mawz wrote:
Registration distance is irrelevant to the issues in question. However non-retrofocus lenses do have telecentricity issues which can lead to both CA and soft corners. This is an ongoing problem with some wide-angle lenses when adapted to the m43 cameras, as well as on the M8 and M9 (although the offset microlenses on the Leica's significantly reduce this problem)
Well, please bear in mind also that the poster is an avid pro-4/3 speaker - that has no history whatsoever of admitting the built in faults of the Oly 4/3 system (as compared to anything at all, from P65+ through D3x/5D2 down to compacts)... A disturbingly high percent of 4/3-users are totally oblivious to the outer world. Extreme brand-loyalism is even more prevalent here than with Nikon/Canon/Leica users.
Regarding the angle issues, we'll just have to wait and see. It all depends on the filter-package that Sony decided to throw onto the really brilliant base sensor (that's actually quite a lot better base than the Leica's Kodak offerings... Colour quality remeains to be seen though). The seemingly lackluster performance of the first lenses from Sony has no bearing on this subject at all.
Now IF the problem was coincidence angle related as opposed to lens-related we would see the Leica-typical extreme field colour casts and also Leica-typical extreme vignetting caused by the rather incompetently designed (and il-cheapo-constucted!) filter package in the M8/9 placed on top of the very angle-sensitive Kodak sensor base. As there's none of that, I'd say the corner problems are strictly lens-related.
The biggest dampening on my enthusiasm regarding the first two NEX's is not about optical/sensor performance, but rather on the manual exposure control system. Direct mechanical input (wheels) for the two all-important aperture and shutter-time settings would be preferable. I've already set aside money for the Nex 5, but we'll see if any details regarding the Nex 7 surfaces before the "5" model hits the shelves. I just might wait for that if the exposure controls don't work smoothly for me.
Otherwise, as the mount seems to provide ample space for a quick&simple wide-converter (or even a simple 10mm flat float insertion for a 20% WC) the new E-mount seems brilliant. As long as the WC optic can take it from a quality POV, there's easily room for at least an F/1.5 limited converter. I've already sketched two versions for prototyping...
Joakim you sure seem to have a strong opinion about M8/M9 and their incompetently designed (albeit quite different) sensor systems.
Just to clarify a few of your not quite qualified statements.
really brilliant base sensor (that's actually quite a lot better base than the Leica's Kodak offerings... Colour quality remeains to be seen though).
I'm sure it will be a great sensor but there is very little information to class it as 'brilliant' before any real shots from the system have been taken and shared by real photographers.
It has a high ISO advantage over m4/3, do you have some other information with regards to dynamic range, low ISO noise and sharpness that you aren't sharing? Colour reproduction would be quite a definite factor in my evaluation of a sensor, and how it deals with high incident angles. Which leads me to the next point.
Leica-typical extreme field colour casts and also Leica-typical extreme vignetting caused by the rather incompetently designed (and il-cheapo-constucted!) filter package in the M8/9 placed on top of the very angle-sensitive Kodak sensor base. As there's none of that, I'd say the corner problems are strictly lens-related.
The angles of incidence on the M9, having a full frame sensor can be as extreme as 45 degrees to the film plane, or even more severe on the Hologon for example.
There are no vignetting issues with the M9, it vignettes exactly as film did with the same lenses. I have first hand experience with this. Please share your experience if it is to the contrary.
Leica also chose a thinner IR filter as a thicker one would degrade image clarity and contrast, this may be to blame for the red colour shift on the M9.
We also don't know if Sony are using software correction for known lenses in-camera. Leica employ a similar system for their own system lenses (which works great for all "Leica" lenses except the 21mm super angulon), and we might get a nasty surprise when we mount alt glass on these NEX cameras, but no one knows for sure. The cropped format will certainly help though.
The "smearing" in the corners on the NEX could be a result of simply transplanting the A550 sensor into the NEX cameras without a thought to microlens redesign... But that's unlikely as it's so fundamental in a system with a short registration. The requirement of telecentric lenses on the m4/3 system is one way to avoid colour shift (the severe crop helps as well), and the smearing with ultra-wides like the voigtlander 12mm on the m4/3 cameras is a clear indication that the microlenses don't have an extreme enough offset for such lenses on those cameras.
You do seem to have quite a biased opinion, for someone who comes in guns a-blazing against someone with brand loyalty.
thrice wrote:
Joakim you sure seem to have a strong opinion about M8/M9 and their incompetently designed (albeit quite different) sensor systems.
Just to clarify a few of your not quite qualified statements.
I'm sure it will be a great sensor but there is very little information to class it as 'brilliant' before any real shots from the system have been taken and shared by real photographers.
It has a high ISO advantage over m4/3, do you have some other information with regards to dynamic range, low ISO noise and sharpness that you aren't sharing? Colour reproduction would be quite a definite factor in my evaluation of a sensor, and how it deals with high incident angles. Which leads me to the next point....Show more →
Actually, the sensor is a known item, being a minor variant of the sensor in the A450/A550. It really is an excellent sensor, although there are indications that Sony does not know how to get the best out of it (the closely related 12MP version performs noticeably better in Pentax's K-x than in Sony's A500)
Expect significantly better noise performance from ISO 640 and up than the m43, dynamic range which is competitive with the best of APS-C, excellent sharpness and good(but not class-leading) colour. If you want a good indication of what the sensor will deliver look at the reviews of the A550.
Yakim Peled wrote:
This looks to me like a definite proof that the NEX can't accept a FF sensor.
Happy shooting,
Yakim.
I'd disagree, the extra room in A mount is occupied by the aperture mechanism, which E mount doesn't have. E mount appears to be slightly larger diameter than M mount, which is known to be able to take a FF sensor.
i was going to say... the NEX mockup looks fairly similar to the M9 in the relationship of the sensor to the mount so I wouldn't rule it out... its just a matter of how they would deal with short register effects on a FF sensor (with a register that is even shorter than M-mount)
morpheus2891 wrote:
i was going to say... the NEX mockup looks fairly similar to the M9 in the relationship of the sensor to the mount so I wouldn't rule it out... its just a matter of how they would deal with short register effects on a FF sensor (with a register that is even shorter than M-mount)
The mount itself is quite wide in diameter and the lenses themselves can be significantly "fatter" than M lenses, which would help the light rays to hit the sensor at a less extreme angle. Somewhat like what Olympus claims for its designed for digital lenses.
morpheus2891 wrote:
i was going to say... the NEX mockup looks fairly similar to the M9 in the relationship of the sensor to the mount so I wouldn't rule it out... its just a matter of how they would deal with short register effects on a FF sensor (with a register that is even shorter than M-mount)
The 'short register effects' is actually due to optical design and lenses with a rear element close to the sensor. The register itself is immaterial to it (the reason you don't see it in SLR designs is due to mirror clearance requirements, not the long register of the mount although the former does drive the latter as well). Any mirrorless design with symmetric wide-angle lens designs will have to deal with this issue regardless of mount register.
yea...adapters for M to E shouldn't be a problem... the only thing that sucks is that its 1.5x so you loose the wide angle for lenses like the ZM 18/4... well, that and the fact that there isn't a high quality viewfinder (EVF please... so we can do macro and telephoto)..... seeing the pictures that Thrice makes with his M9 and ZM 18 makes me drool... obviously the tallent of the photographer brings out the potential of the equipment but my... I love the potential
According to DPreview, Sony says that the 16mm lenses that are out for review are all preproduction lenses and not the final versions, which might account for the appearance of images from those lenses.
j4ake wrote:
According to DPreview, Sony says that the 16mm lenses that are out for review are all preproduction lenses and not the final versions, which might account for the appearance of images from those lenses.
"What? You don't like the lens? Ummmm... that's not the REAL lens we're making. The real one will be much better than this piece of junk!"
I think Sony was a little surprised at the unfavorable response to the 16/2.8 samples, issued this rather unprecedented statement in hopes of shutting down talk of soft lenses and too-short register distance, at least until launch. Just how I see it....
mawz wrote:
I'd disagree, the extra room in A mount is occupied by the aperture mechanism, which E mount doesn't have. E mount appears to be slightly larger diameter than M mount, which is known to be able to take a FF sensor.
I'll keep my fingers crossed that you are right and I am wrong.
morpheus2891 wrote:
yea...adapters for M to E shouldn't be a problem... the only thing that sucks is that its 1.5x so you loose the wide angle for lenses like the ZM 18/4... well, that and the fact that there isn't a high quality viewfinder (EVF please... so we can do macro and telephoto)..... seeing the pictures that Thrice makes with his M9 and ZM 18 makes me drool... obviously the tallent of the photographer brings out the potential of the equipment but my... I love the potential
Any M lens wider than around 35mm will probably have softness issues in the corners on NEX. Those lenses just weren't designed for digital (which is why Leica has had to deal with offset microlenses.)
As for manual focusing, that high res screen should be fine, and you could always use a hoodman loupe.
CKrueger wrote:
"What? You don't like the lens? Ummmm... that's not the REAL lens we're making. The real one will be much better than this piece of junk!"
I think Sony was a little surprised at the unfavorable response to the 16/2.8 samples, issued this rather unprecedented statement in hopes of shutting down talk of soft lenses and too-short register distance, at least until launch. Just how I see it....
Yep. I think you have it figured out. I was just trying to be diplomatic.
Chris: if you do; please report back and try and sneak in a SD card and make a few shots with the 16mm to see if they really have tweaked it's performance up to par.
