"The NEX-7's 24MP pixel count pays off in high levels of detail, accurately resolving up to at least 3400 lines per picture height. Beyond this some demosaicing artefacts and colour moiré start to creep in, and past 3600 lph there's little meaningful detail. As we'd expect from the pixel count, this is pretty well as good as it gets without making the considerable step up to medium format - impressive stuff from such a small camera."
Really looks like a great Alt. camera. They mention in the conclusion the usual caveat with NEX - few worthwhile native lenses so far - but for Alt. glass that really isn't an issue at all.
Overall, what this all means is that while the NEX-7 offers an excellent platform for using old lenses, or experimenting with some slightly left-field options for different image looks, the native E-mount lens range is arguably its Achilles' heel at the time of launch (although this situation will undoubtedly improve over the camera's lifetime).
Unfortunately the ZM Biogons are not good candidates for the NEX-7, presumably due to the AA (and possibly other) filter in front of the sensor, nor are other wide angle non retrofocus designs in the Voigtlander and Leica line-ups. Too bad.
Still, given what I've experienced with the NEX-5N and now the GXR, the increased speed and accuracy of focusing on the GXR is not an advantage I am willing to trade in for a NEX-7 in exchange merely for more pixels and improved noise performance at higher sensitivity levels. The more I use the camera the less I can imagine getting rid of it down the road.
michaelwatkins wrote:
Unfortunately the ZM Biogons are not good candidates for the NEX-7, presumably due to the AA (and possibly other) filter in front of the sensor, nor are other wide angle non retrofocus designs in the Voigtlander and Leica line-ups. Too bad.
Well since RF wide performance is totally case by case on the 5n, its way to soon to predict how the n-7 will work with a 35 biogon. The ZM18 may be just fine, the 28 ultrons ARE retrofocus, and there are likely many safe choices at 35mm.
It will be a least a month before we can really get a grasp on the n-7 and RF wides.
Looks like the JPEG engine is fairly heavy handed with sharpening and noise suppression. Hopefully whoever gets their hands on it first can post up some good photos processed from RAW. They seem to like the IQ very much at lower ISO which is all I care about in a camera.
I've been cold on the NEX7 ever since Steve Huff's review but I'm kind of re-warming to this camera. I'll probably just adapt my alt DSLR lenses that aren't even that big and pick and choose RF lenses that play well with this sensor. That or maybe wait and see what 2012 brings.
uhoh7 wrote:
Well since RF wide performance is totally case by case on the 5n, its way to soon to predict how the n-7 will work with a 35 biogon. The ZM18 may be just fine, the 28 ultrons ARE retrofocus, and there are likely many safe choices at 35mm.
It will be a least a month before we can really get a grasp on the n-7 and RF wides.
There is a noticeable difference in corner shift and extreme corner resolution with the ZM 35/2 on the 5 vs. the 5N. I was surprised how much better the 5N was with that lens.
The 28 Ultrons, and most Leica wides, aren't true retrofocus. They're just not totally symmetrical. All have various degrees of issues on a digital sensor, even a Leica sensor, which is why they still need to be coded on a Leica. The ZM 18 seems to do fine on the 5N, but I wouldn't wager that it'll be problem free on the 7.
I am relatively convinced that the 5N is still the better Sony sensor for rangefinder lenses, but we'll see. As Michael has found, the GXR is probably better than both.
Vignetting and magenta cast aside, even at f/8 we aren't seeing the corners *or* edges clean up. To some degree I see the same repeated even on the 5N with the ZM 25/2.8 although the edges do become acceptable enough when stopped down. On the GXR in contrast the edges and corners look great by f/4 and just get better - I get the sense that I am seeing the real performance of the lens, at all apertures. That missing AA filter is accounting for more than just improved resolution it seems.
Poor Rob was pilloried for doing his work, even though he applied what appeared to be a consistent methodology (focus on a single spot for all lenses, manly tripod, etc.) yet his work gave us the heads up and queasy feeling that at least some lens designs - notably the wider Zeiss - were going to be problematic on the NEX-7. In reality he gave us the first useful heads up.
To some extent I'm guilty of hoping for a better result against the mounting evidence regarding the NEX-7. I recall seeing Rob's shots (and the 21 and the 25 and to a lesser extent the 35) and feeling my stomach churn inside because I *so* wanted the NEX-7 to work out with rangefinder wides, not that I was all that attached to Sony but I wanted a camera with a built in EVF.
It turns out firmer edges and corners are more important to me than a built in EVF which principal additional benefit is easier sliding the camera in and out of my bag, and it seems that my bag has learned to accommodate a camera with an external EVF.
(edit: actually it turns out that a lot of things are more important to me than a built-in EVF, and knowing that, I guess, is a good outcome regardless)
Hilariously, I just bought this whole set of FDn lenses, but I cancelled my NEX-7 preorder a few days later. Despite there probably being no issues with the NEX-7 and my new lenses, I'm not ready to give up the form factor of the 5N + half case + EVF. I may be the only person in the world that wants the 5N body with the 7 sensor.
...Even with the resolution bump of the NEX7 the NEX5N does look to have the superior sensor. I guess one could normalize the resolution of either to make it fair but the 5N looks nice and clean at base ISO.
douglasf13 wrote:
Hilariously, I just bought this whole set of FDn lenses, but I cancelled my NEX-7 preorder a few days later. Despite there probably being no issues with the NEX-7 and my new lenses, I'm not ready to give up the form factor of the 5N + half case + EVF. I may be the only person in the world that wants the 5N body with the 7 sensor.
It takes all kinds.
;- )
And re. the above -- you're absolutely right that comparing at comparable output resolutions is the only way to make the comparison relevant. I have to say that the 5n (as well as the Nikon D7000 and D5100) is the first crop camera with base ISO quality that looks even remotely as good as an FX Nikon. Noise in broad gradients and shadows is more common than of us would like to admit. I hope the NEX-7 is as good there.
Are the problems with wide rangefinder lenses something that can be corrected with firmware or is it a design problem? Would it have the same problem with SLR lenses?
With the price of the 5n and EVF, I'd really prefer to have the 7. Whether or not it handles alt wides as well as the 5n may be my deciding factor. I'm anxiously awaiting some FMers to get their hands on one and give us some real samples.
flashinm wrote:
Are the problems with wide rangefinder lenses something that can be corrected with firmware or is it a design problem? Would it have the same problem with SLR lenses?
It's either a the thickness of the IR/AA filter pack or micro lens design. While firmware could potentially help the color shift issue in the future, it can't restore resolution issues. This isn't a problem with SLR lenses, as they are much more telecentric.
