New 12 meg Nikon D200 Specification

Archived from groups: rec.photo.digital (More info?)

Hopefully, it's not true. The D200 with a CMOS sensor means losing the
fast flash sync of the D70. Why always CMOS? So far, Nikon did not turn
into a Me-Too company.

Gregor

Archived from groups: rec.photo.digital (More info?)

On Thu, 18 Aug 2005 21:19:13 +0100, deryck lant wrote:

> Sensitivity - ISO equivalency 100 to 800

That seems increadibly low.

-SL

Archived from groups: rec.photo.digital (More info?)

In article <1orjwr0xf8rpx$.zfz4raa0kgp9.dlg@40tude.net>,
No.Spam@Thank.You says...
> > Sensitivity - ISO equivalency 100 to 800
>
> That seems increadibly low.

That's the standard sensitivity on the D2X.
--
http://www.pbase.com/bcbaird

Archived from groups: rec.photo.digital (More info?)

> The D200 with a CMOS sensor means losing the
> fast flash sync of the D70.

Hopefully, it'll have FP flash-sync, useful for
outdoor fill work.

-- stan

Archived from groups: rec.photo.digital (More info?)

"Skip M" <shadowcatcher@cox.net> wrote:
> "deryck lant" <deryck@deryck.com> wrote in message
>> http://alpseb.free.fr/D200/NikonImagingNikonD200.htm
>> http://www.infodigitalcamera.com/b [...] ication-2/
>>
>> !2meg
>> 3fps at 12meg for 10 Raw
>> 5fps at 6meg for 18 Raw in cropped mode
>> Same AF as D2X - WOW!!
>> 2 1/2 inch CCD
>> 810 grams
>>
>> Its the D2X in the F6 body

I think not: the F6 is 975 grams.

>> The dream has come true - a lightweight D2X
>
> Sounds like Nikon has been reading the alleged specs of the supposedly
> upcoming Canon 5D, or the other way 'round. Except for the Nikon's
> variable res/fps. Should be an interesting matchup, if, indeed, they both
> are real cameras...

A lightweight, US$1500 camera with the D2X chip in it is _far_ more likely
than a US$3000 full frame camera, this year anyway. (I'd think the specs
would be closer to the D70 than the D2x, but there's really nothing wrong
with the D70 specs.)

David J. Littleboy
Tokyo, Japan

Archived from groups: rec.photo.digital (More info?)

In article <0duag1pu0i50n2m6t32cp5nuqdf6qh1st3@4ax.com>,
Father Kodak <dont_bother@IDontCare.COM> wrote:
>There seems to be at least one difference. The "additional
>information" refers to support of Nikon AF lenses. Nothing about MF
>lens, etc, which do get supported with metering on the D2X.

If the D200 gets the F80 body that forms also the basis of the D100,
D70, and the D50 then the usual limitation of using MF lenses will apply.

The D2X is already a digital F6. A consumer DSLR will of course get a
consumer body, and that is the F80.


--
That was it. Done. The faulty Monk was turned out into the desert where it
could believe what it liked, including the idea that it had been hard done
by. It was allowed to keep its horse, since horses were so cheap to make.
-- Douglas Adams in Dirk Gently's Holistic Detective Agency

Archived from groups: rec.photo.digital (More info?)

Odd how both Canon and Nikon had specs for their new cameras on their
websites for about 10 minutes. The only sources always seem to be 3rd party
sites.

"deryck lant" <deryck@deryck.com> wrote in message
news:31303030313839354304FB5155@deryck.com...
> http://alpseb.free.fr/D200/NikonImagingNikonD200.htm
> http://www.infodigitalcamera.com/b [...] ication-2/
>
> !2meg
> 3fps at 12meg for 10 Raw
> 5fps at 6meg for 18 Raw in cropped mode
> Same AF as D2X - WOW!!
> 2 1/2 inch CCD
> 810 grams
>
> Its the D2X in the F6 body
>
> The dream has come true - a lightweight D2X

Archived from groups: rec.photo.digital (More info?)

This is good news. Now Nikon and Canon will have competitive price
pressure. Even if the Canon 5D is full frame, for alot of
people, 12 Mpix is 12 Mpix and I don't think Canon could get away with
a huge price premium for full frame.....at least I hope not.

Archived from groups: rec.photo.digital (More info?)

The message <YMcNe.1485$sw6.1136@fed1read05>
from "Skip M" <shadowcatcher@cox.net> contains these words:

> "deryck lant" <deryck@deryck.com> wrote in message
> news:31303030313839354304FB5155@deryck.com...
> > http://alpseb.free.fr/D200/NikonImagingNikonD200.htm
> > http://www.infodigitalcamera.com/b [...] ication-2/
> >
> > !2meg
> > 3fps at 12meg for 10 Raw
> > 5fps at 6meg for 18 Raw in cropped mode
> > Same AF as D2X - WOW!!
> > 2 1/2 inch CCD
> > 810 grams
> >
> > Its the D2X in the F6 body
> >
> > The dream has come true - a lightweight D2X

> Sounds like Nikon has been reading the alleged specs of the supposedly
> upcoming Canon 5D, or the other way 'round. Except for the Nikon's
> variable
> res/fps. Should be an interesting matchup, if, indeed, they both are real
> cameras...

IFA Show Berlin Sept 2-7 will reveal all.

Deryck

Archived from groups: rec.photo.digital (More info?)

In article <1124474578.873204.144450@o13g2000cwo.googlegroups.com>,
winhag@yahoo.com says...
> I have no idea as to the price. I did see a sub $2k guess I think it
> was on Ken Rockwell's site?

That would shoot down D2X sales... but then again, Nikon needs something
to compete with the 20D. They haven't had a higher end prosumer camera
since the D100.

I don't think Sony/Nikon can do a 12 megapixel sensor for sub $2,000,
but if they do I'll be pleasantly surprised.
--
http://www.pbase.com/bcbaird

Archived from groups: rec.photo.digital (More info?)

Brian Baird wrote:
> In article <1124474578.873204.144450@o13g2000cwo.googlegroups.com>,
> winhag@yahoo.com says...
>
>>I have no idea as to the price. I did see a sub $2k guess I think it
>>was on Ken Rockwell's site?
>
>
> That would shoot down D2X sales... but then again, Nikon needs something
> to compete with the 20D. They haven't had a higher end prosumer camera
> since the D100.
>
> I don't think Sony/Nikon can do a 12 megapixel sensor for sub $2,000,
> but if they do I'll be pleasantly surprised.


Well, I guess, it would be about $3500 then.

Archived from groups: rec.photo.digital (More info?)

>Brian Baird writes ...
>
>I don't think Sony/Nikon can do a 12 megapixel sensor for sub
>$2,000, but if they do I'll be pleasantly surprised.

Why not? Most of the cost is due to the actual size of the sensor in
area, with a very small yield loss in a higher pixel count chip due to
tighter geometries, but very slight. In other words, a 1.5x sized
sensor with 6 Mpixels should cost only pennies more to manufacture than
a 1.5x sized sensor with 12 Mpixels (obviously it cost something to
design the 12 Mpix version, but I'm talking manufacturing costs here).
Or to put it a different way, it will cost them less to make this 12
Mpixel 1.5x sensor than it costs Canon to make the 8 Mpixel 1.3x sensor
used in the 1D Mark II because the Mark II sensor is larger in area.

Bill

Archived from groups: rec.photo.digital (More info?)

"Brian Baird" <no@no.thank.u> wrote in message
news:MPG.1d70220425366ca39898d6@news.verizon.net...
> In article <1124474578.873204.144450@o13g2000cwo.googlegroups.com>,
> winhag@yahoo.com says...
>> I have no idea as to the price. I did see a sub $2k guess I think it
>> was on Ken Rockwell's site?
>
> That would shoot down D2X sales... but then again, Nikon needs something
> to compete with the 20D. They haven't had a higher end prosumer camera
> since the D100.
>
The sales numbers are pretty low on the 20D these days. The 20D and D200 are
a tiny market segment. Consumers are heading for the sub $1000 (usd) dSLR.
All industry trend figures bear that out.

Archived from groups: rec.photo.digital (More info?)

> Brian Baird writes
>
>But more photosites = more chance of manufacturing defect =
>higher production costs.

But the geometries of these sensors are so loose compared to typical IC
chips like memory or logic chips that it won't make a difference. They
are putting a few million sensor sites on a relatively large chip ...
compare to a RAM chip with hundreds of millions of bits, each bit
requiring several gates (devices), jammed into a die smaller than this
sensor ... I doubt it will cost them a nickel more per chip to go from
8 to 12 Mpixels. I designed chips for 20 years and the main cost
factor was always the die size.

>I think Nikon would be shooting themselves in the foot if they
>offer the D2X's sensor in a camera a third of the price.

Canon offers the 8 Mpixel D20 for about 1/3 the cost of the 8 Mpixel 1D
Mark II (I know the sensor is larger on the M II but the performance of
these two sensors isn't wildly different) and while they sell a lot
more D20's than Mark II's at these prices (and a LOT more Rebels than
either, due to price) there is still a strong demand for the M II
because it fills a different market niche.

Bill

Archived from groups: rec.photo.digital (More info?)

In article <de6d1n$2pd$1@nnrp.gol.com>, davidjl@gol.com says...
> I don't see any problem with 12MP for under US$900.

Only time will tell.
--
http://www.pbase.com/bcbaird

Archived from groups: rec.photo.digital (More info?)

On Fri, 19 Aug 2005 22:02:25 -0400, "Darrell" <spam@this.eh> wrote:

>"Bart van der Wolf" <bvdwolf@no.spam> wrote in message
>news:43066606$0$11080$e4fe514c@news.xs4all.nl...
>>
>> "l e o" <someone@somewhere.net> wrote in message
>> news:iPoNe.7544$Wi6.6632@newsread2.news.pas.earthlink.net...
>> SNIP
>>> It's very odd that Darrell had no doubt about this Nikon leak. Because
>>> it's not Canon? Very funny.
>>
>> ;-) Maybe the Nikon leak is also a hoax? ;-)
>>
>I said, if you read my post. Was both leaks are always on a unrelated third
>party site. The Canon "leak" was on some personal Romanian website, the
>Nikon was on an French website with broken links. The D200 trademark has
>been registered (July 2005, USPTO.gov) My speculation weeks ago was a new
>version of the 12.4 D2X sensor. Because of economy of scale the second
>generation 12.4 CMOS would be cheaper to fabricate. This is what Canon did
>with the 10D and 300D. The 350D/XT is the second generation of the 20D CMOS.

Then Nikon could lower the cost of their 12.4 MP sensor still more by
increasing volumes by ... "slipstreaming" that sensor into new models
of the D50 and D70s. and maybe even the D2Hs.

Pere Kodak

Archived from groups: rec.photo.digital (More info?)

"Bill Hilton" <bhilton665@aol.com> wrote in message
news:1124505233.853744.143660@o13g2000cwo.googlegroups.com...
>> Brian Baird writes
>>
>>But more photosites = more chance of manufacturing defect =
>>higher production costs.
>
> But the geometries of these sensors are so loose compared to
> typical IC chips like memory or logic chips that it won't make
> a difference.

I don't know a thing about CMOS imagers or CCDs, but it
seems naive to conclude "it's CMOS, therefore it obeys
the very same geometry/economy/transistor-budget rules of
other CMOS products."

For example, chips that are primarily "Logic" (most chips
in consumer electronic devices) undergo different product-
development and design schedules, versus memory semiconductors
(flash, DRAM, SRAM, etc.) They're all CMOS, but each has its
own nuance.

One of the larger CMOS devices in recent times was the
original Playstation 2's GS chip (graphics synthesizer.)
The large on-chip embedded-DRAM pushed the overall die-area
past 300mm^2, and yet the entire console could be sold for
$299 USD (initially at a slight loss.) (Intel's Itanium CPU
is even larger, but that's hardly a high-volume consumer
product.)

Yet the largest mass-produced DDR semiconductors are a
fraction of that size -- to make them larger would reduce return
on investment. (And DRAMs are already architected with
'spare' rows for defect-substitution after testing.)

What's the sweet spot for a CCD or CMOS imager? I don't know...
For better or worse, the digicam market is headed toward ever
shrinking imagers. (Well, worse for low-light performance!)

> They are putting a few million sensor sites on a relatively
> large chip ...
> compare to a RAM chip with hundreds of millions of bits, each bit
> requiring several gates (devices), jammed into a die smaller than this
> sensor ... I doubt it will cost them a nickel more per chip to go from
> 8 to 12 Mpixels. I designed chips for 20 years and the main cost
> factor was always the die size.

I've only been in the ASIC industry for the past 5 years, but
it's pretty obvious to me that Moore's Law has outpaced
what logic designers can effectively do with extra transistors.
Specifically, doing larger designs requires better (EDA) software
tools, and incurs larger NRE costs (like the photomask set).
If products had an infinite product lifespan, then companies could
just ignore the one-time R&D and manufacturing-setup costs...
Finite product lifespan, higher NRE costs, etc., now "smaller" isn't
necessarily cheaper (over the lifecycle of the product.)

There are a lot of examples of mass-produced consumer electronic products
fabbed at less than state of the art foundry-nodes (such as 0.18/0.15u),
because transitioning them to the next smaller node (0.13u or below)
won't increase profit. (DVD players and many other portable electronics
are good examples. Set-top boxes, DVD recorders, DVRs, tend to use
silicon closer to the 'cutting edge'.)

Superficially, it may seem like imagers are a more straightforward
(though still challenging) design task, as imager chips have
inherently regular structure. But the imager vendors have to suffer
through the same manufacturing difficulties faced by the rest of
the ASIC industry.

I'm not saying Nikon/Canon can't/won't push progress forward. I
think at some point, they'll hit a practical limit on the sensor
specifications, and simply have to innovate 'horizontally'. You know,
add useless luxuries like a built-in cellphone, WiFi connectivity,
etc. (or combine a high Mpix stillcam into HDV camcorder...ugh!)

Archived from groups: rec.photo.digital (More info?)

> Padre Kodak inquires ...
>
>Out of curiosity, is the main cost issue with a full-frame sensor the
>greater chance of a defect or the lower number of sensors per disk
>of raw silicon?

Greater chance of a fatal defect ... here's a short course in
semi-conductor economics that illustrates why ... assume you have a
wafer (raw silicon starting material) that's say 8 inches diameter, so
there are roughly 50 sq-in of area (in real life some of the edge area
is wasted but to make the math easier let's say 50 sq-in). Say they
have a chip that's .1 x .1 inches in size so they get 100 in a square
inch or 5,000 total possible sites on the entire wafer with this chip,
so they use this small chip to calculate the overall defect density for
the entire process.

Say that they run many wafer lots with this chip and find the average
good die per wafer is 4,950 with 50 bad ones. So the average defect
density is one fatal defect per sq-in.

Now say we designed a chip for the mini-Hasselblad that's 1x1" or 1
sq-in, so there are 50 possible die sites on an 8" wafer. But the
average defect density is 1 per sq-in so in theory we could get zero
yield if the defects are spread out just so. In practice you'd get a
few good die because the 50 errors aren't spread out evenly across the
entire wafer (they were probably caused by a few locally catastrophic
rips or tears in photo-resist or similar) but you likely won't get a
lot of good ones with a design this large. Anyway, for the sake of
argument say you are getting 1 good die on the entire wafer at this die
size because two of the fatal flaws happen inside one die site, leaving
one site error-free ... now if you reduce the die size in half you have
100 possibles instead of 50 and you should get at least 50 good die
(chips) instead of zero or slightly more than zero. If you reduce the
die size to 1/4 the area you have 200 possibles and at least 150 should
be good. So if the wafer cost is $500 then the raw cost of one good
large chip is $500 but for a chip half that area it's $10 and a chip
1/4 that area it's $3.33. That's why I spent many years working on
getting chip sizes smaller

So with a linear decrease in die size you get exponential increases in
yield and lower costs. That's the problem the full-sized sensors face
(which are actually greater than 1.5 sq-in once you add in the
peripheral circuitry and bonding pad), they are not that far from the
near-zero good die part of the yield curve.

What Brian and I were disagreeing about was the effect on the defect
density of going from say an 8 micron pixel pitch to a 5.5 micron pixel
pitch as you jam more sensor sites into the same area. My point was
that most of these CMOS processes were designed for much smaller
geometries for logic and memory chips, say .2 microns or less, so you
won't see a big change in the defect density number.

Bill

Archived from groups: rec.photo.digital (More info?)

deryck lant <deryck@deryck.com> wrote:
> http://alpseb.free.fr/D200/NikonImagingNikonD200.htm
> http://www.infodigitalcamera.com/b [...] ication-2/

Do you think Sony will make the CMOS sensor for the D200,
as they did for the D2X?