PentaxForums.com - View Single Post - Pentax

mattdm · 01-26-2008, 08:45 PM

Originally Posted by 24X36NOW

I'd argue that 4/3 is not intended to, nor will it ever, compete with FF 35mm sized digital for IQ, because whatever sensor technology advances there are that will imrpove the ability to put more on a given chip size will also benefit FF 35mm size chips, i.e., they will always be better.

Clearly. But currently, that's a practical disadvantage because there's trouble getting ISO 800 (let alone 1600) performance. In the not-so-far future, the IQ at that sensor size will be so good that the theoretical gains from getting any better don't apply in practice to the vast majority of the market.

And there certainly is huge amounts of research being poured right now into increasing image quality on tiny sensors -- look at all the P&S cameras (not to mention camera phones) sold with sensors smaller than a little fingernail. Image quality sells in that market too, so those problems will be solved.

Looking at how AMD's x86_64 crushed Itanium out of the marketplace is instructive -- Intel's next-generation 128-bit CPU architecture is clearly technically superior to the kludged-onto-1970s-tech x86_64 design -- but x86_64 could be sold in the mass market where Itanium had no chance, so modern x86_64 chips are now actually faster than Itaniums except in specific niches.

I'm not personally putting any huge bets on either horse (although I'm perfectly secure in buying 1.5× crop Pentax lenses today), because I won't be really surprised to see the camera market go either way. If full-frame wins the day, I don't think it'll really be on image quality (even though that claim is true) but because Canon and Nikon will see dumping huge amounts of marketing money into it as a way to shake off the pesky little companies. (As they're trying now with their campaigns for lens-based SR.)

01-26-2008, 08:45 PM	#7
mattdm Pentaxian Location: Boston, MA Gallery Photos: 8 Posts: 2,012	Originally Posted by 24X36NOW I'd argue that 4/3 is not intended to, nor will it ever, compete with FF 35mm sized digital for IQ, because whatever sensor technology advances there are that will imrpove the ability to put more on a given chip size will also benefit FF 35mm size chips, i.e., they will always be better. Clearly. But currently, that's a practical disadvantage because there's trouble getting ISO 800 (let alone 1600) performance. In the not-so-far future, the IQ at that sensor size will be so good that the theoretical gains from getting any better don't apply in practice to the vast majority of the market. And there certainly is huge amounts of research being poured right now into increasing image quality on tiny sensors -- look at all the P&S cameras (not to mention camera phones) sold with sensors smaller than a little fingernail. Image quality sells in that market too, so those problems will be solved. Looking at how AMD's x86_64 crushed Itanium out of the marketplace is instructive -- Intel's next-generation 128-bit CPU architecture is clearly technically superior to the kludged-onto-1970s-tech x86_64 design -- but x86_64 could be sold in the mass market where Itanium had no chance, so modern x86_64 chips are now actually faster than Itaniums except in specific niches. I'm not personally putting any huge bets on either horse (although I'm perfectly secure in buying 1.5× crop Pentax lenses today), because I won't be really surprised to see the camera market go either way. If full-frame wins the day, I don't think it'll really be on image quality (even though that claim is true) but because Canon and Nikon will see dumping huge amounts of marketing money into it as a way to shake off the pesky little companies. (As they're trying now with their campaigns for lens-based SR.)