philber wrote:
Another issue that is worth noting IMHO is that, because 24 ZA is a native E-mount, it is very likely that the NEX 7 has multiple in-cam lens corrections for it which are already visible in the menu of 5N...
Also, nobody commented on it, but MR showed a small pic of an angel, quality of which was mouth-watering.
Hard to tell sharpness since that shot of baby Jesus is only a ~500 x 300 pixel thumbnail shot. The crops from the church look pretty good for far distance stopped down f4 and more.
FlyPenFly wrote:
It should be pretty easy to tell from the RAW files in the same way you can see MFT does some huge amounts of corrections.
Why are MFT users always so self conscious?
Every time, every time MPX goes up the same old arguments come up, every time.
Because the same old arguments are mostly still valid. I did a print comparison between files from an E-5 and an A900 earlier this year (identical motives), and although the A900 showed more detail, the differences were surprisingly small. High resolution requires high quality lenses and the larger the sensor the larger the lens needs to be. The cost of making those lenses tend to increase even more.
One reason why I tend to use 4/3 and m4/3 rather than one of my 5 F-mount bodies is that the lenses are better unless I spend money on the latest, greatest Nikkor primes. The PanaLeica 14-50mm f/2.8-3.5 is simply an amazing lens with hardly any weak points at all. I can't say that I have found any "normal" zoom lens from any manufacturer (except the Zuiko 14-35mm f/2.0, and that's mostly at f-stops larger than f/2.8 ) that betters it. It's a large lens compared to the sensor, but it does show that telecentricity works.
When that is said, m4/3 lenses aren't strictly telecentric the same way as the 4/3 lenses are supposed to be. Software corrections and microlenses on the sensor have to a certain degree been able to replace the optical principle. With the 5N and 7, Sony seems to have succeeded with this too, but I cannot say that I think they have succeeded delivering 24MP of imagery like in 150% more than 16MP or 100% more than 12MP imagery, and I'm quite certain that the problem is the lens and/or the lens/sensor relationship. MR's comparison clearly shows that there's a lot to be extracted from that sensor and that the potential for those who want to print really large is clearly there.
"...the lenses available from Sony in NEX mount are struggling with 24MP."
"Doubling the number of pixels improves the transfer function even if
the sensor resolution is better than the resolution of the lens."
"The curve for the poor lens on the 24 MP sensor is almost as good as the
curve of the good lens with the 12 MP sensor."
Last two quotes are from Dr Nasse of Zeiss. Sony also understands optical theory. It is not a common understanding, nor is it at all well appreciated.
High Mp cameras always 'struggle' their way to better image quality regardless of lens quality - for a given level of lens quality - as compared with lower Mp sensors.
Dr Nasse again:
"Thus concerns that today's good lenses may in general not be able to cope with a
24 MP sensor appear somewhat exaggerated. The reason for an overly pessimistic view is the misconception that only the resolution limit of the system determines the image quality and that it is identical to the resolution of the weakest link of this chain. This is not the case, though, since the curves are multiplied..."
So a hypothetical user with a set of merely average to good lenses will see a significant boost in image quality from moving up to 24Mp from say 16Mp...with the usual caveats. The 'bottleneck' theory of image quality is a fallacy, however intuitively correct it may seem. Also, storage is cheap, drives are high capacity and modern computers are fast...and cheap. One reads very few posts on Sony forums about the need for much better lenses to 'keep up with' their high Mp cameras...this is the reason why.
It's hard to disagree with that. Still, buying a better lens for an old camera body will in my experience almost always improve image quality more than buying a new body with more megapixels. That is of course not contradictive to what the gentleman from Zeiss states and neither is my claim that 100 more pixels doesn't necessarily give 100% more resolution using the same lens.
The question for me, and for many other photographers, is what gives more value for money. For me, buying a better lens for one of the camera bodies I already have almost always makes more sense. That way, I can improve the "look" of my photos. Since some of my 12MP photos are already printed wall size to great satisfaction for my clients, I struggle to see the advantage of even more pixels.
Mind you, I don't say that more pixels are bad. I just can't see how it makes sense to spend money on it when I could spend that money on another great lens or for an airline ticket to Myanmar or Borneo where people couldn't care less how many megapixels or command wheels my camera has.
That is unless I adapt the philosophy of a PJ I met years ago who owned one lens, a 35mm, for his F5 and who pointed the camera in the general direction of the subject, letting the editorial staff do the necessary cropping
FlyPenFly wrote:
I have the 5N and the 7 on order.
Do you want to take bets on which will have more detail using the same Contax G lenses?
i bet the 45/2 has no trouble and out resolves the nex-7's sensor. if what little (and poorly executed) we've seen from the nex-7 and rangefinder lenses holds true, i bet your 28/2.8 will have more detail in the corners on the nex-5 then the nex-7 at any given print size.
) that betters it. It's a large lens compared to the sensor, but it does show that telecentricity works.