JoeV
Thin Air, Bright Sun
Er, not exactly. This would be equivalent to taking a hacksaw to a circuit board in the hopes of making a smaller circuit board that still functions. Individual die on a silicon wafer are cut out, using diamond saws, at assembly/test, but they are cut along the scribe lines between individual chips. You can't hack a photo-diode array and hope that the remaining cells of the array will still function. The thickness of the saw cut is much larger than the micron-sized diodes. There are also read-out and processing circuitry in the periphery of each die surrounding the diode array; so hacking the chip destroys the peripheral circuitry, which functions effectively as a custom graphics processor.
Individual die cost in semiconductor manufacturing are most efficiently improved by engineering improvements in the manufacturing line that increase line yield (the percentage of wafers that make it to end of line intact) and die yield (the percentage of die on each wafer at end of line that function properly). It is highly likely that kodak's chip factory makes the S2, M8 and supposed M9 image sensors in the same manufacturing line on the same machines, but they are each processed in distinctly separate lots; the lithographic masks for each type of sensor are distinctly different, so different reticles (the litho mask patterns at each layer) are used in the steppers for each different product type.
Where cost savings for the S2 chip are made with the M8 and/or M9 chips is most likely due to the same machines being used to make all three, so that a higher profit-margin product like the M8 chip (smaller die size, more die per wafer, higher revenue per wafer due to smaller defect levels due to decreased area size) offset the higher cost, reduced yields of the larger sensors.
Also, any process improvements in the line that increase S2 die yield will also improve M8/M9 die yield (since they are all three made on the same machines), improving profitability for these other products, too.
~Joe
EDIT: I must also add that it has been a common practice for years to employ a "masking ROM" in parallel with an image sensor, whereby dead pixels in the sensor array (FYI: there are no perfect sensors) are mapped in a custom-burned ROM that accompanies each sensor, such that the live pixels surrounding the dead ones are read and their voltage levels averaged, so as to synthesize the dead pixel's appropriate voltage. So the output from the sensor chip is processed through the masking ROM to mask, or hide, the dead pixels. This goes a long way to making more die per wafer practically salable.
An example would be my Lumix G1, which has a menu feature whereby the sensor can be remapped and dead pixels effectively can be eliminated from the image. This most likely employs a masking ROM that is flash-memory based; the image sensor is remapped and the new map of dead pixels written into the flash memory masking ROM.
FYI: It is also probable that masking ROMs may now be manufactured within the periphery of each image sensor die, rather than packaged from a physically separate ROM chip; the size of a ROM memory cell can be made much smaller than a CCD or CMOS image sensor pixel, since light collection area is not an issue with memory arrays.
Similarly, in microprocessor manufacturing, the die with more defects usually operate at slower speeds, so they are binned out and sold as lower end products. The processors that test at higher speeds are conversely sold as higher-end products.
Last edited:
Wiyum
Established
My math puts an m9 FF chip with pixel pitch identical to the S2 at 24MP... A good bit more than I'd want, if ISO performance is to suffer.
jaapv
RFF Sponsoring Member.
Well, to be quite honest, I do not put much credence in the chip recovery rumour either. The same machine/technology aspect seems to be far more logical to me. It would be great if they used the same basic architecture for both the M9 and S2 sensor with a lower pixel count for the M9 - it would do wonders for noise performance
bean_counter
Well-known
It would make sense that the same sensor would be used in both the S2 and M9; particularly if the sensor fits in the new (larger?) M9. Then, there is no need to trim it down physically; just turn off excess pixels.
Regardless, I would doubt there would be enough falied S2 chips to meet M9 demand. More likely scenario is that S2 chips would be cherry-picked from the best of the run.
Regardless, I would doubt there would be enough falied S2 chips to meet M9 demand. More likely scenario is that S2 chips would be cherry-picked from the best of the run.
Richard Marks
Rexel
I am a bit cocerned that if the economicsof the m9 production are linked to the S2, if the S2 fails it will not exactly help the M9. I do not have any close contacts in Leica, but common sense tells me S2 sales are not going to be too great. Fashion and advertising pros would be the main target Market and a number are struggling at the moment. It probably is not the time to sell ones H3d and start a whole new system with a company who can not offer service centre pro support in most countries.
Richard
Richard
climbing_vine
Well-known
I have to think that the manufacture cost difference between a 35mm "full frame"-sized sensor and the one in the S2 is much greater than the cost of doing separate runs that can largely use the same equipment. There's a reason that manufacturers with a wide range of sensor sizes in their cameras don't do this...
JoeV
Thin Air, Bright Sun
The die cost (cost of each chip) is related to the area of each chip in relation to the entire wafer's area. It costs the same the run the wafers through the production line, regardless of what's patterned onto them. So larger die (like the S2 chip) take up more surface area; the production costs and expected profit have to be factored into the cost of these larger, fewer chips. They will therefore be more expensive to produce; and this is ignoring yield issues caused by larger sized die having more total defects than smaller die (defects are usually related to surface area of the die).
Conversely, with a smaller die like the proposed M9 FF chip, the die size is smaller than the S2; more can fit on each wafer; they will yield higher at end of line; thus can be priced appropriately for the product range.
But putting a dumbed-down S2 die in place of an actual FF die is bad business; it costs the same as a full-resolution S2 chip (i.e. it has the same surface area; you can only fit so many of them on a wafer) and therefore you'd have to sell them for the same money as an S2.
It's smarter business management to run a higher volume product (like an M8 chip or M9 FF chip) in the same line as an S2 chip; each product gets their own litho reticle masks, they all share the same production equipment, and the higher volume production of the smaller chips helps offset the cost of the S2. Semiconductor fabs run more efficiently at high volumes.
Just to clarify, silicon wafers are ran in production lots of 25 wafers each; a lot in a plastic box with a cassette holding each wafer in its appropriate slot. The newest processes use 300mm (12 inch) diameter silicon wafers, while older processes use 200mm (8 inch) diameter wafers. A fab can run multiple types of products simultaneously; this is usually a sound business decision, as reduced sales in one product can be offset by increasing sales in others. Don't put all your eggs in one basket. As long as the basic transistor process is compatible between product types (i.e., they all use the same process technology) the only difference between product types is the litho reticle masks that pattern the wafers.
So the business cost of a chip fab is related to cost per wafer. And the number of die that can fit on each wafer is related directly to its surface area. Thus the larger chips will intrinsically cost more.
Also, chips are rectangular; wafers are round. So there is dead space near the wafer's edge where you can't fit a complete chip pattern. The larger the chip, the more dead space near the wafer's edge; this has the effect of biasing production efficiency for smaller sized die, since more complete ones can be patterned near the wafer's edge. Hence the photography manufacturers have pushed large megapixel, small-sized sensors for the purposes of maximizing production efficiency; more of them can be fit on each wafer, and more megapixels is a selling point that can position the product at a higher price point in the market.
~Joe
Conversely, with a smaller die like the proposed M9 FF chip, the die size is smaller than the S2; more can fit on each wafer; they will yield higher at end of line; thus can be priced appropriately for the product range.
But putting a dumbed-down S2 die in place of an actual FF die is bad business; it costs the same as a full-resolution S2 chip (i.e. it has the same surface area; you can only fit so many of them on a wafer) and therefore you'd have to sell them for the same money as an S2.
It's smarter business management to run a higher volume product (like an M8 chip or M9 FF chip) in the same line as an S2 chip; each product gets their own litho reticle masks, they all share the same production equipment, and the higher volume production of the smaller chips helps offset the cost of the S2. Semiconductor fabs run more efficiently at high volumes.
Just to clarify, silicon wafers are ran in production lots of 25 wafers each; a lot in a plastic box with a cassette holding each wafer in its appropriate slot. The newest processes use 300mm (12 inch) diameter silicon wafers, while older processes use 200mm (8 inch) diameter wafers. A fab can run multiple types of products simultaneously; this is usually a sound business decision, as reduced sales in one product can be offset by increasing sales in others. Don't put all your eggs in one basket. As long as the basic transistor process is compatible between product types (i.e., they all use the same process technology) the only difference between product types is the litho reticle masks that pattern the wafers.
So the business cost of a chip fab is related to cost per wafer. And the number of die that can fit on each wafer is related directly to its surface area. Thus the larger chips will intrinsically cost more.
Also, chips are rectangular; wafers are round. So there is dead space near the wafer's edge where you can't fit a complete chip pattern. The larger the chip, the more dead space near the wafer's edge; this has the effect of biasing production efficiency for smaller sized die, since more complete ones can be patterned near the wafer's edge. Hence the photography manufacturers have pushed large megapixel, small-sized sensors for the purposes of maximizing production efficiency; more of them can be fit on each wafer, and more megapixels is a selling point that can position the product at a higher price point in the market.
~Joe
climbing_vine
Well-known
JoeV, that was very informative. Good show.
ferider
Veteran
One addition to Joe's well-written summary:
Unless you consider products with several 10-thousand wafers output per month (for example memories), the cost of the final die in a modern fab is largely set up, for the process and its ramp, the masks, etc. In particular for a sensor as used in M8, M9, etc. If you make a sensor with a custom process for a single end customer and a few thousand end products (cameras), the sensor price will be very high - similar to printing a book in a few hundred copies only.
If the M9 sensor is manufactured in a modern 300mm fab, both Leica and the fab will have high interest in use of the sensor or of the identical process in other products as well (for instance via the shuttles that Joe mentioned); only increased die count means reasonable price per die. If this can not be done, the sensor would be more expensive than a few thousand dollar per die. In other words a custom manufacturing process for Leica only would make no economic sense.
I am convinced a standard process is being used for both S2 and M9 sensors. Maybe the same. The S2 does not require any vignetting correction via new sensor technology. Its registration distance is long enough. If the M9 sensor has enough bits/pixel and sensitivity, it will be able to correct vignetting via software. The necessary lens coding has already been customer-accepted. So you will use a few bits (up to 4 ?) with wide angles. Big deal if you have, say 16bit/pixel and use a new 21mm Summilux
No Super Angulon use at 16000 ASA though
Leica marketing (and associates ) said a digital RF was not possible and then produced the M8.
Then they said a FF digital RF was not possible due to short registration distance and "un-avoidable" vignetting on sensor.
Go figure.
Cheers,
Roland.
Unless you consider products with several 10-thousand wafers output per month (for example memories), the cost of the final die in a modern fab is largely set up, for the process and its ramp, the masks, etc. In particular for a sensor as used in M8, M9, etc. If you make a sensor with a custom process for a single end customer and a few thousand end products (cameras), the sensor price will be very high - similar to printing a book in a few hundred copies only.
If the M9 sensor is manufactured in a modern 300mm fab, both Leica and the fab will have high interest in use of the sensor or of the identical process in other products as well (for instance via the shuttles that Joe mentioned); only increased die count means reasonable price per die. If this can not be done, the sensor would be more expensive than a few thousand dollar per die. In other words a custom manufacturing process for Leica only would make no economic sense.
I am convinced a standard process is being used for both S2 and M9 sensors. Maybe the same. The S2 does not require any vignetting correction via new sensor technology. Its registration distance is long enough. If the M9 sensor has enough bits/pixel and sensitivity, it will be able to correct vignetting via software. The necessary lens coding has already been customer-accepted. So you will use a few bits (up to 4 ?) with wide angles. Big deal if you have, say 16bit/pixel and use a new 21mm Summilux
No Super Angulon use at 16000 ASA though Leica marketing (and associates
) said a digital RF was not possible and then produced the M8.Then they said a FF digital RF was not possible due to short registration distance and "un-avoidable" vignetting on sensor.
Go figure.
Cheers,
Roland.
Last edited:
bwcolor
Veteran
Your logic makes much sense. We can now calculate the MPixels of this new FF and might even suggest that a third, even smaller, but higher volume camera could be in the works. After all, you want to maximize profit and minimize costs.
Richard Marks
Rexel
All this means if S2 sales are poor, the m 9 is very much at risk.
Richard
Richard
jaapv
RFF Sponsoring Member.
Fortunately not the case. A post by Dave Farkas on LUF (being a major dealer) tells us that sales are excellent, with the price meeting the expectations of the intended customer group. Having said that I am convinced that there are contingency strategies in place if there are problems of that kind.
. A post by Dave Farkas on LUF (being a major dealer) tells us that sales are excellent, with the price meeting the expectations of the intended customer group. Having said that I am convinced that there are contingency strategies in place if there are problems of that kind.
Last edited:
jaapv
RFF Sponsoring Member.
Your statement contradicts Leica's that was expressly that the S2 did use shifted microlens technology.One addition to Joe's well-written summary:
Unless you consider products with several 10-thousand wafers output per month (for example memories), the cost of the final die in a modern fab is largely set up, for the process and its ramp, the masks, etc. In particular for a sensor as used in M8, M9, etc. If you make a sensor with a custom process for a single end customer and a few thousand end products (cameras), the sensor price will be very high - similar to printing a book in a few hundred copies only.
If the M9 sensor is manufactured in a modern 300mm fab, both Leica and the fab will have high interest in use of the sensor or of the identical process in other products as well (for instance via the shuttles that Joe mentioned); only increased die count means reasonable price per die. If this can not be done, the sensor would be more expensive than a few thousand dollar per die. In other words a custom manufacturing process for Leica only would make no economic sense.
I am convinced a standard process is being used for both S2 and M9 sensors. Maybe the same. The S2 does not require any vignetting correction via new sensor technology. Its registration distance is long enough. If the M9 sensor has enough bits/pixel and sensitivity, it will be able to correct vignetting via software. The necessary lens coding has already been customer-accepted. So you will use a few bits (up to 4 ?) with wide angles. Big deal if you have, say 16bit/pixel and use a new 21mm Summilux
Leica marketing (and associates
Then they said a FF digital RF was not possible due to short registration distance and "un-avoidable" vignetting on sensor.
Go figure.
Cheers,
Roland.
Pickett Wilson
Veteran
If I were a major dealer in Leicas, that's what I would say, too. "Man, these S2's are blowing out of here. Better get your's before they are gone!"
I guess if you are a company that only sells 140 total cameras a month, though, maybe even 10 orders for the S2 would seem brisk.
I guess if you are a company that only sells 140 total cameras a month, though, maybe even 10 orders for the S2 would seem brisk.
TJV
Well-known
While I think David is most likely an honest man, I'm not sure how many people have paid in full for these S2's yet considering there are many details regarding full specs etc yet to be released. Not to mention no one has seen the RAW files. Not until the cash is in the till and the cameras out the door do I believe "sales are excellent."
I do wish Leica well though and, honestly, the S2 is a system I'm looking at for my personal work. I'll wait a while to see the results though!
I do wish Leica well though and, honestly, the S2 is a system I'm looking at for my personal work. I'll wait a while to see the results though!
Fortunately not the case
jaapv
RFF Sponsoring Member.
Here is the link. Post 286
http://www.l-camera-forum.com/leica...available-october-starting-price-15-a-10.html
http://www.l-camera-forum.com/leica...available-october-starting-price-15-a-10.html
Ben Z
Veteran
Well I've met David on several occasions and he strikes me as a candid and straightforward individual, not at all the type who slings bull. He's also probably the most photographically-knowledgeable of any Leica dealer I've met. I believe what he says in terms of the waiting list for the S2.
If any suspicion can be cast, it would be toward the people on his list. I don't doubt there are plenty of amateurs who just must be the first on their block to get one, in the same way there are many who will just have to have an M9 before anyone else. Here in South Florida where I am and David is, the economy sucks for average folk, but I still see brand-new Rolls Royces and Ferrarris running around with paper tags, and the parking lot is always full at Bal Harbour Mall (way upscale) and all the uber-expensive trendy restaurants, so plenty of people still have plenty of disposible income. But it shocks me that a pro would commit to trade his working system for one that nobody has ever seen a file from. As someone in an equipment-based profession I would never consider risking losing clients by being an early adopter of untried technology, no matter how sterling the manufacturer's past reputation has been. And in terms of teething troubles, Leica's reputation isn't that stellar.
I'm happy if a company like Leica can turn a profit regardless of whether their prices fit my definition of sane or they don't, and I'm happy if guys like David and Tony Rose and others can make a profit too. However I'm not interested in an S2 (wouldn't be if it was half its price), and they'll have to wait to sell me an M9 when they get demos. I'm just not shooting as much anymore as I once did, so it doesn't make much sense to keep so much cash locked up in a closet.
If any suspicion can be cast, it would be toward the people on his list. I don't doubt there are plenty of amateurs who just must be the first on their block to get one, in the same way there are many who will just have to have an M9 before anyone else. Here in South Florida where I am and David is, the economy sucks for average folk, but I still see brand-new Rolls Royces and Ferrarris running around with paper tags, and the parking lot is always full at Bal Harbour Mall (way upscale) and all the uber-expensive trendy restaurants, so plenty of people still have plenty of disposible income. But it shocks me that a pro would commit to trade his working system for one that nobody has ever seen a file from. As someone in an equipment-based profession I would never consider risking losing clients by being an early adopter of untried technology, no matter how sterling the manufacturer's past reputation has been. And in terms of teething troubles, Leica's reputation isn't that stellar.
I'm happy if a company like Leica can turn a profit regardless of whether their prices fit my definition of sane or they don't, and I'm happy if guys like David and Tony Rose and others can make a profit too. However I'm not interested in an S2 (wouldn't be if it was half its price), and they'll have to wait to sell me an M9 when they get demos. I'm just not shooting as much anymore as I once did, so it doesn't make much sense to keep so much cash locked up in a closet.
Last edited:
Pickett Wilson
Veteran
Well, I'm waiting for a deal on all those used Hasselblads that folks are dumping to order S2's that nobody has seen a raw file from. I can't believe working professionals would do such a thing.
jpfisher
Well-known
That is simply because the M8/M8.2 is not an SLR. It's listed on B&H's Rangefinder page.
Andrew Sowerby
Well-known
That's interesting. A very reasonable size for a bigger-than-FF camera.
... not that it makes any difference to me. I'm about as likely to buy an S2 as I am to buy my own airplane.