D300s Successor-D400, what and when

Paperman

Commercially available silicon wafers currently max out at 200mm and 300mm. The 450mm wafers won't be available until 2018. My math gets 135 DX vs. 45 FX on a 300mm wafer. The cost differential ratio will be roughly the same, roughly 3:1.

I have seen the highest numbers of around 75 DX vs 24 FX and even as low as 18Fx sensors per wafer. I'm not sure how old those numbers were as manufacturing processes have improved year over year. Or as PB_PM said with the error rate, maybe that is all the usable ones they get out of it.

\-------------

Just googled CMOS wafer prices . I had to, as I simply could not believe a piece of electronics size of a stamp could cost as much as an iPad,

300mm Wafer $ 3200
200mm Wafer $ 820

Ironheart's figures bring a FF sensor cost to +/- $70, TTJ's to +/- $130 . That brings the FX-DX sensor cost difference to $50-$90 .

Any comments ?? Remember I was talking about pure material cost; let's not go beyond that.

PB_PM

Except there is more than a sensor that seperates DX and FX. You also need a larger mirror, shutter, and pentaprism.

Paperman

Add another $50 for that if you want ...( A DX DSLR with all these included + lens goes for $400-500.)

I can't imagine the material cost of any electronic/high tech consumer item being more than 10-20% of the retail price. Most of the cost is R&D, labor costs, marketing ( + logistics, servicing, etc... ).

You simply can't have a $500 sensor in a $1700 or a $2500 DSLR .

And you can't compare DSLRs to phones which are subsidized by the phone companies. Nikon has to sell every unit with a profit as it is not an item that carries a contract that will bring additional revenue.

ChromiumPrime

Commercially available silicon wafers currently max out at 200mm and 300mm. The 450mm wafers won't be available until 2018. My math gets 135 DX vs. 45 FX on a 300mm wafer. The cost differential ratio will be roughly the same, roughly 3:1.

I have seen the highest numbers of around 75 DX vs 24 FX and even as low as 18Fx sensors per wafer. I'm not sure how old those numbers were as manufacturing processes have improved year over year. Or as PB_PM said with the error rate, maybe that is all the usable ones they get out of it.

\-------------

Just googled CMOS wafer prices . I had to, as I simply could not believe a piece of electronics size of a stamp could cost as much as an iPad,

300mm Wafer $ 3200
200mm Wafer $ 820

Ironheart's figures bring a FF sensor cost to +/- $70, TTJ's to +/- $130 . That brings the FX-DX sensor cost difference to $50-$90 .

Any comments ?? Remember I was talking about pure material cost; let's not go beyond that.

Hahaha what a nostalgic time ) ... new forum, new D400 thread but still same old sensor cost argument. Maybe somebody should change the topic to "D400 (FX vs. DX sensor costs)" )

Well, I'm all talked out of this topic this time around but for a reference, people should read from this post on the old froum down to about the bottom of the next page.

Msmoto

@ ChromiumPrime
Things do seem to get off topic....I wonder if anyone is planning on purchasing a D400 or any new Nikon body anytime soon. Nikon may have done such a PR job in dealing with the D800 and D600 problems that no one is going to by a D...anything. I certainly will have to wait on this until I see what the new product is like. And, in spite of a Nikon loyalty of nearly 100%, my next lens will be a Sigma no doubt. The fact Nikon has not let anything slip out about a D300s replacement is very curious. Even the leak six months ago was apparently grossly inaccurate.

Let's see....native ISO of 12,800, 9 FPS, focus like D800 (corrected for all the points) and full metal body. Oh, 18 mp sensor. USD $1995.00.

Ironheart

Not to belabor the point, but I should have said "at least" 3:1. Wikipedia puts the ratio at 20:1 DX:FX which seems a bit high to me, but I could easily believe 5 or 10:1. As far as the cost of the 300mm wafer, I assume that is before exposure and etching, basically a raw slice of silicon. Modern sensors probably require multiple passes for the various layers, and as others have mentioned there will be yield issues due to defects which will impact the viability of FX greater than DX due to the larger footprint.

All in all, I would peg the cost of a fully packaged and tested FX sensor somewhere between $500-$1000. DX is probably in the $100 to $150 range.

donaldejose

Sorry Ironheart but I have a hard time believing Nikon can pay $500 to $1000 dollars for an FX D600 sensor and all the other parts of the camera only cost $1,500 to $2000. Maybe so, but it seems highly unlikely to me.

Here is another twist on the length of, and interest, in this thread. Many people have argued the old market for the D300s no longer exists because that market niche has been split by the new D600 and the yet to be released D7200. In other words, when those two bodies are available very few people will buy a D400. Maybe, but the length of this thread suggests a lot of people are quite interested in a D400.

ChromiumPrime

@Msmoto I believe something is going to change (and quick) with Nikon within the next few months. The sales report rankings are finally starting to come in and things aren't looking favorable for Nikon. Up until this year, Nikon had been steadily gaining ground on Canon in the DSLR market. This year, however, a year that was supposed to be so good for Nikon, not only did Nikon lose a bit of market share and Canon gain more market share but the gap widened from 7% the year before to 17% this year.

While sales of D4/1Dx, D800/5DIII and D600/6D were probably a small part of those shares, the companies handling to various issues within those lines most likely contributed much more to the bottom line than other factors. The difference between the way Canon addressed their "light leak" issue (an issue that affects a small percentage of overall users) and the way Nikon is handling their left-side AF and dust issues is stark and I'm more than positive that this had a lot of influence in overall sales (consumer confidence, word of mouth/brand recommendations... etc).

It looks like Canon have already learned from their 1DIII AF fiasco. Nikon doesn't look like the type to learn from other's mistakes and so now they have their own lesson. If they want to stay competitive, they have no choice but make some changes - and quick.

Ironheart

A quick Internet search on repair costs for replacing an FX sensor shows a range of $1000 to $2200. Obviously this includes labor, etc, but I bet the part is close to the range I suggested. A Intel Core i7-3960K goes for $1000 just for the CPU, and it is 435mm^2 compared to FX which is 864mm^2

I guess I'm just crazy...

donaldejose

Ironheart: You are not crazy, but how can Nikon pay $1000 for a D600 sensor and sell the whole camera for only $2,100. That is what I cannot understand. Doesn't seem to make sense to me. Surely, all those other parts must cost more than half the final price of the camera, especially since Nikon will not be getting the full $2,100 for the camera in the first place. The camera store will get some money, the importer/distributor will get some money, the shipper will get some money. Seems to not be enough money to go around if that sensor costs Nikon $500 to $1,000. When the D7200 comes out with the same build as the D600 and the same features but just with a DX senor perhaps we will have a yardstick by which to measure the DX to FX differential. Perhaps the only real difference will be the DX vs FX sensor and then the cost difference can be directly attributed to the additional cost of an FX sensor. It may be the same with the D400 and the D800. Perhaps the build quality and feature set will all be the same with the only difference being the sensor used. Then we would have a second yardstick by which to measure the DX vs. FX cost differential (assuming all other parts are essentially the same

PB_PM

Easy, cost cutting. The use of lower end parts throughout the body of the D600. If the extra shutter, mirror and pentaprism are only $50 extra over DX (as you claim) then it's easy to see. In fact the margins on the D600 could be $200-400 a unit. $1000 sensor, $600 body (all other parts). After all the D600 is just a slightly modified D7000 with an FX sensor, yet the D7000 body only started at $1099. Cut the margins out of the D7000 (release price), and that almost perfectly accounts for the $500 more for the FX sensor equipped D600.

donaldejose

Maybe you are correct. It would be nice if we had equivalent cameras with the only difference being the DX vs. FX sensor to compare. I think the right comparison to a D600 will be the D7200 when it arrives rather than the D7000. Both D600 and D7200 will have the same Expeed 3 processor, for example. The D7000 has the Expeed 2 processor.

Paperman

"All in all, I would peg the cost of a fully packaged and tested FX sensor somewhere between $500-$1000. DX is probably in the $100 to $150 range"

So a single 24x36 mm sensor ( sold in millions ) now more expensive than iPad ? Let's get real guys !

I wish we had someone among us working in the electronics industry that could tell us how little the "parts" cost of an electronic item is when compared to R&D, labor, marketing, distribution costs of that item.

Hundreds of highly paid execs, scientists and engineers, thousands of management /admin/ marketing /production staff. Shipment from one end of the world to maybe 50 countries, warehousing and distribution there , dealer commissions , claims/returns/servicing, advertising ( same in any business ) ... Yet, just a single body part costing 1/2-1/3rd the price of a DSLR ?? . I will be surprised if the WHOLE cost of a DSLR is more than 1/3rd of retail price ...

PB_PM

The sensor itself may not be $1000, but since Nikon outsources actual sensor production, the unit cost is a little higher. Canon on the other hand, develops and produces their own sensors.

Ironheart

I did say "fully packaged". I used the cost of a CPU that is made in 10 or 100 times the volume as a comparison point. I also happened to work in the electronics industry and have actual knowledge of the costs involved. Chips are expensive to make, and that cost goes up exponentially with the size.

Thom hogan on Feb 15 last year said "...that it was $5, $50, $500 (compact, DX, FX)" http://www.bythom.com/2012 Nikon News.htm

Thom was talking about the cost of a unpackaged sensor chip. Adding in other costs could triple the final component cost.

For additional comparison here are some Intel chip sizes with costs:
294mm^2 (M1 stepping) $294
435mm^2 (C1 stepping) $999
A DX sensor is 370mm^2 compared to an FX sensor which is 864mm^2. I think Thom and I are very close to the true cost of goods for the sensors.

Finally I think you are off by quite a bit on the profit margin, perhaps the other way around. Costs are closer to 2/3 with profit 1/3. As a data point the 163mm^2 A5X costs $23 (iPad 3 processor). That does not include Apple's R&D cost as they designed the chip in house. That is what Apple pays Samsung for the chip. The total parts cost for a 32GB iPad 3 is $365 which gives apple about a 50% profit margin, considered a premium in the electronics industry. Oh, and I'm not sure Nikon has made a million of any one of their FX sensors yet.

http://appleinsider.com/articles/12/03/16/high_cost_of_new_ipad_components_drive_down_profit_margins

Anyway, this is just my viewpoint, take it or leave it. Some day we may know the real costs on these things and I will either owe you all a beer or you me :-)

donaldejose

I think it is an interesting discussion.

But a slight correction, Thom acutally said "It used to be that it was $5, $50, $500 (compact, DX, FX)." Note, he said "used to be" not "is." On December 5th Thom said "Even if an FX sensor is only US$250 fully burdened in large quantities" which suggests the price of an FX sensor today is higher than $250 but lower than $500. Thom did say FX was 10x more costly than DX. If FX is $500 and DX is $50 there would be a cost differential of $450 if all other parts of the camera body cost the same. But we are seeing a price differential of more like $900 for the "FX premium."

Paperman

@Ironheart

Finally I think you are off by quite a bit on the profit margin, perhaps the other way around. Costs are closer to 2/3 with profit 1/3.

You have got me wrong. When I say the cost of a DSLR being maximum 1/3rd of retail price, I mean the parts/material cost - I think that was clear enough when I mentioned all additional costs. The R&D/ labor/ marketing/ logistics etc not included and will be at least another 50%. You, on the other hand, are the ones saying a FF sensor may cost $500-$1000 .

When you add all the costs , Nikon as any company should be happy to end up with 5-10% profit from gross sales.

Anyway, we are going on and on with issues like this simply because Nikon hasn't given us a D400 yet :-) and all we can do is speculate and speculate and talk nonsense .... Must be the longest thread ( inc. the old one ) .

DaveOl

I've had more trouble registering for this forum.
I'm glad that someone restarted this discussion on the D400. I was the person who started in on the old forum. I never expected the popularity of that subject.
I'm still waiting on an announcement by Nikon for this camera. I am tempted by the D 700, but I like the crop factor of the DX cameras.

TaoTeJared

Commercially available silicon wafers currently max out at 200mm and 300mm. The 450mm wafers won't be available until 2018. My math gets 135 DX vs. 45 FX on a 300mm wafer. The cost differential ratio will be roughly the same, roughly 3:1.

I have seen the highest numbers of around 75 DX vs 24 FX and even as low as 18Fx sensors per wafer. I'm not sure how old those numbers were as manufacturing processes have improved year over year. Or as PB_PM said with the error rate, maybe that is all the usable ones they get out of it.

\-------------

Just googled CMOS wafer prices . I had to, as I simply could not believe a piece of electronics size of a stamp could cost as much as an iPad,

300mm Wafer $ 3200
200mm Wafer $ 820

Ironheart's figures bring a FF sensor cost to +/- $70, TTJ's to +/- $130 . That brings the FX-DX sensor cost difference to $50-$90 .

Any comments ?? Remember I was talking about pure material cost; let's not go beyond that.

You can not talk "about pure material cost" - sorry but that is absurd. That is just like telling a photographer their not worth selling their services since photo paper cost $0.10 a sheet. If the sensor is purchased from Sony - as everything points to it, then you have to include all costs as Sony is not going to sell it at cost. A blank wafer sheet is just that, a blank wafer sheet.

I found your article you are referring to: you should have read further down:
In a similar manner to wafers, survey participants reported that after decreasing for two consecutive quarters there was a sequential increase in the median mask set cost for 200-mm CMOS wafers. Participants indicated that the median cost, at $79,200 increased 21 percent QoQ and 20 percent YoY. Mask sets for 300-mm wafers had a median cost of $510,000, a decrease of 1.8 percent QoQ, but up 2.0 percent YoY.


If you use Ironheart's numbers, 45FX sensors, with the numbers above - that is $600 per sensor. That's with zero losses.

Paperman

How do you get $600 per sensor ? ... x 45 makes $27,000 - not a figure I can see in the article... Surely you are not suggesting the price of 300mm wafer rose to $27,000 or to $510,000 ? ( from $3200 in the same year ) . I don't have a clue what those figures in the article are by the way.

TaoTeJared

*palm to face -

So where the heck is the D400!?!

Msmoto

Yup, we are after all looking for a D300s replacement....but maybe Nikon is being run by young children who have no idea of the true heritage of the Nikon brand....mmmmm? ( ( (

sevencrossing

Yup, we are after all looking for a D300s replacement....but maybe Nikon is being run by young children who have no idea of the true heritage of the Nikon brand....mmmm? ( ( (

I think it might make more sense if the children are looking at something to compete with Samsung's rumored NX-R mirror less full frame camera
mirrorless and full frame is the future for the Professional
mirrorless and 13.2x8.8mm ( nikon 1) for the rest

donaldejose

Lets hope the delay is caused by some new technology being added to the D400 to make it very attractive, even compared to FX such as the D600. What advantage could a DX have over an FX for about the same price? Maybe better IQ at high ISO such as the D4 has at and above ISO 12,800? Maybe a high ISO IQ like the D4 in a DX sized (and priced) sensor? Dare I say it? A "D4 lite?"

PB_PM

I don't think it is a delay at all. If Nikon follows the usual release pattern for this class of camera the D400 would be announced in late August and start shipping in September.