing perspective a full frame flagship makes sense for canon.

I know it's popular to attack Americans lately, but that was a big assumption on your part, an unfounded attack (like most attacks are), and only showed your lack of tolerance of others that may be different than you or have different experiences than you.

This is not the first time I have seen accusations of Americans thinking only of themselves on this forum (all unfounded like this one), but no accusations the other way even though I routinely see Europeans saying something that doesn't apply to the US. We all seem to accept that peope are different all over the world. I guess that maybe one difference is that some Europeans are intolerant and antagonistic, the same things they seem to be accusing Americans of??

Let's all try to understand where the other guy is coming from before making remarks like that.

toughluck said:

rennie12 said:

This developed into 35mm etc etc - BUT the shape is not really
logical and normally requires considerable cropping for prints.

Click to expand...

In USA, yes. In Europe, hardly any.

Click to expand...

--
MjN

N

paulbod2 said:

Maybe Kodak FF have different microlenses than Canon, so that CA is
more under control.
Is Your C or N ?

Click to expand...

--
http://www.gavincato.com

Might be that N lenses have a larger coverage circle.. very probable.

paulbod2 said:

1) film accepts light from a wider angle than a digital sensor does
for correct reproduction

Click to expand...

This is the concept that some people keep coming back to, but the bottom line is that nobody has ever demonstrated that images from full frame digital cameras show these problems to a greater degree than film cameras. Can you see vignetting or chromatic aberation in FF digital images? Of course. Is it any worse than with a film camera? Not really.

paulbod2 said:

2) Last Photokina C mentioned the problem on photosite size (there
was an article in Let's go digital when visiting Canon booth but
cannot find it right now) please bear in mind that CMOS photosite
is surrounded by circuity

Click to expand...

Searching the Let's Go Digital website does not reveal any article of this nature.

Of course each photosite is surrounded by circuitry, but the density of circuits on a CMOS image sensor isn't anywhere close to the density of other CMOS devices. Also, the current CMOS image sensors aren't using the smallest available process sizes anyway. The bottom line is that there's still room to reduce the size of the circuitry around each photosite.

paulbod2 said:

3) Canon Cmos has small sensors but microlenses cover a larger
area, this is an obvoius highway to CA

Click to expand...

Chromatic aberation is the failure of the lens to properly focus different frequencies of light onto the same point. This is a product of the image projected by the main camera lens and it simply can't be caused by microlenses. It's either already there when it gets to the microlens, or it's not.

The simple truth is that most lenses that exhibit any significant amount of chromatic aberation do so at the corners of the image, not the center. Any full-frame image will demonstrate this, film or digital. The solution is to use better lenses, or otherwise use one of the various filters available to correct the problem. But avoiding full frame image sensors to avoid CA is a drastic overreaction. You have to keep in mind that the full frame image is inherently better quality in the first place.

paulbod2 said:

4) there is no higher cost to the manufacturer, only to the final user

Click to expand...

Where would get the idea that there's no higher cost to the manufacturer to make larger chips? Of course there is. The fact that the cost may get passed onto the final customer doesn't change that.

The manufacturing cost of any sort of chip is based solely on how much it costs to produce a wafer and how many good chips each wafer yields. All other things being equal, you'll get a higher yield with smaller chips than with bigger ones. The reason is simple. Each wafer invariably has defects on it, so the larger the chip, the more likely it is there will be a defect that causes it to be unusable. If a wafer costs X dollars, and you can get 20 functional FF chips out of it, then the cost of each FF chip is .05X. But you might get 80 APS-sized chips out of the same wafer, so the cost of each of those would be just .0125X, or 1/4 as much.

The price difference in the real world might not be proportional to my example, but it's there, nevertheless. And that cost is a factor in the design, just like everything else.

Mike

No text

I also think Canon has a major decision ahead. I suspect that they will drop the FF and go 1.3x on future pro DSLR models.

Ilkka

digital_ray_of_light said:

Canon is only making 2,000 FF DSLRs per month. This hasn't changed
from the move from Mk I to II. Whether the limiting factor is
supply or demand or a combination, I don't know.

Kodak sold 80,000 FF DSLRs in 2004, roughly three times the number
of FF DSLRs Canon produced in 2004.
http://forums.dpreview.com/forums/read.asp?forum=1000&message=13175982

In contrast, smaller (cheaper) DSLR output has sky-rocketed. Take
for example the D70. It started in the 70K/mo output, then went to
90-100k. Now the next iteration (D50 and D70s) are a combined
220k/month.
http://forums.dpreview.com/forums/read.asp?forum=1034&message=13730876

So asking the question is not ridiculous. I am not saying it will
happen, I am just saying that asking the question is not ridiculous
at all.

Peter said:

There is absolutely no link between kodak abandoning DSLR and canon
abandoning FF.

This whole thread is ridiculous.

Click to expand...

--
My brand new photography blog:
http://www.livejournal.com/users/photographyetc/

Click to expand...

Ilkka Nissilä said:

I also think Canon has a major decision ahead. I suspect that they
will drop the FF and go 1.3x on future pro DSLR models.

Ilkka

Click to expand...

All that has happened is that Canon have moved from being the producer of the most flexible FF camera, to the sole provider of this segment.
I can't see any rationale for CAnon stopping this successful product.
They can do one of two things:
Keep costs high and maximise profits,
or drop prices over time as they accuire production expertise.

Of course Moore's law does not apply, as it refers to the number of transistors not area of silicon, but just the same Canon are good at engineering cost reductions.

From a loss-leader when the 1Ds was introduced, I suspect that Canon now make very good profits on the 1DII.

Their own statements seem to indicate that they will unify the 1-series, and sice they have also said that they intend to keep all 3 sensor sizes, it will presumably be at FF, and likely sell at around $5k.
All this announcement has done is make the market for such a camera bigger
--
Regards,
DaveMart
Please see profile for equipment

some reflections :

• kodak is restructuring its activities; coming from film, they are trying to make money in digital now

they had DSLR; finally they decided not to continue in DSLR, they think there is more future for them in consumer digicams
such global decisions occur in companies of that size

• kodak doesn't have a lens line up such as canon; kodak only makes money on the camerabody

• kodak hasn't been able to take its DSLRs to a same point of "perfection" as the japanese DSLRmanufacturers

• canon may or may not continue in FF; why should their decision be influenced by kodaks decision

1Dsm2 is more expensive than kodak DSLR, more camera also
don't expect canon to loose money on their 1Dsm2 at the moment

• IMHO canon has no reason to stop FF DSLR at this moment.

• not a single professional canon lens exists specifically for the 1,6 crop factor

photogaustin said:

Roger Krueger said:

photogaustin said:

I think many of you are just forcing your wishfull thinking on
technology, instead of looking where it is going.

Canon's 350 has less noise as well as smaller pixels than the 1DS
MK II.

Click to expand...

Lower noise? Says who? Accordinng to dpreview it doesn't even have
lower noise than the 20d!

Click to expand...

Your right. I had looked at pictures from the IDS, not the Mark
II. I don't really trust the noise charts, but if you look at the
one on the Nikon D2X you'll notice the D2X, 1DS Mk II and 20D are
all very close. These are great results. With film I normally
shot 400 or slower. Though I did sell some concert pictures with
800 pushed to 3200. All these DSLRs blow away where film was.
Look at the charts, but better look at the picutures and decide for
yourself if the smaller pixels make noise much worse.

Click to expand...

Noise can be a lot worse than the 1dsII and still be a lot better than film. No arguement from me that the D2X blows away film on noise. But that's not really relevant to the FF vs. 1.5x discussion.

photogaustin said:

I think legacy 35 mm film size does not make much of a standard in
the digital age. That was my point.

Click to expand...

Is it a silly arbitrary standard? Sure. Plenty of standards become that way. But since the vast majority of lenses out there cover 24 x 36 it'll probably be a standard for some time to come.

photogaustin said:

photogaustin said:

Longer lens designs will probably out resolve sensors for some
time. It's just the corners of wide designs that are trouble.

Click to expand...

I mostly agree. The top of the line canon and nikons are very
good, and the long primes and zooms should be fine.

But technology is already giving us 12 megapixel at a 1.5 x camera
(D2X). Zooms that go to the normal or wide end that were not
designed for these sensors will be the limiting factor in quality.

I totally disagree about Contax and Leica though. Bad execution,
poor manufacturing practices, bad marketing, and lack of inovation
are what killed these companies. Quality of lenses had nothing to
do with there businesses goning down hill.

Click to expand...

Yeah, Contax and Leica both have a variety of arrows stuck in the corpse, but the point is that ultra-high end lenses, lilke Leica's wonderful but way out of reach $6,000 15/3.5, just aren't very good moneymakers. Canon's 14/2.8 isn't nearly as good a lens, but they sell enough at $1,500 to make some money at it. Canon isn't stupid. They could certainly make that caliber of lens if they chose to. They just understand that the market's not going to reward them. Big buck telephotos sell because there are a variety of paying pursuits where they're really necessary, and have no substitute. Not so for the Leica 15/3.5 or Contax 21/2.8.

photogaustin said:

I expect canon will continue to make full frame even if they lose
money on the line. From a purely marketing perspective a full
frame flagship makes sense for canon.

Click to expand...

That's another big difference vs. Kodak--Kodak didn't have a product line appropriate to bask in the glow of an FF flagship anyway. Kodak for years had the lead in uber-expensive PJ cameras, but never even came close to developing a midrange line to make real money from their reputation. They also didn't have lenses, an area where I'm guessing Canon makes far better margins than their cameras.

This is something Sony should pay very close attention to if the rumors of a Contax N resurrection have any weight. Having a gonzo flagship is pointless without a high mid-range camera like the 20d, a nice entry-level like the dRebel or D70, and a high-margin lens lineup, to actually make you money. Fixed-lens prosumer cams just don't catch much benefit from a flagship so far away.

The only company besides Canon that could remotely justify an FF flagship would be Nikon, and they've apparently said "no" for now. If Leica could make an FF M body that might sell some high-margin lenses--there are some real benefits to a rangefinder vs. an SLR--but their manual-focus-digital-SLR project has about as much commercial potential as liver-n-onion Pop Tarts.

They will propably (guessing) drop the 1.3 x camera, and leave the FF and 1.6x cameras.

Image quality wise the 1.3x camera does not bring anything special compared to say 20d. Now that Canon has dedicated lenses to size of sensor, it would not be smart to have as much cameras as possible to sell to go with them (and vice versa).

1ds (and MK2) has been a blazing succes to Canon (they kinda bought certain pro market with that camera totally), so NOT keeping it going would be silly!

Juuso

Ilkka Nissilä said:

I also think Canon has a major decision ahead. I suspect that they
will drop the FF and go 1.3x on future pro DSLR models.

Ilkka

Click to expand...

Ilkka Nissilä said:

Ilkka Nissilä said:

--
My brand new photography blog:
http://www.livejournal.com/users/photographyetc/

Click to expand...

Click to expand...

--
http://www.pbase.com/juuso

The 1DII has moved Canon into a position where they have dominance of the sports market - you don't scrap success, and the 1.3 sensor is a success.

You get most of the benefits of FF without the edge issues and at greatly reduced cost.

As Canon have repeatedly stated, they will continue to produce all three sensor sizes.

The 1.3 crop is the one which makes it easy to match the resolution of the D2x whilst still getting better High ISO.
They aren't going to drop it, or the FF either.

Juuso said:

Image quality wise the 1.3x camera does not bring anything special
compared to say 20d. Now that Canon has dedicated lenses to size of
sensor, it would not be smart to have as much cameras as possible
to sell to go with them (and vice versa).
1ds (and MK2) has been a blazing succes to Canon (they kinda bought
certain pro market with that camera totally), so NOT keeping it
going would be silly!

Juuso

Ilkka Nissilä said:

I also think Canon has a major decision ahead. I suspect that they
will drop the FF and go 1.3x on future pro DSLR models.

Ilkka

Click to expand...

Juuso said:

Juuso said:

--
My brand new photography blog:
http://www.livejournal.com/users/photographyetc/

Click to expand...

Click to expand...

--
http://www.pbase.com/juuso

Click to expand...

--
Regards,
DaveMart
Please see profile for equipment

on the other hand, the 1DmkII replacing camera could be easily made on the smaller format too, without sacrificing practically anything and gaining the usabilty of say 10-22 mm zoom.

nevertheless, i think that 1.6 and FF formats are "safe" but the 1.3 format may be dropped at some point of time.

that´s just me of course...

juuso

--
http://www.pbase.com/juuso

The only ones who don't seem to know the 1.3 format is as good as dead are Canon!

They have said whever asked that it is staying, and considering that it maximises the area you can use without running into edge issues and does so at reasonable cost it is not surprising.

They 'might' be able to get 12MP in a 1.6 sensor and still get very good high ISO, but that is all difficult, expensive, cutting edge technology.

It is hugely easier to just use a bigger sensor, and it is still nothing like as expensive as ful frame.

My guess is that their game plan is to gradually migrate the larger sensor down the chain.

Look for a camera with the same number of pixels as the D2x but much better high ISO in around the fall or spring.
What makes this relatively easy to do is the large sensor size.

All the time they have this, they have a built-in advantage over the Nikon 1.5, and Nikon have it all to do with all the difficult engineering to just keep up.

Juuso said:

on the other hand, the 1DmkII replacing camera could be easily made
on the smaller format too, without sacrificing practically anything
and gaining the usabilty of say 10-22 mm zoom.

nevertheless, i think that 1.6 and FF formats are "safe" but the
1.3 format may be dropped at some point of time.

that´s just me of course...

juuso

--
http://www.pbase.com/juuso

Click to expand...

--
Regards,
DaveMart
Please see profile for equipment

The reason is simple. Pro photographers will demand higher and higher resolutions. At some point APS-C will hit the wall. Larger FF sensors will have a lot more resolution and therefore will be preferred by pros.

I do see FF becoming a "niche" so to speak and I don't see it coming down in price at all simply because there won't be the sales volume to sustain profitability at a lower price. Nikon will also have to go FF at some point in the future when Canon releases something like 24MP FF camera and the difference in pixel count in top level bodies becomes truly ridiculous.

David Martin said:

The only ones who don't seem to know the 1.3 format is as good as
dead are Canon!

Click to expand...

Why not? Take the 24─70/2.8 lens. On a 1.3x format (which is actually more like 1.25x), it becomes ca. 30─87.5 equiv. FOV. Still VERY nice for general photojournalism, and without the edge issues. Take the 17─40/4 or 16─35/2.8. They make for (note the word 'for') very nice 21─50/4 and 20─44/2.8 respectively. They are still wide angle without the edge issues. A note for Canon:

Put the 135 format sensor into a future EOS 1Dn. Make it 16 megapixel @8 fps, or a slightly cropped mode (akin to the Sony CMOS sensor in the Nikon D2x), to 10-12 mpix, and that @8 fps (or even more). You'll still have a winner. And put that future 12 megapixel (@5 fps) 1.3x crop sensor into a EOS 3 class of body. Great for PJ, wedding, product, sports, etc., without the premium.

Member said:

They 'might' be able to get 12MP in a 1.6 sensor and still get very
good high ISO, but that is all difficult, expensive, cutting edge
technology.

Click to expand...

Just as it was unthinkable to put over 10 megapixels in an APS-c sized sensor. Voila, enter the Nikon D2x. "Cutting edge," indeed. For year 2005. In year 2006 it will be on the shelves, and in 2007, it will be yesterday's news. Canon (with their 20D sensor) already prove that it is possible to make a sensor with photosites the size of the ones in Olympus E-1 sensor (and even smaller─interline vs. full frame transfers), and much smoother, with more dynamic range. Unthinkable just yesterday.

Member said:

My guess is that their game plan is to gradually migrate the larger
sensor down the chain.

Click to expand...

Forget it. Too expensive if 1.6x provides for enough image quality, and will always be much cheaper than 135 format. Look at it this way:

On a circular wafer, you can etch some 6─8 135 format sensors. In the same area, you can fit two times as much APS-c sized sensors, plus you get more use of the periphery, so imagine you can stack over 20 APS-c sensors on the same wafer. Now, let's say there are 6 flaws on the wafer. The yield of 135 format sensors will be anything between 25% (2 sensors) and 87% (7 sensors). Much nearer the first figure, anyway, since there are many wafers, and flaws are not usually concentrated. With those 20 APS-c sized sensors, you can get anything between 70% (14 sensors) and 95% (19 sensors). With 1.25x sensors, I believe they could fit some 12 sensors (50% yield, 6 sensors). So, assuming that:
─ we take the first figures;
─ the cost of making sensors is the same for small and large ones;
─ the cost of processing a wafer is fixed at $4,000.
We get:
─ 14 APS-c sized sensors, $286 each, or:
─ 6 1.25x crop sensors, $667 each,
─ 2 135 format sensors, $2,000 each.

Migrate down the chain indeed. And there's no point in making just slightly larger sensors. You'd get incompatibility for future bodies. Want wide angle? Much easier to make one lens fit all sensors (same size), than many sensors for backwards compatibility with older lenses.

Most photographers, anyway, have already bit the bullet, and purchased those EF-S lenses. There is no more legacy in keeping those older lenses in your pouch. You just have alternatives now.

Member said:

Look for a camera with the same number of pixels as the D2x but
much better high ISO in around the fall or spring.

Click to expand...

Very true.

Member said:

What makes this relatively easy to do is the large sensor size.

Click to expand...

Relatively easy, but absolutely MUCH more expensive. Especially to the end user, who has already invested in EF-S lenses, and now learns that his new Canon 25D, with 1.4x crop sensor, does not support the EF-S mount, and forcing it causes the mirror to shatter on impact with the rear element on the lens.

Member said:

All the time they have this, they have a built-in advantage over
the Nikon 1.5, and Nikon have it all to do with all the difficult
engineering to just keep up.

Click to expand...

Keep up? Their 12 megapixel D2x is ENOUGH for MOST uses. For those that it is not enough for, the 16 megapixel EOS 1Ds2 is probably ALSO NOT enough. Even the 25 megapixel MF backs might not be enough. We're talking 4x5" or 8x10" film here.

What do you need those 16 megapixels for anyway? What do you need that 15% greater resolution for? Is it worth the $3,000 (66%) more? I don't think so.