 |
 | Home | Contact Us | FAQ | Search & Site Map | Link to Us |
 | Sign In | Join | Other 45 Sites in Network |
| PhotoKB HomeDiscussion Groups Digital Photography Digital PhotoDSLR CamerasZLR CamerasPoint & Shoot CamerasFilm Photography 35 mmLarge FormatMedium formatDarkroomFilm and LabsOther EquipmentPhoto Technique Nature PhotographyPeople PhotographyTechnique GeneralGeneral Photo Topics General TopicsAustralian PhotographyUK PhotographyDirectoryPhoto Clubs |
Photo Forum / Digital Photography / DSLR Cameras / July 2005New backs to give 35mm format shooters the "drools"
31 Mpix not enough? How about 39 Mpix? http://www.dpreview.com/news/0507/05071802phase1_3backs.asp Doesn't state the bodies, but prob. Hassy and Mamiya models offered. Price not stated... but as the man said, if you need to ask, you can't afford it. Cheers, Alan > 31 Mpix not enough? How about 39 Mpix? > [quoted text clipped - 6 lines] > Cheers, > Alan It's $30K (U.S.), - what a bargain! BTW, you can use the back on any camera that accepts Hasselblad A or H series backs, or Mamiya backs. You can also use the back on view cameras via adapters. I'm sure they'll sell like crazy! <not> Walt >> 31 Mpix not enough? How about 39 Mpix? >> [quoted text clipped - 16 lines] > >Walt Lets compare cost to resolution: Phase 1 39mp $30k = 1300 pixels per $1. Canon 16mp $8k = 2000 pixels per $1. Olympus E-300 8mp $800 = 10000 pixels per $1. With lenses! Cheapo Fuji 4mp $250 = 16000 pixels per $1. With a lens! What does this mean? Nothing. :) As this comparison shows, it's not the number of pixels, but the size of the sensor that drives the price. The challenge is to produce a defect-free piece of silicon. If the manufacturing process has an average of one defect per area the size of a medium format sensor, half of them will end up bad. On the other hand, if the same process is used to make the tiny sensors of a P&S, the same piece of silicon will yield perhaps 99 good sensors and one bad. A much cheaper manufacturing process, like one having ten defects over the same area will still yield around 80-90 good small sensors, allowing to produce them and make good money. But no matter how many times one tries, such process will practically never deliver even a single defect-free medium format sensor. Not exact numbers by any stretch of imagination, but you can see the logic behind the price ratio. > Lets compare cost to resolution: > Phase 1 39mp $30k = 1300 pixels per $1. [quoted text clipped - 3 lines] > > What does this mean? >As this comparison shows, it's not the number of pixels, but the size of the >sensor that drives the price. The challenge is to produce a defect-free [quoted text clipped - 8 lines] >format sensor. Not exact numbers by any stretch of imagination, but you can >see the logic behind the price ratio. True enough, but ... Most semi-conductor manufacturers practice "yield" management. For microprocessors, it means testing all "dies" at the highest speed rating (for your current product line). Those that pass get packaged up and labeled at that fastest speed, and sold for a correspondingly high price. Those that don't pass, they get tested at the next speed rating, "one down" from the fastest. The same yield process. Those that don't pass this slower speed test get rejected and then tested again at the next lower speed, and so on and so forth. If the manufacturing process is good, not that many "dies" are actually defective and have to be discarded. Same process for memory chips, and also for photosensors, CMOS and CCD. So in principle all sensors can be tested for that 54 MB Hasselblad-back part. Those that fail, which is probably most of those parts, will be sliced up and then tested again as ???? size the manufacturer also sells. Part of the logic behind the pricing is that there is a market at that high price. At some point, the price for all these sensors will drop as volumes increase and manufacturing processes are tweaked to improve yields. It has always been this way in semiconductors and will probably always be this way. Back in the long-ago day of single-sided and double-sided floppy disks, Verbatim and others practiced the same yield management for floppy disk media, Father (grandfather??) Kodak [] > Most semi-conductor manufacturers practice "yield" management. For > microprocessors, it means testing all "dies" at the highest speed [quoted text clipped - 14 lines] > those parts, will be sliced up and then tested again as ???? size the > manufacturer also sells. Not for imaging chips, though. You can't re-slice them. Idon't know if any manufacturers make a large imager from two smaller chips physically abutted - I would guess not for the market segments we are considering. Whether there might be anything to be gained by putting, say, four sensors at a single "pixel" site, and being able to discard one dead one, I don't know. I suspect not, as the dead ones amy come in clumps. David >[] >> Most semi-conductor manufacturers practice "yield" management. For [ ... ] >> Same process for memory chips, and also for photosensors, CMOS and >> CCD. So in principle all sensors can be tested for that 54 MB [quoted text clipped - 5 lines] >any manufacturers make a large imager from two smaller chips physically >abutted - I would guess not for the market segments we are considering. To provide a line on which they could be divided, you would have to leave a section of the image uncovered, so there would be either a cross-hairs through the center of the image, or a bunch of processing to synthesize the missing pixels. And a practical line for parting would be many rows of pixels wide. Remember -- the sensors are not all that is on the chip. There are also the addressing or shifting circuits which lead to readout devices (and perhaps even to A/D converters) on the borders of the sensor chip. Slice out a bad area, and you may wind up with a good sensor with no way to get the signals off the chip. And whenever a chip is sliced, there is some damage, so a border is needed to keep the damage from reaching the active circuitry. The two most common ways (at least back when I knew about it) were to scribe and then flex to break into individual chips, or to use a fine wire saw, carrying an abrasive slurry to cut through the material. >Whether there might be anything to be gained by putting, say, four sensors >at a single "pixel" site, and being able to discard one dead one, I don't >know. I suspect not, as the dead ones amy come in clumps. Aside from that, each sensor must use up surface area. Four sensors in a single pixel area would mean 1/4 the area, and thus a corresponding increase in noise. (Though that could be reduced somewhat by averaging the four when all four are good.) This would also increase the processing needed to make the raw image. And, as long as you can get four sensors in the space of a single pixel, you can just as easily increase the resolution of the sensor by a factor of four, so which way do you think the manufacturers will go, given how many people consider a simple increase from 6MB to 8MB to be a "killer improvement". :-) Enjoy, DoN.  SignatureEmail: <dnichols@d-and-d.com> | Voice (all times): (703) 938-4564 (too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html --- Black Holes are where God is dividing by zero --- > Most semi-conductor manufacturers practice "yield" management. For > microprocessors, it means testing all "dies" at the highest speed > rating (for your current product line). Those that pass get packaged > up and labeled at that fastest speed, and sold for a correspondingly > high price. Lately the yield management in CPUs has focused more on binning by lowest power. You pay a big premium for the lowest power chips. The el-cheapo no-name notebook PCs are using the highest power parts, both in CPUs and graphics, which results in more heat, and shorter battery life. >> Most semi-conductor manufacturers practice "yield" management. For >> microprocessors, it means testing all "dies" at the highest speed [quoted text clipped - 6 lines] >el-cheapo no-name notebook PCs are using the highest power parts, both >in CPUs and graphics, which results in more heat, and shorter battery life. True enough. Very good addition. The main point is that the "binning" process always allows a manufacturer to eke out revenues from parts that are not the "best." F K > As this comparison shows, it's not the number of pixels, but the size of > the sensor that drives the price. The challenge is to produce a [quoted text clipped - 8 lines] > a single defect-free medium format sensor. Not exact numbers by any > stretch of imagination, but you can see the logic behind the price ratio. And this is why anyone who thinks cheap full frame sensors are just around the corner are dreaming. SignatureStacey >> As this comparison shows, it's not the number of pixels, but the size of >> the sensor that drives the price. > >And this is why anyone who thinks cheap full frame sensors are just around >the corner are dreaming. Except that you have to take the costs of producing a 'small' sensor into account. If we assume that Nikon makes a profit on the D50, then APS-C sensors can't cost all that much. The step from APS-C to full frame 35mm is not all that big. Even though relatively speaking larger sensors will be more expensive than smaller sensors, that does not imply that larger sensors have to be very expensive in an absolute sense. SignatureThat was it. Done. The faulty Monk was turned out into the desert where it could believe what it liked, including the idea that it had been hard done by. It was allowed to keep its horse, since horses were so cheap to make. -- Douglas Adams in Dirk Gently's Holistic Detective Agency > 31 Mpix not enough? How about 39 Mpix? > [quoted text clipped - 3 lines] > Price not stated... but as the man said, if you need to ask, you can't > afford it. I thought Leaf already made a 52MB (?). Can someone please explain to me just what a "back" is, as compared to a camera "body"? Many thanks, CC > Can someone please explain to me just what a "back" is, as compared to a > camera "body"? > > Many thanks, > > CC The "back" is the portion of the camera that holds the film. With medium format and large format cameras, they are typically interchangeable, so you can have multiple backs for one camera body. The digital backs in question simply replace the film backs on existing cameras, so you don't buy an entirely new camera to go digital, you just buy the back. However, these backs are extremely expensive because of low volume production and very high pixel counts to satisfy the needs of demanding commercial photographers. Walt Yikes! So you're saying you can spend $30,000 and then you STILL need a camera and a lens to be able to take photos? > > Can someone please explain to me just what a "back" is, as compared to a > > camera "body"? [quoted text clipped - 16 lines] > > Walt > Yikes! > > So you're saying you can spend $30,000 and then you STILL need a camera > and > a lens to be able to take photos? I think it is safe to assume that anyone in the market for this back would already have an extensive collection of bodies and lenses. Walt |
Reply to this Message |
 Sign In
 Join
 My Latest Posts My Monitored Threads My Blog My Photo Gallery My Profile My Homepage Start New Thread Enable EMail Alerts Rate this Thread
|
©2009 Advenet LLC Privacy Policy - Terms of Use
This website includes both content owned or controlled by Advenet as well as content owned or controlled by third parties.