Focal plane vs. leaf shutters in MF SLRs

I think it's worthwhile to point out that there's also a 3D MF subculture out
there shooting with Sputniks, ancient Rolleidoscops and twinned MF cameras.
In additon I've read that there's a new MF 3d camera due out of China this
year with a $1300 price tag. They're also selling a coin operated viewer
(you can just guess what that's for). From what I've read 3d MF slides are
breathtaking. I've done superslide 3d, but have not spent the bucks for a
full frame (i.e. 50mmX50mm) 3d viewer.

the key point y'all missed is that it looks unlikely that a 35mm format
64MP sensor is likely, based on CMOS developer Carver Mead's comments at
end of article fundamental size limits in wavelength of light see
http://www.dpreview.com/news/0009/00...foveon16mp.asp

the 16.8MP foveon chip is 22mmx22mm square format ;-) now, so 64MP would
be 44x44mm square, right? The trend is towards larger sensor sites to
improve images and reduce noise. The Fuji 22 MP sensor is circa 40x51mm; a
32 MP sensor would be decidedly MF sized and not 24x36mm.

Most of us shooting MF SLRs with interchangeable backs also have full kits
of 35mm SLRs, including the pros.

So if 64MP becomes available, I see it requiring MF sized cameras and
lenses, and not 35mm SLR sized chips, even full frame ;-) What I don't see
is any "need" or killer application which requires 64MP sized chips. Now
if HDTV cameras required a 64MP or larger sensor, I would expect cheap MF
digital backs as a side product. But they don't. Other than specialty
military/science applications, I don't see a requirement which will
justify the multi-billion dollar fabrication facilities needed to make
42x56mm or 56x56mm sized sensor devices. Do you?

Again, my argument is that the 35mm SLRs, as QGdeB noted, are far more
popular than the more pricey 16MP digital backs for MF. If 16MP offered a
serious advantage over 8 MP or 11 MP for digital users, then I would
expect to see a lot more digital back users with 16MP backs. And we don't.

So I argue that there is a series of "sweet spots" here, around 2 MP for
consumers with cell phone cameras for 90% of photos to websites, and 4-5
MP for $200 cameras for family digicams doing home prints up to 11x14" or
so, and a modest market (in terms of multi-million $ R&D and chip plant
costs) for higher end DSLRs up to 11 to 16 MP or so. There isn't and
hasn't been much demand for 16MP sensor MF digital back images, right? So
where is the demand for 64MP images which will drive the mfgers to make
billion dollar investments to provide such chips in high enough volumes to
make 64 MP digital backs possible?

In short, I see MF gear being needed beyond 16MP (cf Fuji's 22 MP sensor).
That's the good news. The bad news is I don't see enough of a market for
such gear (today, or in future) from the handful of MF using pros to
justify the mfger costs to deliver digital backs in high enough volumes to
get the custom made costs down to mass produced limits.

In the above article, National Semiconductor figures they can make 16MP
chips for "disposable" digital cameras as cheaply as today's disposable
film cameras (actually recycled), i.e., under $10 or so each 16MP CMOS
sensor chip. The problem is that a 64MP chip at today's volumes will still
cost $10,000, not $40 ;-)

my $.02 again

bobm
--
************************************************** *********************
* Robert Monaghan POB 752182 Southern Methodist Univ. Dallas Tx 75275 *
********************Standard Disclaimers Apply*************************

digital isn't film ;-) with film, you can simply cut the large rolls of
film into 120 rollfilm, or 4x5" or 8x10" sheets or 35mm film, as needed.

With digital, you have to build multi-billion dollar IC fab plants to make
a 16MP or even more for a 64MP CMOS sensor chip. You can't afford to build
a multi-billion dollar 64MP sensor making plant, or spend tens of millions
on integrating that 64MP sensor into a digital camera back with software
etc., unless you have a market that can repay those investment costs.

If national semi's CEO is right, and they can sell 100 million annually of
Foveon style CMOS 16MP sensor chips, then they can get the costs down to
$10 or less (even $2-3 per chip, making "disposable 16MP digital cameras"
feasible to compete with today's disposable $10 film cameras. Again, see
http://www.dpreview.com/news/0009/00...foveon16mp.asp article etc.

The same isn't true of a 64MP sensor, for reasons I suggested. That is why
you can't say what happens with cell phone cameras and 35mm DSLRs is
unconnected to what we can do with digital in the MF niche market. MF is
niche market (50,000 film cameras sold per year worldwide, maybe 1-2% of
them with 16MP digital backs costing $10-25k, yes?). You can't justify
billion dollar plants and tens of millions in R&D on digital back sales in
MF (look at costs of adapting existing sensors in today's digital backs
($25k..) vs. costs of sensors in somewhat higher volume Kodak DSLRs for
35mm format etc.

So nobody is going to make multi-millions or billions in investments in
order to provide MF users with low cost 16/32/64+MP digital backs ;-)

So we _are_ dependent on what the masses do, because IC sensor mfgering is
a really mass production process. It costs as much to build a IC plant to
make a thousand 64MP chips as 100 million 16MP chips, but that implies a
$1 million cost per 64MP chip vs $10 for " " ;-) ;-) Can we afford that?

Is there a mass volume application which justifies and demands 32MP and
64MP chips? So far even pro photographers seem satisfied with 8 MP and 11
MP DSLRs vs 16MP digital backs at $25k or so. Unless there is such a
demand and market, I doubt 64 MP chips will be developed.

In short, QGdeB's argument that what happens in the mass consumer market
doesn't limit us on the MF niche market is false, because sensor chips
require huge investments in R&D and IC fab plants which mandate huge sales
volumes to get the costs down.

I don't see any such huge mass market or demand for 64MP sized devices, so
you? If there is one, then the good news is that we can probably hope for
low cost 64MP sensors of MF sizes which can be adapted for MF digital
backs. These 64MP sensors probably won't be compatible with 35mm smaller
formats (cf. Fuji's 40+x51+mm chip for 22 MP sinar backs etc.).

So high end (22+MP) digital photography may be the preserve of MF or
larger systems? The problem is consumers and even pro photographers are
happy enough at 8 or 11MP to make 16MP cameras a likely "sweet spot". So
there won't be any photographic demand for larger resolution digital
cameras past 16 MP or so?

The caveat here is that it may be possible to synthesize a 32MP or larger
equiv. sensor using multiple cheaper 16MP chips (or a 48 MP in 35mm format
using tri color filtering etc.). But a native 64MP chip does not look to
have the manufacturing volumes needed to make them cheap due to lack of
demand?

my $.02+ '-)

bobm
--
************************************************** *********************
* Robert Monaghan POB 752182 Southern Methodist Univ. Dallas Tx 75275 *
********************Standard Disclaimers Apply*************************

Hi Michael!

Let us say that present trends in scanner improvements continue. And that
future color printers can produce rather larger prints inexpensively, as
picoliter technology is implemented better. Films might also continue to
be improved to be better matched to scanner's needs, producing higher
resolution and Dmax and so on. Sounds reasonable, doesn't it?

Now you can buy a $750 MF camera with auto-exposure and shoot 645 slides,
then blow them up optically or scan them to 30MP or 64MP or better
resolutions and make large poster prints. Perhaps Kinkos would have a
large printer where you make poster sized prints from your CDROM.

Or you can spend $1k+ on a DSLR and generate an 8MP or 11MP or at best a
16MP image (again, per Foveon's analysis, see
http://www.dpreview.com/news/0009/00...foveon16mp.asp )

Presumably, a 30-64MP sensor equiv. scan can generate a larger poster
print than the 8 to 11MP (or 16MP) sensor DSLR, yes? As I have noted in
this thread, I don't expect 64MP sensors. But a film scanner that can
produce clean 64MP images from MF film is feasible, yes? ;-)
===============


We have a new series of color printers able to use much larger widths, and
unlimited lengths in some cases, of paper to produce color prints. I think
as larger MP digicams come available, people will want to produce larger
prints, including panoramics and wall sized photos, if prices are right.
Today, you can get 11x14" digital prints for rather less $ than an optical
print, and even less in larger (panoramic) formats. I suggest that a
future 20x30" print could cost less than today's 11x14" print with new
printer technology, and even 40x60" might be possible - with the demand
coming from thousands of kiosks and companies wanting to be able to do
posters and large prints for banners, conferences, and all that ;-)

The key point is that 35mm sized DSLRs of 8Mp or 11MP or the max of 16MP
still can't match film's potential resolution, which future scanners can
tap, and provide much larger digital prints (just as larger optical prints
are possible today, yes?).

So MF continues to enjoy an enlargeability and quality gap in the future
;-) Moreover, as larger print making costs decline, the demand for larger
print capability may promote the use of MF with film scanners to achieve
those needs, yes? ;-)

grins bobm

PS - the counterargument? in some cases, you can merge multiple shots (not
action shots, but many subjects are fixed) from an 8 MP camera to achieve
16 MP or larger equiv. image quality without going to film and a scanner.

--
************************************************** *********************
* Robert Monaghan POB 752182 Southern Methodist Univ. Dallas Tx 75275 *
********************Standard Disclaimers Apply*************************

Bob Monaghan wrote:

Today, you can get 11x14" digital prints for rather less $ than an optical
print, and even less in larger (panoramic) formats. I suggest that a
future 20x30" print could cost less than today's 11x14" print with new
printer technology, and even 40x60" might be possible - with the demand
coming from thousands of kiosks and companies wanting to be able to do
posters and large prints for banners, conferences, and all that ;-)


The problem with this hope is that it requires large paper to be kept in
stock. Not only do you need to keep multiple paper sizes in stock you'll
need too change paper rolls. It stops being cheap when you need a human.


Nick

On 23 May 2004 22:21:57 -0500, (Bob Monaghan)
wrote:


the key point y'all missed is that it looks unlikely that a 35mm format
64MP sensor is likely, based on CMOS developer Carver Mead's comments at
end of article fundamental size limits in wavelength of light see
http://www.dpreview.com/news/0009/00...foveon16mp.asp


And who appointed Carver Mead as the authority on this topic?
Carver's got a specific product to sell, and so far it's been a
very hard sell. Foveon continues to play a very small role in
the digicam market.

I think MF and LF may continue to have a role in niche
applications. It will be many years (if ever) before a silicon
sensor can return the sort of pixel counts that I get from
scanning either of these. The only real issue I see is
how long will Kodak and Fuji (et al) continue to make
film in these formats?


rafe b.
http://www.terrapinphoto.com

On 23 May 2004 23:15:14 -0500, (Bob Monaghan)
wrote:


We have a new series of color printers able to use much larger widths, and
unlimited lengths in some cases, of paper to produce color prints. I think
as larger MP digicams come available, people will want to produce larger
prints, including panoramics and wall sized photos, if prices are right.
Today, you can get 11x14" digital prints for rather less $ than an optical
print, and even less in larger (panoramic) formats. I suggest that a
future 20x30" print could cost less than today's 11x14" print with new
printer technology, and even 40x60" might be possible - with the demand
coming from thousands of kiosks and companies wanting to be able to do
posters and large prints for banners, conferences, and all that ;-)

The key point is that 35mm sized DSLRs of 8Mp or 11MP or the max of 16MP
still can't match film's potential resolution, which future scanners can
tap, and provide much larger digital prints (just as larger optical prints
are possible today, yes?).

So MF continues to enjoy an enlargeability and quality gap in the future
;-) Moreover, as larger print making costs decline, the demand for larger
print capability may promote the use of MF with film scanners to achieve
those needs, yes? ;-)



The larger printers are great fun but not cheap.
Hard to believe that your typical "consumer" will ever
learn how to make satisfactory 16x20" prints at home,
or learn how to scan MF film or invest in the gear to do
any of that. That's just dreamin'.

Tabloid printers (13x19") are available for the
desktop starting at around $400 new, but that is
still rather expensive. A printer to do 16x20" will
cost you $1500 to $1800 - hardly a commodity item.

Quality isn't the only factor that drives markets, it's
often not even near the top of the list. For photography,
convenience and portability are *huge* factors, and
that applies nearly across the board, except maybe for
the very highest end studio gear.


rafe b.
http://www.terrapinphoto.com

Raphael Bustin wrote:

On 23 May 2004 22:21:57 -0500, (Bob Monaghan)
wrote:


the key point y'all missed is that it looks unlikely that a 35mm format
64MP sensor is likely, based on CMOS developer Carver Mead's comments at
end of article fundamental size limits in wavelength of light see
http://www.dpreview.com/news/0009/00...foveon16mp.asp

And who appointed Carver Mead as the authority on this topic?
Carver's got a specific product to sell, and so far it's been a
very hard sell. Foveon continues to play a very small role in
the digicam market.

He's got the physics on his side. The wavelength of light isn't
changing. You can't use smaller than five micron pixels, and even those
are pretty noisy. 64MP with 7x7 micron dots is 6x6cm in size, even
without the electronics between the pixels.

I think MF and LF may continue to have a role in niche
applications. It will be many years (if ever) before a silicon
sensor can return the sort of pixel counts that I get from
scanning either of these. The only real issue I see is
how long will Kodak and Fuji (et al) continue to make
film in these formats?

Film will be competitive for a long time. When you start handling 64MP
(or 400MB) images, you'll soon notice some things:
- memory isn't cheap, and you need lots of it
- digital processing needs memory all the way from scanning workstations
to permanent storage
- permanent storage isn't permanent, unless it is refreshed every five
years or so
- the rest of the equipment gets obsolete even faster...

In terms of total cost of ownership, MF is still hard to beat. Digitals
win when you need the speed.

-- Lassi

Hi,

Recently, Lassi Hippeläinen posted:

In terms of total cost of ownership, MF is still hard to beat.
Digitals win when you need the speed.

I agree with half of this... the cost of MF is hard to beat for creating
high-quality images. However, speed is not always in digital's favor. As I
pointed out in another post, downsampling a 64 MP image to be used at, for
example, 3" x 3" in a publication can take much longer than scanning MF
film to that finished size. Combined with the archival issues, digital's
advantages are in producing images for short-term uses and for unique
features such as the ability to shoot at 4-5 fps without having to change
film / backs.

Neil

Bob Monaghan wrote:

the key point y'all missed is that it looks unlikely that a 35mm format
64MP sensor is likely, based on CMOS developer Carver Mead's comments at
end of article fundamental size limits in wavelength of light see
http://www.dpreview.com/news/0009/00...foveon16mp.asp

Carver Mead is enjoying great visionary succes lately, isn't he?

It was deemed very unlikely that people would ever fly. Or that anybody
would ever need more than 512 kB of memory. Or that "still-video" would ever
replace film. Or...
;-)

Anyway, perhaps sensors must grow to be able to produce more pixels.
But unless that will happen tomorrow, it will be too late to help keep MF
alive today.
So those 4x4 cm 64MP sensors will have to be put into something which will
look stunningly like a 35 mm format SLR camera of old (have you ever
compared the sizes of a Nikon F5 and a Mamiya 645? ;-)), or perhaps one of
those "bridge" cameras produced in the late 70s.
Not in a MF digiback, since there will not be a MF camera left to put that
on.

The above is perhaps a bit "over the top", but there's nothing wrong with
the realism of notion it is expressing.
We will not need MF cameras. They will simply put another housing (perhaps
even a telephone ;-)) around that chip. Put a lens mount on it, a display on
the back, and Bob's your uncle.

[...] popular than the more pricey 16MP digital backs for MF. If 16MP
offered a
serious advantage over 8 MP or 11 MP for digital users, then I would
expect to see a lot more digital back users with 16MP backs. And we don't.

But not (!) because there is no advantage. Only because we are expected to
pay far too much for it. The things as the are today are way out of
proprotion. That's why the lower end will win. That's why the lower end will
bring in the revenue that will help the lower end to gain even more of an
advantage over the upper end by becoming more and more like the upper end
while at the same time holding on to its competitive edge (disproportionally
low price), which will increase the low end's winning potential no end. Etc.

The only thing to break this spiral would be a rather huge correction of the
high end's prices.
If that will not happen, the current MF and LF oriented high end will simply
cease to be.

I guess we fundamentally disagree. I do believe there will definitely not be
a place for MF in the future unless they start carving it out today.
Relying on present technology's shortcomings to ensure a MF future is the
most foolish thing MF industry (both camera- and digital back manufacturers)
can do. It would indeed be so stupid that not even shouting at the top of
your voice could put enough stress on that simple fact.

In the above article, National Semiconductor figures they can make 16MP
chips for "disposable" digital cameras as cheaply as today's disposable
film cameras (actually recycled), i.e., under $10 or so each 16MP CMOS
sensor chip. The problem is that a 64MP chip at today's volumes will still
cost $10,000, not $40 ;-)

They also though the D-Finity was a break-through product. And that with
their invention of a three layer chip, noone would want to buy those crappy
Canons and Nikons anymore.
;-)