Wow , a 100 posts. All in a few days too. I 'll add my two cents.
To me 15 (pick a number) megapixels are 15 megapixels. Assuming the camera and lens are in an acceptable range, I find the prints to be largely comparable. This assumes lighting conditions, a lens sharp enough to resolve a sharp image on the sensor, a sound non vibrating platform (tripod), and good exposure technique. Thats what techniqually makes an image. The photographer adds the interpretation and composition.
Now the way I see it is " the devil is in the details".
All cameras will work well on an acceptable foundation, so that parameter is ok.
Exposure is somewhat chip dependent, but largely on just three parameters, noise (speed), number (of pixels), and cost ($). I'll come back to this point.
The photographer is a wild card.
Whats left is the lens. Thats why when I so often read, that Nikon is dead, I just laugh. There is "no" way that a major, high quality, glass company with all the cost advantages of large scale production of quality product is going anywhere but to be a survivor in an industry shakeout. Any company who is a player in high quality glass has the ability (potential) to survive.
Now back to the chip thing. For the point of the moment lets put aside cost. So, then all we want a sensor to do is gather data. How much is dependent on final print size. So we can, based upon our largest print size we expect to print, select the amount of data needed. (We can now rationalize uprezing if we bring cost back in).
So much data to collect, whats the best way to collect it. I know I need _?_ pixels. What does a bigger chip do for us? It makes life easier on the lens. Big pixels are easier on the lens requirement than small pixels. That seems fairly intuitive.
What about pixel quality? This is where I'm convinced significant progress will be made. And in two ways, hardware and software. Hardware will somewhere out there run into a wall (the darn physics thing), I think we have a little way to go yet. Software (either on chip or camera) or in the raw (jpeg) converter has the potential to make chicken soup out of chicken ****. This has a long way to go IMHO. Not only manipulating the data but allowing better yield in chip making. Locking out bad sectors has been happening in Hard Drives for a long time. Elegant ways of locking out bad transistors on a chip will become feasable at some point to improve yields someday.
On to the final point, we should expect better quality but were getting close to enough is enough. We should see significant breakthroughts in cost, with the lens and sensor coming into equilibrum with final requirements meeting the output needed.
I for one find it "nuts" to pay the rediculous prices the industry has sold the photography industry on. Although early adopters have funded all technology companies. I remember my first PC, A DEC rainbow, $7000+ with two floppies and a 10 Meg HD. Great monitor but monochrome. In 1982 dollars no less.
I think the brand arguments are silly, the format sizes interesting, but in the end we just want quality data. The path is lens first, chip second.
Two points before I close.
The idea that Canon will kill everyone else because they are bigger is not how it works. Canon is a great company with a broad array of products. They invest their R&D dollars in a number of different product categories, from camera's to copiers, to printers, to.... You get the picture. Each division gets R&D (or marketing) dollars because of two reasons, 1)greatest return or 2) defending the turf against outsiders. I doubt cameras get unlimited funds
Lastly, the chip guys when there competitively behind will gang up on the leader. Sony getting in bed with Nikon and others is an example. Canon jumped out there, there going to have to fight to keep it. Otherwise we'll never have a reasonably priced highpixel camera to create the LL.
Whew, I don't know where that came from, just started typing, Thanks for your indulgence.
