D400 discussion continuation

Started May 26, 2009 | Discussions
jfriend00
Forum ProPosts: 11,116
Like?
Re: I don't think I'm going to convince you...
In reply to Thom Hogan, May 30, 2009

Thom Hogan wrote:

jfriend00 wrote:
We want
them to design to the customer need. If I were in charge of the PM
for a "consumer indoor sports camera" the very first thing I'd want
to know is what the customer failure point is. I'll make you a wager
that it's not noise. Oh, it might be perceived as noise, but
underexposure produces noise, too. This is actually one of the
reasons why I hate Scene exposure modes and go out of my way to
document a better version of every Scene exposure mode in my books.
Nikon's engineers only went halfway on virtually every one. This
doesn't actually lower the failure rate--it ensures that everyone
fails by at least some amount. (Failure being defined as "less than
optimal image.")

Focus, Exposure, Light Color are all potential failure points. An FX
sensor doesn't necessarily fix any of these. You have to be careful
that you're not putting a BandAid on the real problem.

I'd agree with the point that Nikon should understand where the real weak points are for their target audience and make sure they solve those well and if noise isn't one of the top issues, then focus elsewhere. As we've said before, that will include making and marketing the right lens for that use too.

I would guess that AF accuracy on a moving subject in a low light situation is a tough problem for volleyball/basketball moms and dads and we know that it ruins photos even more so than noise. Perhaps future advances in AF integration with the exposure sensor will improve the ability to track a person on-the-go through the frame in a consumer/prosumer camera.

Reply   Reply with quote   Complain
Anthony Medici
Veteran MemberPosts: 5,682Gear list
Like?
Re: I don't think I'm going to convince you...
In reply to CeeDave, May 30, 2009

CeeDave wrote:

A 135mm f2 DC instead of the 200mm f2; about 5 times less money and
much smaller. Or, compared to the 70-200 2.8, get a smaller, cheaper
lens --- with better IQ -- and DX wins a stop back. Not VR, of
course, but AF is fast and you get DC if you want it.

You also give up isolation. The 135 needs to be an F1.4 lens on DX to give the same isolation as the 200 F2.0 does on FX.

A 180mm f2.8 instead of the 300mm f2.8; used, the 180mm is about 6 or
8 times less and (again) smaller and easier to handle. The used 180
is also cheaper than the 300mm f4 (but does not have AF-S), giving a
stop "back" to DX.

Again, a pale imitation. The 200 F2.0 is the equivalent DX lens for the 300 F2.8 on FX. The 180 is slightly short and doesn't give the isolation of the other.

300 instead of 400 or 500, and so on.

The 300 is a long 400 or a short 500. Against the 400, it doesn't have the isolation again since it needs to be F2.0 to match the F2.8 of the 400. (Unless your using a 200-400 then the 300 is just a little long.) And although it can match the isolation of the 500 F4.0, it doesn't have the length needed.

I've found the 300 on DX to either be too long for what I need or too short as compared to using a 400 or 500 on FX.

So DX allows use of a very high quality camera (D300 class)
with truly fast, high-quality "pro" glass --- at a small fraction of the total
cost including lenses.

Unless you're looking to take advantage of the crop by doing things that you can't do on FX. Like getting the angle of view of a 750 mm lens by using a 500.

Because I don't need best-in-class frame rates, I prefer the smaller body, and
find the D300 battery life adequate. Voice annotation would be nice,
though.

I like smaller bodies too. I just don't like the dropped feature set because the body isn't top of the line. I loved the 2x crop in the D2X. It's gone in the D300. I like the mirror black out time of the D3 but I'm stuck with a longer one in the D300. Certainly Nikon could have used the same feature set and choose not to. That's been my complaint all along. It is the cheapening of the features that always bothered me.

 Anthony Medici's gear list:Anthony Medici's gear list
Nikon D800 Nikon 1 V2
Reply   Reply with quote   Complain
Thom Hogan
Forum ProPosts: 13,659
Like?
Re: May end up with fragments of a road map here
In reply to SNRatio, May 30, 2009

SNRatio wrote:

A good 7MP ISO 6400 on the D400 will be more than
enough to make lots of people use that instead of FX alternatives, I
think.

Sorry, but I don't buy that notion. I see little difference between a 7mp and 6mp downsampled image competing against a 12mp image. If this [better high ISO capability through downsampling] were the primary goal of a 14mp sensor versus the existing 12mp sensor, it would be a strange goal in my book.

Sure for most people today, but we must count on standards being
raised a bit.

The emphasis is on "a bit." The problem is that we've hit an interesting intersection. Most who want a DSLR have one. Many who've had DSLRs for a long time are tired of updating. Going from 12 to 14mp is not a significant boost. The worldwide recession means buying is down. And, if that weren't enough, there are signs that even if the recession were to end, consumer spending wouldn't ratchet back up. "A bit" is not going to move metal.

So you imply that the upcoming model is a D300s? I don't think so.

You know, when I wrote my predictions for the year back in November, I almost wrote D300s instead of D400. Something's been nagging me about the Dxxx line for awhile. It all evolves around whether you think the D200 was an anomaly or not. I'm beginning to think it was. Nikon really wants to keep the basic production of the Dxxx model going for four years if it can. What I'm hearing about what's being tested seems to be more D300s than D400. So, do I imply that the upcoming model is a D300s? Probably ; )

Yes, and I think one important question now is how much more raw
processing power may help in the new AF systems.

Well, I have some thoughts on that, since I used to design a camera that used the CPU of the desktop machines ; ). It's tough to brute force focus. You have multiple problems you're dealing with simultaneously, one of which is the physical mass in the lens itself. One reason why phase detection works so well--especially with wave motor lenses--is that a single sample tells you which way you're out of focus and by how much (at least to a relatively small tolerance). Thus, you can instruct the lens to "just go there." When you sub-sample along the way, you have some intersecting problems that mean you have to do very careful calculations on before deciding what to do about moving the lens. Consider that you're on a moving platform, the subject is moving, and you're zooming. The math gets pretty complex, and if those things are all changing speed, too, the math gets more complex as you start trying to smooth the change curves.

Granted, there are some tricks, and some of the compact cameras are now using those tricks. And you can build dedicated ASICs that do the math (once you've figured out what the math is and are sure you're not going to change anything ; ). And things like face detection actually help with the problem (as does Nikon's skin tone detection--you ought to be able to build a focus model from that data stream). So it's not impossible to improve contrast detection systems. But I've yet to see one that can perform at phase detection speeds, which means that until you break that barrier, you'd be going backwards if you deployed one in a prosumer or pro DSLR.

I was reasoning exactly from your premises :-). The question is, how
much of the D700 sensor cost is in the silicon, and how much is in
the topping and wrapping?

If we're only talking about the sensor, it doesn't matter. What matters is yield, and both the actual fab of the die and the post production on the die (e.g. continuous microlenses) have not-so-great yields if you try to mass produce them (i.e. speed up the production and do it in large quantities). I haven't seen anything that indicates to me that the DX and FX sensor prices are changing at all, let alone relative to each other.

price difference could be about halved (to about $600) when they
produce the chips in larger volumes. With a simpler body, we might
therefore end up around $1500, I think.

If you're suggesting that a US$600 sensor can be profitably put in a US$1500 retail price body, you're way, way off. Add up all your costs, now multiply by 3.5. You're in the ballpark of what a consumer electronics firm would want for product margin. Thus, if an FX sensor costs US$400 (which is lower than I think it currently is), US$1400 of the final price of the product is consumed by the sensor. If the camera is to retail for US$2199, that means that ALL THE OTHER PARTS and the labor to assemble it have to cost US$228. Even the 10-year old N80 had US$100 worth of parts and assembly in it, and it didn't have the digital sensor and support electronics (and we'd have to scale that for inflation, too ; ).

-- hide signature --

Thom Hogan
author, Complete Guides to Nikon bodies (19 and counting)
http://www.bythom.com

Reply   Reply with quote   Complain
Thom Hogan
Forum ProPosts: 13,659
Like?
Re: D3 - D300 AF comparison
In reply to maljo@inreach.com, May 30, 2009

maljo@inreach.com wrote:

Your comments that there is little difference between the D300 and D3
is interesting. My impression is that the D3 is noticeably better
in low light, and in accuracy with moving targets.

In general, the D300 and D3 are close in acquiring initial focus on central subjects, enough to call them equal. The D3 does have advantages for tracking focus during continuous bursts. I'm not sure I've shot enough with the D300 in really low light conditions to say whether tracking for a single shot is better or worse than the D3.

Using the same lens, say a 200 f2, with both cameras,
the field of view is quite different. If one frames the image
similarly the subject will be farther from the D300 than the D3,
making the D300's job easier.

Not necessarily. I actually have my most trouble with Nikon's AF system with long distance shooting. Up close, I rarely have issues (assuming I set the camera right, of course ; ).

For a moving target, the closer
it gets to the camera, the harder it is for the camera to keep it
in focus.

You would think so if you were trying to focus like your brain wants you to, but phase detection systems have advantages our brains don't. Close movement actually triggers a more reliable phase difference.

-- hide signature --

Thom Hogan
author, Complete Guides to Nikon bodies (19 and counting)
http://www.bythom.com

Reply   Reply with quote   Complain
jfriend00
Forum ProPosts: 11,116
Like?
Re: D3 - D300 AF comparison
In reply to Thom Hogan, May 30, 2009

Thom Hogan wrote:
Not necessarily. I actually have my most trouble with Nikon's AF
system with long distance shooting. Up close, I rarely have issues
(assuming I set the camera right, of course ; ).

For a moving target, the closer
it gets to the camera, the harder it is for the camera to keep it
in focus.

You would think so if you were trying to focus like your brain wants
you to, but phase detection systems have advantages our brains don't.
Close movement actually triggers a more reliable phase difference.

Both of your comments here match my experience in soccer and track exactly. The shot that my D300 and 200-400 will not focus reliably is the shot moving towards me at distance. My best shots are closer in with the subject filling at least 1/2 to 2/3 the frame. While the subject may be moving relatively faster when closer (and the DOF even shallower), I don't see the camera having difficulty with those shots.
--
John
Popular: http://jfriend.smugmug.com/popular
Portfolio: http://jfriend.smugmug.com/portfolio

Reply   Reply with quote   Complain
René Daniël
Regular MemberPosts: 248
Like?
Re: I'm ready
In reply to Anthony Medici, May 30, 2009
n/t
Reply   Reply with quote   Complain
SNRatio
Regular MemberPosts: 339
Like?
Re: May end up with fragments of a road map here
In reply to Thom Hogan, May 30, 2009

Thom Hogan wrote:

I see little difference between a
7mp and 6mp downsampled image competing against a 12mp image. If this
[better high ISO capability through downsampling] were the primary
goal of a 14mp sensor versus the existing 12mp sensor, it would be a
strange goal in my book.

Sure! You just didn't get my point, which is about 2x downsampling, not about 7 vs 6MP. I'm not at all arguing for 14MP per se, I just take that as a natural, minimal, upgrade. And the difference may be a little more significant when downsampled. But, if there is no D400 this year, I think the 14MP variety may be skipped altogether.

Most who want a DSLR have one. Many who've
had DSLRs for a long time are tired of updating. Going from 12 to
14mp is not a significant boost. The worldwide recession means buying
is down. And, if that weren't enough, there are signs that even if
the recession were to end, consumer spending wouldn't ratchet back
up. "A bit" is not going to move metal.

Not by itself. But I don't think we shall underestimate the effect of buyer confidence here - many may have postponed the decision about upgrading, but when there is an economically attractive DX alternative that gives them virtually everything they want.... I have always felt the sensor is the weakest point in my D300 (not very weak at that, though), and a modest quality boost over the D5000 will make the D300 successor, whether D300s or D400 or whatever, a very complete camera. And as improved high ISO s most important, the pixel density may be safely left where it is.

You know, when I wrote my predictions for the year back in November,
I almost wrote D300s instead of D400. Something's been nagging me
about the Dxxx line for awhile. It all evolves around whether you
think the D200 was an anomaly or not. I'm beginning to think it was.
Nikon really wants to keep the basic production of the Dxxx model
going for four years if it can. What I'm hearing about what's being
tested seems to be more D300s than D400. So, do I imply that the
upcoming model is a D300s? Probably ; )

I get your point, and I agree. I have had the rather fixed idea that the 's' upgrades must be mostly cosmetical, and a bigger upgrade would, at least, require a new sensor. But the necessary items to upgrade for the D300 doesn't really qualify for a new generation, and the whole D300 concept is very viable. So I think 4 year cycles may be the most natural for this line. Which makes your initial question all the more interesting: Should Nikon do a "D4DX", a "D400", or both, next time? Maybe the smartest would be to do the D400 first, with a dual D4/D400 release, and somewhat later the pro DX body?

Yes, and I think one important question now is how much more raw
processing power may help in the new AF systems.

Well, I have some thoughts on that, since I used to design a camera
that used the CPU of the desktop machines ; ). It's tough to brute
force focus. You have multiple problems you're dealing with
simultaneously, one of which is the physical mass in the lens itself.
One reason why phase detection works so well--especially with wave
motor lenses--is that a single sample tells you which way you're out
of focus and by how much (at least to a relatively small tolerance).
Thus, you can instruct the lens to "just go there." When you
sub-sample along the way, you have some intersecting problems that
mean you have to do very careful calculations on before deciding what
to do about moving the lens. Consider that you're on a moving
platform, the subject is moving, and you're zooming. The math gets
pretty complex, and if those things are all changing speed, too, the
math gets more complex as you start trying to smooth the change
curves.

Well, when we throw more raw processing power at a problem, it is normally not just brute force.. Rather, it is often used for performing a much more thorough analysis. For instance, may the weaknesses of phase detection be partially overcome by an approach involving more computational statistics? I haven't looked into this, I just have a feeling that, for example, Bayesian methods might be of some use - if they are not employed already. Different kinds of information must be utilized, and non-deterministic approaches probably have a place in tracking.

price difference could be about halved (to about $600) when they
produce the chips in larger volumes. With a simpler body, we might
therefore end up around $1500, I think.

If you're suggesting that a US$600 sensor can be profitably put in a
US$1500 retail price body, you're way, way off. Add up all your
costs, now multiply by 3.5. You're in the ballpark of what a consumer
electronics firm would want for product margin. Thus, if an FX sensor
costs US$400 (which is lower than I think it currently is), US$1400
of the final price of the product is consumed by the sensor. If the
camera is to retail for US$2199, that means that ALL THE OTHER PARTS
and the labor to assemble it have to cost US$228. Even the 10-year
old N80 had US$100 worth of parts and assembly in it, and it didn't
have the digital sensor and support electronics (and we'd have to
scale that for inflation, too ; ).

Now you're playing the calculating tricks here We were talking in the sales prices scale. In terms of sensor prices, we could use US$500 and 300. And the sensor may well be 20% percent of retail price. If we use your factor backwards, we would come to about $290 for the D90. Subtract $80 for the sensor, add $300 for the FX variety, and multiply again. We end up between 1500 and $2000.

As a check, the price difference of the sensor assembly between D300 and D700 should be $250-350 by these principles - not too far off, I guess.

But of course, if there is no way to reduce the sensor production cost substantially, this is purely hypothetical. I would find that strange, though.

Reply   Reply with quote   Complain
Thom Hogan
Forum ProPosts: 13,659
Like?
Re: May end up with fragments of a road map here
In reply to SNRatio, May 30, 2009

SNRatio wrote:

Which makes your initial question
all the more interesting: Should Nikon do a "D4DX", a "D400", or
both, next time?

That was my whole reason for continuing the discussion. I doubt Nikon is locked into 2011 decisions yet, though they are well on their way ; ).

Maybe the smartest would be to do the D400 first,
with a dual D4/D400 release, and somewhat later the pro DX body?

I say if you're going to do all three, do them simultaneously. By 2011, many of us are just going to be out shooting ; ). Nikon needs a Big Bang, much like the D3/D300 was, to get our attention back.

Well, when we throw more raw processing power at a problem, it is
normally not just brute force..

If you have to use more processing power, by definition you're using brute force. What you look for in designs is elegant simplicity wherever possible. Nikon doesn't have that for contrast AF, in any camera.

Rather, it is often used for
performing a much more thorough analysis.

That's throwing more CPU (brute force) at more data in my book. Sometimes that's the right way to do it. Something tells me that this is not the right way for AF.

I just have a feeling that, for example, Bayesian methods might be of
some use - if they are not employed already.

They are.

Different kinds of
information must be utilized, and non-deterministic approaches
probably have a place in tracking.

That's where simplicity enters the picture. If I knew what the subject was and where it was, I could probably calculate a reasonable exposure by looking at one pixel in the metering sensor ; ). That was the greatness of phase detection: one value gave you lots of information.

Now you're playing the calculating tricks here We were talking
in the sales prices scale.

And the way a manufacturer determines sales price is to look at costs and then figure out what profit margin they want to obtain. That's why Parts*3.5 is a very good rough estimate. The "Parts" portion is real and easily calculated, the "3.5" part is a rough approximation of all the marketing, sales, and distribution costs + profit.

In terms of sensor prices, we could use
US$500 and 300.

You might. I see nothing that gets us to an FX sensor price of US$300 in the foreseeable future. Nothing. As I noted, my US$400 was even optimistic.

If we use your factor backwards, we would come to about $290
for the D90.

Wanna guess what my parts analysis says ; ).

Subtract $80 for the sensor,

Too high.

add $300 for the FX

Too low.

We end up between 1500 and $2000.

Back in the late seventies when we started selling lots of Apple II's with spreadsheets to middle managers in big companies, we came up with a term for what the delusions they entered into: Visicalc Mentality. After all, if the numbers were correctly calculated by the computer, the result must be right ; ). Only problem is: if you put garbage in, you get garbage out the other end.

As a check, the price difference of the sensor assembly between D300
and D700 should be $250-350 by these principles - not too far off, I
guess.

As I've tried to say several times but apparently wasn't heard, it appears to me that Nikon is taking less margin on a D700 than they normally would want to. There are many reasons why a company chooses to do that, but with Canon also taking less margin on the 5D (now 5DII) and having a competitor that would be unmatched by Nikon if Nikon used their usual pricing, one can see at least one reason why they might have priced the D700 as they did. Most of Nikon's products are priced with consistent (and relatively high) product margins. The D700 isn't.

-- hide signature --

Thom Hogan
author, Complete Guides to Nikon bodies (19 and counting)
http://www.bythom.com

Reply   Reply with quote   Complain
TOF guy
TOF guy MOD
Forum ProPosts: 10,768Gear list
Like?
Re: another D400 discussion cont'd
In reply to jfriend00, May 31, 2009

John wrote:

Read this thread and see the sample shots:
http://forums.dpreview.com/forums/read.asp?forum=1030&message=31789476
and see if you change your mind.

I've used VR quite often on my longer lens and have not yet observed this even once ! I wouldn't give up on VR because of a malfunction which is either rare or possibly a lens malfunction.

Believe it or not, VR in some
lenses messes up the quality of the background. The theory is that
VR works by decentering a piece of optics to try to stabilize the
in-focus portion of the image. But using the outer edges of this
optic degrades the out of focus areas of the image (the bokeh).

No, something totally different is at play here.

There is also a theory that VR has a limited range over which it can
correct and, at some point, it has to "snap" back to the center of
it's range in order to continue doing corrections.

That theory is a misunderstanding of how it works: once the correction has reached the maxmum range it resets itself * during composition * (while the shutter is half-pressed). It also systematically reset itself right after fully pressing the shutter, while the mirror is moved and * before * opening the shutter to maximize the VR efficiency during the pic (all according to Nikon's description of VR). What Nikon does not tell you is that this comes at a price: the pic is not necessarily taken as it was composed because resetting VR changes it. This is the price to pay for stablizing the image during composition, and why I am not so enamored of stabilization during composition.

VR does not reset itself * while the picture is recorded * if it reaches its maximum range: simply the the image is not stablized anymore at that point.

In continuous shooting mode, VR resets itself between each picture while the mirror is moved (answer to your comment in another post about 8 fps, which I have to do here as we've reached 150 posts).

If the shutter
fires while it's moving during the "snap-back" before it stabilizes
again, it could impact your shot.

It does not reset during the shot, only during composition / shutter half-pressed.

I am aware of no coordination
between shutter and VR to prevent this.

Now you know it is a non issue

It might be rare given the
timing,

More like never !

-- hide signature --

Thierry

 TOF guy's gear list:TOF guy's gear list
Nikon D800E Nikon AF-S Nikkor 70-200mm f/2.8G ED VR Nikon AF-S DX Nikkor 12-24mm f/4G ED-IF Nikon AF-S Nikkor 24-70mm f/2.8G ED Sigma 150mm F2.8 EX DG OS Macro HSM +3 more
Reply   Reply with quote   Complain
TOF guy
TOF guy MOD
Forum ProPosts: 10,768Gear list
Like?
Re: another D400 discussion cont'd
In reply to Billx08, May 31, 2009

Billx08 wrote:

John wrote:

. . .
There is also a theory that VR has a limited range over which it can
correct and, at some point, it has to "snap" back to the center of
it's range in order to continue doing corrections. If the shutter
fires while it's moving during the "snap-back" before it stabilizes
again, it could impact your shot.

Bill replies:

It's not a theory. I recently read some page on Nikon's website that
explained that one reason why Nikon's VR was better than other
manufacturer's IS and OS lenses was because it could snap back to the
center position just before the shutter opened,

That's right: before the shutter opens, never after.

providing the maximum
amount of "correction" during the time the shutter is open.

Resetting the correctivel element position does maximize VR efficiency, but at the price of changing the composition (which Nikon does not mention), i.e. there is a price to this.

The
camera and lens VR are evidently in synch,

Of course resetting VR never occurs while the pic is taken. Only 1) during composition (menaing while half-pressing the shutter) and 2) while the mirror is moved, right before the shutter is opened.
--
Thierry

 TOF guy's gear list:TOF guy's gear list
Nikon D800E Nikon AF-S Nikkor 70-200mm f/2.8G ED VR Nikon AF-S DX Nikkor 12-24mm f/4G ED-IF Nikon AF-S Nikkor 24-70mm f/2.8G ED Sigma 150mm F2.8 EX DG OS Macro HSM +3 more
Reply   Reply with quote   Complain
SNRatio
Regular MemberPosts: 339
Like?
Re: May end up with fragments of a road map here
In reply to Thom Hogan, May 31, 2009

Thom Hogan wrote:

SNRatio wrote:

Which makes your initial question
all the more interesting: Should Nikon do a "D4DX", a "D400", or
both, next time?

That was my whole reason for continuing the discussion. I doubt Nikon
is locked into 2011 decisions yet, though they are well on their way
; ).

I see. To sum up: D300 is done now, D400 will probably have both a new sensor and a new AF system. We want these parts also in a bigger brother, which continues the D2X line. With faster electronics for handling the DX sensor readout, a high-speed crop mode and otherwise whatever non-FX-specific that goes into the D4. Did I miss out anything important?

Maybe the smartest would be to do the D400 first,
with a dual D4/D400 release, and somewhat later the pro DX body?

I say if you're going to do all three, do them simultaneously. By
2011, many of us are just going to be out shooting ; ). Nikon needs a
Big Bang, much like the D3/D300 was, to get our attention back.

They will surely get attention from us existing users with a new dual release only. But competition-wise, a triple release might be a very good move. Maybe even an urgent one, to match a new 1.3X crop offering from Canon.

Well, when we throw more raw processing power at a problem, it is
normally not just brute force..

If you have to use more processing power, by definition you're using
brute force. What you look for in designs is elegant simplicity
wherever possible. Nikon doesn't have that for contrast AF, in any
camera.

You are using a non-standard definition of "brute force" here. It usually means "low-algorithmic" approaches, in that all available information is not fully utilized. I would, for example, not call syntactic pattern recognition "brute force". But it can be very computationally intensive.

I just have a feeling that, for example, Bayesian methods might be of
some use - if they are not employed already.

They are.

And in this context, they will often use lots of processing power...

In terms of sensor prices, we could use
US$500 and 300.

You might. I see nothing that gets us to an FX sensor price of US$300
in the foreseeable future. Nothing. As I noted, my US$400 was even
optimistic.

I still wonder about the different parts costs here: chip, filters and off-chip circuitry.

As a check, the price difference of the sensor assembly between D300
and D700 should be $250-350 by these principles - not too far off, I
guess.

As I've tried to say several times but apparently wasn't heard, it
appears to me that Nikon is taking less margin on a D700 than they
normally would want to. There are many reasons why a company chooses
to do that, but with Canon also taking less margin on the 5D (now
5DII) and having a competitor that would be unmatched by Nikon if
Nikon used their usual pricing, one can see at least one reason why
they might have priced the D700 as they did. Most of Nikon's products
are priced with consistent (and relatively high) product margins. The
D700 isn't.

Using Norwegian net prices as of today, and a factor of 3.5 for the D300 and 3.0 for D700, I end up with a difference of about $330. I doubt the real difference is very, very much higher, because the D300 is also loaded with expensive components.

Alternatively, which I guess is more like Nikon's own take on this, the markup on the rest of the body is normal, while it is about halved on the sensor assembly. And if they do the same for the "D90FX", with a sensor assembly cost of $400, guess where we end up

Therefore, in addition to the business-as-usual D4/D400, I think we might very well see an enthusiast "D90FX" in the $1500-$2000 range launched at around the same time. With a "D4DX" as well, Nikon would then have two complete lines, and the need for a road map might be less urgent.

Reply   Reply with quote   Complain
Keyboard shortcuts:
FForum MMy threads