[deejjjaaaa]

This is difficult to say as most sources are closed. One source I found is this:

from Cypress, a 9MP APS-C sized CMOS sensor which they project will sell for around $90 and which should be available in quantity in early 2006

So, APS-C was 90$ two years ago.

Another source from last year is this:

Sony [...] US $493.3 million to expand its Kumamoto Technology Center (TEC) in Kyushu over the next three years. [...] Production capacity will be ramped by about 20% at the site, run by subsidiary Sony Semiconductor Kyushu Corp., to the equivalent of >20,000 wafers/month for both CCD and CMOS sensors

Assuming they earn 250 million per year (750 for 3 years) just on the 20% increase, a wafer would sell for about 5000 $. A 300mm wafer holds about 150 APS-C chips assuming a yield of 80%. I don't buy the bad yield argument because CMOS sensor chips, unlike CPU chips, are allowed to have faulty sensor cells. Those numbers therefore end up to 35$ per APS-C sensor.

Taking both sources and a certain price curce into account, I would say that an APS-C chip today has manufacturing costs of about 50$.

From this, I would say that an FF chip can be produced for 150$ (**).

Disclaimer: All this is deduced from public sources. I don't know what camera makers really pay. And in the case of Sony they may not even know themselves

(**) The (cost of FF) = (2/y)*(cost of APS-C) , where y is the yield for APS-C production.

does not look like FF sensors are actually cost that much if you in fact own the fab (Samsung, Sony, Canon) ... and the numbers shown are in single digit percentage of the camera cost - am I wrong ?

[kmccanta]

i used to think FF was overrated

but as i mentioned in another thread, after using a Pentax Super Program, i changed my perception.
...
but again, as i mentioned in another thread, its near impossible to explain to a person what 100% FF viewfinder LOOKS AND FEELS like, you have to go out and try it for yourself to really understand.

from the definitive source: Pentax Camera Bodies

The superProgram has a viewfinder with coverage x [magnification] of

0.82 x [92%]= 0.75

The MX has

0.97 x [95%] = 0.92

the K100 has this, but the sensor is smaller so we have to divide by 1.5 (or x.667)

0.85 x [96%] = 0.816 => 0.816 x 0.667 = 0.54

K10 = 0.95 x [95%] = 0.9025 => 0.9025 x 0.667 = 0.60


So it seems the K100 has better vf image than SuperProgram?? Unless the APS-C vs. FF factor is used... then the SuperProgram compares favorably (on paper).

I agree about seeing in person. I LOVE the view through my SuperProgram, but I could hardly tell the difference between K10 v.s. K100. It was different, sure, but not dramatic.

[falconeye]

and the numbers shown are in single digit percentage of the camera cost - am I wrong ?

Probably yes. I guess a K20D is about 300$ to build. 17% of it would be the sensor then.

For FF, if we assume 150$ sensor and 350$ for the rest (up from 250$), we get 30% of a total of 500$.

So, street prices could be:
900$ for K20D and 1500$ for K2D.
But because FF market is high end, it would more likely run at:
900$ for K20D and 2000$ for K2D.
The extra 500$ is extra profit for Pentax and still cheaper than the competition. List prices would be ~1.5x.

[deejjjaaaa]

here is the link where you can see the difference w/ your own eyes

Standard versus Enhanced Viewfinders - Size Matters :: Wetpixel.com

scroll down to the table

[beaumont]

One of the biggest benefits of FF in my opinion is that hopefully there will be an increase in the number of used DA lenses being offered at firesale prices from those making the leap

[cpopham]

I don't buy the bad yield argument because CMOS sensor chips, unlike CPU chips, are allowed to have faulty sensor cells.

I'm not convinced you're correct here. Things like Graphics Processors have whole pipelines disabled on chips which didn't come out perfectly. Desktop Processors frequently have cache blocks disabled, and sometimes even whole cores. However, this still does not result in a 100% yield, because there are still plenty of locations where critical faults occur, or other failures can ruin the whole wafer. With larger chips, you need to produce more wafers to produce the same number of chips, thus a lower yield rate.

Surely, even if applying your criteria of disabling defective photosites, there must be a criteria in place for assessing a chip which is good enough compared to one which is defective. Consider for a moment an LCD panel - a few defective pixels scattered on the screen are "acceptable", but a cluster of defective pixels are not.

With silicon chip manufacturing, in general, as the physical size of the circuit features decreases, the rate of defects also decreases. This means that in general, for the same resolution, an APS-C sensor should have a lower number of defects than a Full Frame sensor. If one was applying the exact same criteria for judging an acceptable chip, compared to an unacceptable one, then you'd have a higher percentage of rejected Full Frame sensors.

If you change your evaluation criteria you're now producing what is effectively a lower quality sensor and you can't realistically compare the yields anymore.

Also, the equation you've used for calculating the manufacturing cost of a Full Frame sensor has a minor flaw in it - a silicon wafer is circular in shape, thus a smaller chip such as an APS-C can actually fit more than twice as many individual chips onto a wafer compared to a chip twice it's size. What I'm not sure of, however, is what the actual ratio would be.

Don't get me wrong, I'm not trying to say that full frame sensors are insanely expensive - what I'm saying is that they *are* most definitely more expensive than APS-C sensors, and that I believe it is a mistake to trivialise the cost of the sensor.

does not look like FF sensors are actually cost that much if you in fact own the fab (Samsung, Sony, Canon) ... and the numbers shown are in single digit percentage of the camera cost - am I wrong ?

A given fab is able to produce less Full Frame sensors than it would APS-C (or compact camera sensors for that matter), simply because a fab is capable of handling silicon wafers at a certain rate, and there are less FF sensors in a wafer.

A fab costs a lot of money to set up - and there is always demand for fab capacity from third parties. It could be that this Sony subsidiary could actually make more money producing other chips for other companies (in fact, they actually do produce things for other companies) presumably that means Sony Imaging has to pay something close to a fair market price....

[falconeye]

cpopham, I think we don't disagree. I assumed an FF sensor would be three times the price for twice the surface. Pretty fair, I guess.

BTW, I changed my body building cost estimate above to 300$ as I confused $ for €. I also added an FF body building cost estimate.

With larger chips, you need to produce more wafers to produce the same number of chips, thus a lower yield rate.

Maybe you don't mean it the way you said it. However, the way you said it this statement is wrong. The yield rate y is not the number of chips per wafer. This is the yield. The yield rate is the number of good chips per total number of chips produced. Typical yield rates range from 50% to 90%.

If you double the surface of a chip, the yield rate y' for chips twice as large is as follows:

• y' = y^2 for chips w/o redundant structures, i.e., a single defect ruins the chip.
• y' = y for chips with full redundant structures, i.e., where only failure density must stay below a threshold.

CPU chips have a higher surface fraction belonging to the first category than CMOS sensor chips. That was my point. Nevertheless, I used the second formula with y=0.8.

Consider for a moment an LCD panel - a few defective pixels scattered on the screen are "acceptable", but a cluster of defective pixels are not.

Yes. For cluster defects one must apply the second formula. But y will be close to 1 for cluster defects, so y^2 will be too.

as the physical size of the circuit features decreases, the rate of defects also decreases

I assume APS-C and FF to have the same pixel density. IMHO, it otherwise makes no sense to produce FF sensors. Of course, the camera may store with less pixels. It has been shown many times that downsampled images from high pitch sensors outperform original low pitch sensor images (for same surface sensors). This is because of lower read-out noise and better bayer interpolation. Therefore, all my arguments assumed a constant pitch (pixel density). An FF sensor will have more pixels and more defects.

Also, the equation you've used for calculating the manufacturing cost of a Full Frame sensor has a minor flaw in it - a silicon wafer is circular in shape

Very minor, indeed. The exact mathematical formula is intricate, you basically have to draw the layout. The area too close to the border is called "waste". Typical waste ratios are 12-20%. With 36mm chips on 300mm wafers, it will be larger. The increase though, will only be about 5% absolute, like in 15 to 20%. I allowed to ignore this effect.

they *are* most definitely more expensive than APS-C sensors, and that I believe it is a mistake to trivialise the cost of the sensor. [...] It could be that this Sony subsidiary could actually make more money producing other chips

I didn't say otherwise. I guessed an FF sensor is 3x the price or ~150$. A Fab will obtain higher margins on FF chips than APS-C chips because price elasticity is smaller.

BTW, which is why DSLR is more profitable than P&S. Or why Pentax needs to withdraw from P&S and need an FF body: they would generate much higher margins from those (why Airbus had to make A380 when Boing made all its profit from 747).

[cpopham]

Insightful and well written response

My apologies, it seems you know more about these things than me, and had in fact thought it all the way through.

Thanks for taking the time to post an explanatory follow up - it's always embarrassing to be proven wrong, but nice when someone posts a polite and detailed explanation.

I have to say this is probably the only FF speculation thread on these forums with a positive rather than negative tone...

[Michael Barker]

Viewfinder brightness is greatly affected by the open aperture.
I remember hating the view through a Sigma 28-200/3.8-6.3 on my K1000 when I compared it to my K55/1.8. On digital it was so bad - (small and dim!) that I had to get rid of it.
I love the light weight of the DA 50-200 - this is a great advantage of APS-C. Pentax has done such a good job on APS-C that it certainly makes up for the best-in-class but small & dim viewfinders on digital.
I still crave a bigger, brighter viewfinder. I understand it's hard to do, and may even require brand new optical technology, or even ... gasp ... electronic viewfinders. In the end the easiest way may be to go back to FF. I don't know. Actually, maybe the easiest way would be to get some laser eye surgery that gives you better than 20/20 The price of that is coming down too...

But you know, another thing that would have made it all better would have been a plastic, soft-in-the-corners, soft-wide-open, 35mm/1.7 DA for $125 or as a kit lens.

Still nursing that grudge against modern camera makers, ever since I looked through the first Digital Rebel with kit lens.