The part about "steppers and lenses capable of exposing an entire full frame sensor at once" was something that was speculated about by one guy and was later shown to have little foundation in reality. Nobody knows for SURE what Canon has but there are good reasons to believe that they are still stitching full frame sensors. Also, the other large reason--that being how many FF sensors a wafer yields--still is highly relevant.
Finally a single sensor is not built from "one wafer", the wafer is a large-ish disc that is used to produce a bunch of 'em.
Thank you for the clarification pingflood. It sounds like you know much more about this than I. Yes, I remembered something about one sensor being built at once, but it sounds like you are indicating that might be a myth. I wonder if that as well might lend to the FF sensor being more expensive as well?
theophilus wrote:
I am a EE in the semiconductor industry - the difference in price between chip A and chip B (this is a real example)
Chip A = 18 sq mm
Chip B = 44 sq mm
Theoretical maximum yield of Chip A is 96% for a given defect density for a wafer
Theoretical maximum yield of Chip B is 92% for the same defect density for the same size wafer
It gets exponentially worse as the die size increases.
-------------
A FF sensor is 24x36mm = 864 sq mm.
I bet the yield is 25% or lower.
On top of that cost to manufacture difference they charge more because they can....Show more →
It gets even worse when you jump from APS-H to FF. APS-H fits on one stepper mask. FF requires three exposures for each sensor, and they must be precisely aligned. This is why Canon reports a extreme increase in the cost moving to FF, multiple exposure lithography is expensive and low yield - beyond what you'd expect from defect densities.
At really large sensor sizes (MF backs) they apparently give up on trying to achieve successful alignment across so many lithographic exposures. Instead the sensor is divided into sub-arrays and the RAW converter must use special processing to account for the misaligned sub-arrays.
wjmeyer wrote:
So, for Olympus, if they thought they could sell a whole bunch of FF OM-D's and make up for all the costs associated, they probably would, but knowing how technology is exponentially increasing with sensor design and knowing their huge investment in four thirds to date (many, many lenses) the question becomes why? Simply to satisfy the desires of some photographers who would like a mirrorless or small format DSLR in FF, with small pancake lenses, etc. to boot?
obviously they could sell a ton of FF OM-Ds, the question is whether they could sell a ton of them at a profitable price. i don't know anybody who would prefer to keep their µ4/3 camera if they could have the identical camera/lens ergonomics and size with a FF sensor for the same price. if it is possible, somebody will do it and bury olympus. the issue is that it is not the same price, so each company has to look at their estimated supply and demand curves to determine whether it would be profitable.
wjmeyer wrote:
The fear I think they all have is would it be profitable. For Sony, they already make a bunch of APS-C sensors, so for the NEX-5 and new 7 series, they are "sharing" sensors between products and not coming out with something completely new. If Sony wasn't already using the sensors for their NEX series in their DSLR series would they have still made the NEX, that is questionable because if the NEX flops, they haven't lost out on all the effort to build the sensor specifically for that body, but if Olympus puts forth the effort for a FF sensor and it flops, they stand to lose a lot more....Show more →
sony also makes a ton of FF sensors (though obviously not as many since sony sells more aps-c sensors than the total number of FF cameras sold by anyone i think). none of these companies except sony have to put forth any research on FF sensor development. sony already makes them and sells them to other camera companies. oly can just stop buying sensors from panasonic and start buying FF sensors from sony (obviously i'm over simplifying). the pieces already exist. oly would just have to look at the cost of the parts and decide if the product is feasible (obviously factoring camera manufacturing and design costs).
Wow, I'd love to sit down with a bunch of you guys to understand more about the technology behind the images (seriously). As a photographer coming from the film days I've always been concerned mostly with the quality of images that I create. That is what caused me to prematurely tell my colleagues and others that I would never shoot digital in the early 2000's, I naively thought that it would take 10-20 years for digital to catch up with film, boy was I wrong. But digital has its own pandora's box as well. Back in the film days I had two labs locally that I could trust to turn my 35mm negatives and slides into fantastic enlargements, but I had no idea what they did back there in the darkroom; however, with digital, by and large we (the photographers) have become the darkroom gurus and have to do our own touchups and editing, so while digital has provided some great opportunities for photographers, it also has some more demands on us as well. I must admit, I love playing around with Lightroom and Photoshop, but there are also days that I wish I could hand my work off to someone, tell them what I want and they give me a print that blows me away. Yes, I'm sure there's some boutique shops that do this, but I have found them hard to find and honestly, now that "I" have control, and because of my perfectionistic tendencies, I usually spend the time myself these days. In fact, I spend a whole lot more time in post production work today than I ever did when I shot weddings with film
sebboh wrote:
sony also makes a ton of FF sensors (though obviously not as many since sony sells more aps-c sensors than the total number of FF cameras sold by anyone i think). none of these companies except sony have to put forth any research on FF sensor development. sony already makes them and sells them to other camera companies. oly can just stop buying sensors from panasonic and start buying FF sensors from sony (obviously i'm over simplifying). the pieces already exist. oly would just have to look at the cost of the parts and decide if the product is feasible (obviously factoring camera manufacturing and design costs)....Show more →
Good point, and hopefully something that might happen with the latest Sony/Olympus hand shaking. Who's to say that if the OM-D line sells very well with MFT that they won't come out with an OM-D FF series... this is exactly what Canon did with the 1D and 1Ds, so maybe down the road we'll see an OM-Ds and a line of new lenses
wjmeyer wrote:
Good point, and hopefully something that might happen with the latest Sony/Olympus hand shaking. Who's to say that if the OM-D line sells very well with MFT that they won't come out with an OM-D FF series... this is exactly what Canon did with the 1D and 1Ds, so maybe down the road we'll see an OM-Ds and a line of new lenses
It would be fun if things eventually get cheap and simple enough that some entrepeneur could make e.g. low volume backs for OM cameras, where you'd just replace the hinged back with a slightly thicker one containing a sensor and related electronics. Then you would have to figure out how to turn the sensor on for exposure, of course, which would be a challenge it itself... maybe an electronically linked soft release button that lets the electronics know when you are about to take a picture.
the moment for me was when i realized that shooting a few hundred extra frames cos me nothing but my time. and i was so glad i had ebayed the color JOBO, the Imacon scanner, the medium format lenses and bodies and now i could be a bum again with a 20D
Beats me! He's comparing 8x10 film to the 80mp IQ180... We're talking about 35mm stuff.
the moment for me was when i realized that shooting a few hundred extra frames cos me nothing but my time. and i was so glad i had ebayed the color JOBO, the Imacon scanner, the medium format lenses and bodies and now i could be a bum again with a 20D
For me it was when I tallied the prices and digital hit the 4mp mark. But I wasn't really waiting or wanting to make some kind of "change". It just happened naturally. And I still don't feel like I've made some "change"... I'm just using a digital camera is all.
But it is 8x10 of film vs. a comparatively "tiny" digital sensor! 8x10 should "win" that contest, and will for sometime to come probably.
At the same format size, film lost the battle for most uses awhile ago. For the upper echelons of resolution it now comes down to cost really - if you don't shoot many exposures 4x5/8x10 film is hard to beat. If you shoot a lot the IQ180 is hard to beat. If you are constrained by size in any way digital is really hard to beat.
Not operating in the stratosphere of resolution digital took the title from film for me sometime ago.
Well - maybe I shouldn't post this - wandering way off topic in to done to death "film vs. digital" land. Against my better judgment I'll hit the "send reply" button.
For me it was the Canon 5D that killed 35mm film for me. But even now, when I'm doing taking my time doing serious landscape stuff, the Mamiya 7 still beats my 5Dll (confirmed by the test Ken considers to be done correctly) and I'm just scanning at 4000 dpi not 8000 dpi. However, if I had a 60+ mp MF back, I might not use the Mamiya 7 -- but then I wouldn't use the 60 mp back enough to justify its price, so even on an economic basis (for me) the Mamiya 7 wins.
Now, back to our regularly scheduled programming.
;-)
kwalsh wrote:
It gets even worse when you jump from APS-H to FF. APS-H fits on one stepper mask. FF requires three exposures for each sensor, and they must be precisely aligned. This is why Canon reports a extreme increase in the cost moving to FF, multiple exposure lithography is expensive and low yield - beyond what you'd expect from defect densities.
At really large sensor sizes (MF backs) they apparently give up on trying to achieve successful alignment across so many lithographic exposures. Instead the sensor is divided into sub-arrays and the RAW converter must use special processing to account for the misaligned sub-arrays.
Not only that, but Canon and Sony have their own fabs. Some of the cost of the sensor is masked in their bottom line as the fixed operating costs and depreciation of their fab equipment. For an outsider like Olympus having to source a FF sensor would be more expensive.
All in all there will be "hurrays" from the users and there will be "what were they thinking" moments. At the end of the day, these companies are in it to make money. Corporations will make decisions based on what they think can help them gain a market edge and ultimately... more money. Olympus is committed to four thirds and most recently to MFT, their lenses have matured and they have one of the best lineup's for anything not DSLR size, where they have struggled is in IQ, and for me, I'd rather invest in a product that has the lenses that I like, need or want with the understanding that the sensors will only get better. How long will it take Sony, or Samsung or Fuji or whomever to get the lineup of lenses that Olympus/Panasonic already have? Olympus has a pretty big headstart and not only that, they have the engineers who know how to make fantastic lenses in a small format. Yes, I am surely keeping my eye on Sony, Samsung and on Fuji now to see how much effort they'll put toward maturing their CSC market, but for right now, I think I made the right choice with MFT, especially if the OM-D lives up to its rumored hype.
My first venture into digital was the Canon 1D Mark II with a 20D, but I was never thrilled with the IQ of the 20D, the Mark III did not tickle my fancy so I stuck with the Mark II and then went to the 30D and was pleasently surprised, but I was always bugged that I shot with FF glass exclusively and I always missed something at the wide end, so I bought into some EF-S alternatives and ultimately decided against them because I didn't like being limited with a lens that could only be used on one type of body. Finally, colleagues convinced me to check out the 5D and when I finally did I wondered why I had never done it before, it had everything I had hoped for and with fantastic IQ. Having been a FF convert I was highly disappointed in the 5D Mark II (for various reasons) and decided to go with Nikon and the D700's and was pleasantly surprised by even better IQ than I had got with the Canon system. No need to discuss all the nuances of my decisions or how I came to these conclusions, my point is this, we ought not to be married to any manufacturer, product or technology because there's always going to be changes, achievements and breakthroughs. Had you asked me last year if I thought MFT could offer the exceptionaly IQ that it does with their lenses I probably would not have believed you, but having used the system now for several months I am a believer, and this is using sensor technology that is a few years old. I have come to realize that Olympus has struggled with its sensor technology, but the Panasonic 16MP sensor shows improvement over the 12 and if rumors are correct and the OM-D uses a completely new sensor then I have high hopes that this will be a big winner.
theophilus wrote:
Not only that, but Canon and Sony have their own fabs. Some of the cost of the sensor is masked in their bottom line as the fixed operating costs and depreciation of their fab equipment. For an outsider like Olympus having to source a FF sensor would be more expensive.
Not so sure about that. Each division has to show profitability Sony sells a lot to outside of Sony, and the sensor sold to these other companies wind up in products that are cost competitive to Sony products, so it is not clear how much of any advantage Sony cameras get from having their sensor manufacturer within the same corporate umbrella.It is also therefor reasonable to assume that Olympus using a Sony sensor would get that at a price that was competitive, and not significantly more expensive. Canon may not sell sensors to ther camera makers, but there are several other suppliers in the market. Further competition in the market would lead offerings from these other manufacturers -- like Kodak's former sensor division (now owned by private equity), Toshiba, Samsung, Dalsa, Aptina -- to be priced competitively for competitive products. Not clear to me that Olympus would have a significant or detrimentally more expensive sensor cost than Sony or Canon.
wjmeyer wrote:
Olympus is committed to four thirds and most recently to MFT, their lenses have matured and they have one of the best lineup's for anything not DSLR size, where they have struggled is in IQ, and for me, I'd rather invest in a product that has the lenses that I like, need or want with the understanding that the sensors will only get better.
+1
That's exactly why it is almost certain that my MILC system will be OM-D + 12/2 + 20/1.7 + 45/1.8.
