300mm is becoming a reality in semiconductor houses, namely nand flash. There will be a day when a 7.2gig file will be a large but not ungodly amount of data to store and transport. In 10 years digital will be at levels previously unattainable today. It would seem to me that rather then make the sensor larger improve the noise and density characteristics.Originally Posted by sog1927
And x86... because they sell in enormous volumes.
All of the BIG semiconductor shops (Intel, TSMC, UMC, Chartered, AMD, and even FreeScale) have been using 300mm wafers for several years now.
That would, however, be big enough for only one 8x10 sensor or maybe 8 4x5 sensors.
Compared to well over 100 Core2 Duos, let alone something as tiny as a NAND chip, and you can see why the costs would go up so much, particularly since you'd lose pixels from every chip when they're that big.
With current sensor technology, we're reaching the limit for density, I think. We'll get better digital to analog conversion, better noise-reduction, and improved read times.There will be a day when a 7.2gig file will be a large but not ungodly amount of data to store and transport. In 10 years digital will be at levels previously unattainable today. It would seem to me that rather then make the sensor larger improve the noise and density characteristics.
Also, current digital sensors are still monochromatic, with the exception of the Foveon x3. That means that the cameras are actually interpolating the vast majority of the data that they emit, since they're producing 10.2 megapixel images from a sensor with 10.7 million photosites, each of which is recording only one color, and ignoring the other two.
Digital imaging is very much in its infancy, it has a LONG way to go. And I don't think that our current technology has legs enough to get us anywhere significant, but there's probably already a lot of R&D effort going into developing a new imaging technology.
Digital technology has all but eliminated the film and analog print business. I'd say that's pretty significant. Technology keeps expanding and every time there is a limit there is a breakthrough. I understand that soon we will have 1TB hard drives on laptops because Hitachi just broke another barrier in the size of read/write heads.
You completely misunderstood my post, apparently. The point is that current technology isn't going to get us much farther.
As I said, there's probably already a lot of work going on to develop a new method of implementing digital sensors that will move us forward. Monochromatic, silicon-based sensors aren't it.Technology keeps expanding and every time there is a limit there is a breakthrough. I understand that soon we will have 1TB hard drives on laptops because Hitachi just broke another barrier in the size of read/write heads.
I think it's possible to get pretty close to large format film resolution with a medium format digital back and a sharp lens. Some guy sent me a shot from his phase-whatever back for hasselblad shot with one of those fancy schneider digital lenses, and it was pretty damned close to the same resolution as the best scan I can get from my V750 desktop scanner with a sharp transparency - possibly better. Because of the ability of a smaller sensor to capture extremely high resolution and allow movements for architecture with various setups I can't see any drive to develop a 4x5 or 8x10 sized CCD or CMOS sensor (which would be pretty cost prohibitive right now, and perhaps for a very long time).
The one thing you can never match with a smaller format digital are some of the optical idiosyncracies of large and ultra large formats (an 8x10 facial portrait is effectively a macro, with narrow DOF and all that), or the ability to dabble in interesting alternative processes (bromoil, chrysotype, van dyke, cyanotype, Pt/Pd, etc) although arguably you can do that with digital negatives. In any case, commercial users probably aren't generally concerned with Jim Galli-esque 8x10 portraiture and artistic fringe stuff like that, so it seems likely that the larger formats will remain (or become even more so) a fringe activity.
I do not doubt for a moment that the primary digital image produced by a high-end MF digital back with excellent lenses will be better than a secondary digital image produced by scanning an excellent LF transparency with a consumer scanner, and I specifically don't want to ignite the digital vs. film flamewar. As I said, it all hinges on what your definition (no pun intended) of "equivalent" is, and that's more a matter of taste than anything else. It's like arguing about when painting will be "equivalent" to sculpture (or, to rephrase it in terms from a century ago, when photography will be equivalent to painting). From a commercial standpoint, it is of course about what gets the job done. I think you're right that the answer to that will involve film less and less.
No high-volume user needs an "exact equivalent" of 8x10 in digital. For that reason, there's absolutely no financial incentive to create such a thing. The market for it would be miniscule and the development costs incredible. It currently costs literally *billions* of dollars to build a new technology semiconductor fab or even update an old one (see:
http://www.news.com/A-fab-constructi..._3-981060.html ).
Steve
I agree with you completely Steve. I was just talking about what makes sense from a commercial standpoint. I think film will eventually have no place at all in the arena of high volume photographers who simply pump out images for advertising and fashion and product brochures and the likes. It's already happened for the most part and I don't think there's a realistic need for sensors too much bigger than the existing reduced medium format digital ones in this area. Unfortunately for us, this is where the money (and hence development incentive) lies.
For the hobbiest and the artist it's a different matter. The quickest and most economical route to some kind of standardized, expected output often isn't the best one.
Well it wouldn't be the first time I've misunderstood a post.
It happens to all of usWell it wouldn't be the first time I've misunderstood a post.
I hope the explanation made sense though.
Digital has gotten much more adoption than its technological maturity implies. After all, until basically this year, our primary tool for image editing has been a graphic design tool with a codebase that dates back to before Windows NT -- and before digital photography was something normal people could afford
I am not convinced a technology has expired until a new technology appears.It happens to all of us
I hope the explanation made sense though.
Digital has gotten much more adoption than its technological maturity implies. After all, until basically this year, our primary tool for image editing has been a graphic design tool with a codebase that dates back to before Windows NT -- and before digital photography was something normal people could afford
How do you think they measure RGB out a single monochrome pixel? Do you think they use some form of programmable color filter to isolate a particular color spectra and then sample the energy in that band? That's the way I would do it provided I could manipulate the color filter accurately and fast enough to scan 'n' mega-pixels. Of course that's why the high end cameras have proprietary digital imaging processors.
Posting Permissions
Reply With Quote
Bookmarks