If I remember correctly Rob indicated he had set focus on the front brick wall.
It would be hard not to nail focus on that wall with a lens like the ZM85 so I'm inclined to take him at his word on that. That said it is hard to imagine the shot at f/4 looking so bad if focus was nailed. Yet isn't that the current controversy being discussed - sorry, I'm not following it that closely so forgive me if I've got it wrong (and please correct me) - that even some longer focal length lenses are being treated badly by the sensor in the NEX-7?
I found the wider angle shots easier to to make conclusions about -- looking at the ZM25 example from that series, wide open and stopped down to f/8 the image shows the same degradation of detail at the edge and corners. The lens probably was and should have been at the infinity stop for that shot with that lens or terribly close to it, and when stopped down to f/8 the area of acceptable focus would be very large. I have this lens and know it can render a subject like this sharp from one edge of the frame to the other and it doesn't have to be stopped down much at all to get there.
Yet even at f/8 the details 1/4 the way all around the frame, regardless of the distance from the camera, are heavily smeared. Given the contrasty light even the centre of the frame at 100% view doesn't look all that sharp either to my eye but maybe a little sharpening in post is all it needs although that won't do much for the edges.
I tested extreme corners with the CV 35/1.2 and contax-g 90/2.8 wide open yesterday, I can't see the problem. I can see soft corners, but that is natural at (especially) 1.2 and 2.8.
michaelwatkins wrote:
If I remember correctly Rob indicated he had set focus on the front brick wall.
It would be hard not to nail focus on that wall with a lens like the ZM85 so I'm inclined to take him at his word on that. That said it is hard to imagine the shot at f/4 looking so bad if focus was nailed. Yet isn't that the current controversy being discussed - sorry, I'm not following it that closely so forgive me if I've got it wrong (and please correct me) - that even some longer focal length lenses are being treated badly by the sensor in the NEX-7?...Show more →
no, there are no reports other than this of the NEX-7 having poor performance in the center. in fact michael reichmann showed it outresolving the m9 with the 50 lux in the center (while losing to the m9 in the corner). there is no explanation i'm aware of for those 85mm shots being so soft in the center other than missed focus. this makes me doubt all the other tests shots by rob as well.
michaelwatkins wrote:
I found the wider angle shots easier to to make conclusions about -- looking at the ZM25 example from that series, wide open and stopped down to f/8 the image shows the same degradation of detail at the edge and corners. The lens probably was and should have been at the infinity stop for that shot with that lens or terribly close to it, and when stopped down to f/8 the area of acceptable focus would be very large. I have this lens and know it can render a subject like this sharp from one edge of the frame to the other and it doesn't have to be stopped down much at all to get there.
Yet even at f/8 the details 1/4 the way all around the frame, regardless of the distance from the camera, are heavily smeared. Given the contrasty light even the centre of the frame at 100% view doesn't look all that sharp either to my eye but maybe a little sharpening in post is all it needs although that won't do much for the edges....Show more →
if the lens actually hits infinity at the lenses infinity stop with the adapter then yes it shouldn't make a difference, however most adapters overshoot infinity when you take the lens to it's infinity stop, which will definitely make a difference. i've seen other shots near infinity with the NEX-7 and some of these wides that don't look as bad (though they still show smearing), so again i don't trust his tests at all.
I think I skimmed over a couple of discussions and some how the assimilated information got tangled. It wasn't centers but longer focal length lenses that were being called into question on NEX-7 - where it seems some users of longer focal length lenses on NEX-7 were observing similar behaviour in the edges as the problematic shorter focal length symmetrical wides deliver on that camera.
nex-7 is looking more and more like a bridge too far
other reports suggest similar issues with alpha zooms and zeiss 85.
and
http://forum.getdpi.com/forum/sony/3...blem-cz85.html
As a result of a discussion on the Photozone camera forum, I tried the CZ85 on both the NEX-5n and the 7. it turns out that the 7 is not only problematic with symetrical designs that have a short focal length, but also at longer focal lengths with a wide aperture.
On the NEX-5n the CZ85 is close to perfect (at 1.4). Sharp from corner to corner. It is getting even better when you stop down a bit, also because there is a lot of LoCA wide open, but it looks very impressive. With the NEX-7 there is a clear bending of the focal plane in the extreme corners and even when focused in the extreme corners, there is less detail than with the 5N.
At say f2.8-f4 the effect is mostly gone. ...Show more →
I'm not trying to import a controversy from somewhere else, only understand what the current thinking is with respect to a camera I'd ordered but then cancelled due to such concerns on the wider end of focal lengths. Maybe I've not been paying attention but it seems news to me to read concerns that longer focal length lenses may show the same "better performance on the 5N than the 7" trait that we saw with symmetrical wide angle RF lenses, and even an A mount SLR lens has a sticky finger pointed at it. Egads!
Michael, it is my brief experience with NEX 7 that it is a very demanding camera. For example, on a Leica WATE at 16mm, a lens of almost unlimited DOF, it will mercilessly reveal the difference between hyperfocally sharp and critically sharp. Much more so that on NEX 5, or than a 21 Distagon ZE on a 5D II. To the extent that, to get critical sharpness form a NEX 7, I need to keep the shutter speed much higher than on the NEX 5N. That, of course, only gets "worse" with longer FLs. This could explain what is happening with the 85. I will test the NEX with my own Contax G 90, to see if I can replicate the issue and prove my theory.
yeah, it seems possible that non-wide-angle fast glass may have similar issues as particularly fast lenses also generate rather oblique angles. from the actual samples i've seen it doesn't seem like a huge issue, but it may very well be worse than the 5N.
my only point was that rob's tests still seem extremely suspect in terms of methodology and results. there is NO reason the zm 85 should look that soft in the center at f/4 other than poor focus.
i'm still on the fence about the NEX-7 myself. i'm not much of a wide angle person and most of the lenses i'm interested in using seem to do fine on it. the zm 18 is my one concern (it would be my widest serious lens), i know it color shifts, but that doesn't bother me. i want to know how much smearing it shows and over how much of the image. in any event, i have some time to find out as i force myself to wait two years between upgrades, so i have till summer to decide.
philber - "a very demanding camera" - now that sounds right to me. Expect less forgiveness as the pixel count goes up, right?
Re the loss of detail at the edges, thanks to the Zeiss white paper speaking about AA-filter induced astigmatism I've assumed the AA filter is different between the NEX-5N and NEX-7, but perhaps the primary difference can be accounted for by the combination of sensor with perhaps even the same AA filter as the NEX-5N.
sebboh - Have to agree - I'm at a loss to understand how that 85 @ f/4 shot could look like that unless there was a focus or infinity stop problem. If the adapter has a hard time hitting infinity for a ZM85 chances are the other ZM's are affected. Mine all seem to hit infinity on a dedicated M mount camera in exactly the same way so if Rob's ZM lenses also follow suit, and the 85 is an indication of focusing short or past (more likely probably) on a given adapter, then they'd probably all be affected in the same manner.
Michael, I owned 3 ZM lenses on different Voigtländer and Novoflex adapters, and have used a number of Leica lenses on the same, and, more often than not, they went past true infinity. So I would put not too much faith on "blind" infinity focusing on the hard stop. Certainly not with a NEX 7 and 85mm lens.
philber wrote:
Michael, it is my brief experience with NEX 7 that it is a very demanding camera. For example, on a Leica WATE at 16mm, a lens of almost unlimited DOF, it will mercilessly reveal the difference between hyperfocally sharp and critically sharp.
I know the EVF has helped me alot on the 5n. Focus accounts for many imperfections no doubt.
I just go back to the LL samples with the 50 lux where the edges at f/1.4 made me gasp.
It's a crop: how could the edges be so bad?
Of course what got the press was the f/8 duel and how the n7 "out-resolved" the m9.
We expect alot from sony. Look at the history of new releases...issues...issues.
There is no doubt the n7 can produce great imagery---for large prints its a real ground breaker.
As a platform for many lenses used in a variety of situations it simply does not live up to our hopes. The first version anyway.
It really seems beyond question now that the 5n is a far more versitile and reliable body for our glass.
It;s going to be fun to see those shots which break through the n7 limitations and perhaps....stun us.
Charlie, I just created a quick-and-dirty Cornerfix profile for the ZM 18, and it cleans up beautifully. No trace of colour shift left, and it is definitely not soft or mushy in the corners. I may not buy the WATE, after all.
As I said above: tested both the 35/1.2 and the 90/2.8 wide open on my nex-7, it preforms very good in the extreme corners. I'm not sure what the problem is with some testers/their cameras, but so far I'm very happy with my nex-7. the only shortcoming so far is the size of the sensor..
philber wrote:
Charlie, I just created a quick-and-dirty Cornerfix profile for the ZM 18, and it cleans up beautifully. No trace of colour shift left, and it is definitely not soft or mushy in the corners. I may not buy the WATE, after all.
I was wondering when you were going to try that---your reports were so dark "shifting horribly"...I was thinking decayed copy of 60s godzilla movie
Finally, and more to the point, the NEX-7′s image quality is indeed much higher than the NEX-5N. Be sure to click the individual frames to view the difference in detail between the two cameras. Here’s the deal: not only does the NEX-7 have 50% more pixels than the NEX-5N, but per-pixel sharpness also seems slightly higher, as if the AA filter was slightly less strong on the 7. I’d love to see large prints compared to the Leica M-9 and Nikon D3x.
ulrikft2 wrote:
As I said above: tested both the 35/1.2 and the 90/2.8 wide open on my nex-7, it preforms very good in the extreme corners. I'm not sure what the problem is with some testers/their cameras, but so far I'm very happy with my nex-7. the only shortcoming so far is the size of the sensor..
The 35/1.2 has been one of the only known 35mm rangefinder lenses that works well on the Nex-7 for a while. It's all of the more reasonably sized 35mm lenses that have been questionable. Granted, I'd still be curious to see how CV compares in the corners to the 5N, because it could be even better at wider apertures.
Sorry, but I don't agree. I see 5N actually being better.
Yes, Nex-7 picture is bigger (more pixels), but it's softer, not sharper
and I actually see more detail in 5N image.
ulrikft2 wrote:
As I said above: tested both the 35/1.2 and the 90/2.8 wide open on my nex-7, it preforms very good in the extreme corners. I'm not sure what the problem is with some testers/their cameras, but so far I'm very happy with my nex-7. the only shortcoming so far is the size of the sensor..
douglasf13 wrote:
The 35/1.2 has been one of the only known 35mm rangefinder lenses that works well on the Nex-7 for a while. It's all of the more reasonably sized 35mm lenses that have been questionable. Granted, I'd still be curious to see how CV compares in the corners to the 5N, because it could be even better at wider apertures.
My 35/1.2 works better with the 5N than with the 5. (I have no 7). "Very good in the extreme corners wide open"? I don't think so (and I also don't think it is important for most users). With both the 5 and the 5N (but less with the later) there is some smearing going on towards the edges due to big difference between sagittal and tangential performance. I don't see how that can disappear with the 7.
Stopped down the 35/1.2 is really sharp, on par or possibly a hair better than the SMC-P FA31/1.8 I compared it to.
All this tells me is that testing methods and the use of words uncontrolled to some degree. What we need is direct comparisons with suitable targets.
Jonas B wrote: ulrikft2 wrote: douglasf13 wrote:
My 35/1.2 works better with the 5N than with the 5. (I have no 7). "Very good in the extreme corners wide open"? I don't think so (and I also don't think it is important for most users). With both the 5 and the 5N (but less with the later) there is some smearing going on towards the edges due to big difference between sagittal and tangential performance. I don't see how that can disappear with the 7.
Stopped down the 35/1.2 is really sharp, on par or possibly a hair better than the SMC-P FA31/1.8 I compared it to.
All this tells me is that testing methods and the use of words uncontrolled to some degree. What we need is direct comparisons with suitable targets.
Very good in extreme corners considering that it is a 35/1.2 being shot wide open.. Not "very good in the extreme corners at 1.2, rivalling the CV 125/2.5 stopped down to f/8 in the center".
Some things goes without saying (or so I thought).
I have revised this post as I have now white balanced the Cornfix blanks, in an effort to minimize the Cornerfix alterations to the DNG file to confine it to the corners, only. To avoid confusion, since I don't have a Heliar 15 revision, I am deleting it from this post.
The NEX 7 has forced me to use Cornerfix on the ZM25, and Summicron 35 ASPH. The Cornerfix blanks below were at infinity, white diffuser over the lens and +2 EV at a white poster board in the shade. All were stopped down 1 stop from wide open.
ulrikft2 wrote:
Very good in extreme corners considering that it is a 35/1.2 being shot wide open.. Not "very good in the extreme corners at 1.2, rivalling the CV 125/2.5 stopped down to f/8 in the center".
Some things goes without saying (or so I thought).
I think the tricky thing is the interpretation of "very good." It could be very good on the 7, and very, very good on the 5N.
