I'm going to quote myself, because I discovered something great.
Jochenb wrote:
Edward, I agree about the sensor stack. However, I don't think the native lenses really suffer from it. I also see the negative effects you describe, but I blame the lens designs. When I put on an adapted lens that I also used on a DSLR I don't really see a difference in the peripheral area. What I do notice is exaggerated field curvature. This is a result of that sensor stack. With the A7 stack you also get sensor reflections as a "bonus".
Last week I added an A7S to my A7. The A7S clearly has a better sensor stack.
1) The extreme field curvature is gone! My "affected" lenses act normal again, just like they did on a DSLR. It's not that I get fooled by the lower resolution that masks the issue. Planar subjects are tack sharp across the frame again. With the A7 and the same lens only the center is in focus (when focusing on the center that is). The edges and corners are out of focus. Resizing them to the resolution of the A7s also confirms it.
2) The sensor flare/reflections are gone.
I'm sorry for slightly going offtopic, but some might be interested.
Tariq Gibran wrote:
Here is another comparison of the Biogon Loxia 35/2 and the ZM 35/2. The ZM diagram was taken from the Zeiss Ikon leaflet PDF.
Just playing complete armchair quarterback but perhaps the compression of the front optical group (the entire thing is compressed it appears) in the Loxia results in greater coverage (expansion of image circle) of the Biogon? Might that push the central sharp area outwards enough but possibly at the expense of a little sharpness?
It would be interesting to hear an experts (theSuede) opinion on the possible effect. Perhaps it's impossible to know with all the other variables ( glass types, etc)?
Yeah, I'd like to see theSuede weigh in on this, as well. I'm not even sure if we can trust these brochure schematics, in terms of exact scale and distances.
Jochenb wrote:
I'm going to quote myself, because I discovered something great.
Last week I added an A7S to my A7. The A7S clearly has a better sensor stack.
1) The extreme field curvature is gone! My "affected" lenses act normal again, just like they did on a DSLR. It's not that I get fooled by the lower resolution that masks the issue. Planar subjects are tack sharp across the frame again. With the A7 and the same lens only the center is in focus (when focusing on the center that is). The edges and corners are out of focus. Resizing them to the resolution of the A7s also confirms it.
2) The sensor flare/reflections are gone.
I'm sorry for slightly going offtopic, but some might be interested. ...Show more →
Wow, that is an interesting point. These Loxia lenses are aimed at both stills and video, so maybe Zeiss used the A7s as their point of reference, and they realized the performance of the ZM 35 and 50 was fine on that camera. If so, it's going to be a compatibility mess.
Jochenb wrote:
I'm going to quote myself, because I discovered something great.
Last week I added an A7S to my A7. The A7S clearly has a better sensor stack.
1) The extreme field curvature is gone! My "affected" lenses act normal again, just like they did on a DSLR. It's not that I get fooled by the lower resolution that masks the issue. Planar subjects are tack sharp across the frame again. With the A7 and the same lens only the center is in focus (when focusing on the center that is). The edges and corners are out of focus. Resizing them to the resolution of the A7s also confirms it.
2) The sensor flare/reflections are gone.
I'm sorry for slightly going offtopic, but some might be interested. ...Show more →
Nothing is off topic here, the Lokia has a lot to prove on a camera many are having difficulties with adapting. As far as your observation about the 7s, very interesting. Somehow I find it hard to believe Sony changed the stack, I just think those big fat pixels are more forgiving.
Jochenb wrote:
I'm going to quote myself, because I discovered something great.
Last week I added an A7S to my A7. The A7S clearly has a better sensor stack.
1) The extreme field curvature is gone! My "affected" lenses act normal again, just like they did on a DSLR. It's not that I get fooled by the lower resolution that masks the issue. Planar subjects are tack sharp across the frame again. With the A7 and the same lens only the center is in focus (when focusing on the center that is). The edges and corners are out of focus. Resizing them to the resolution of the A7s also confirms it.
2) The sensor flare/reflections are gone.
I'm sorry for slightly going offtopic, but some might be interested. ...Show more →
This is fantastic news to me, Jochen! Thanks for sharing! It means that Sony finally realized their mistake and also means the thick cover was not really intentional. I would imagine like Douglas that Zeiss designed the new lenses for the new sensor stack. It would also mean that future models will have a similar stack. The bad news is maybe for A7/R owners but let's wait and see.
mogul wrote:
Nothing is off topic here, the Lokia has a lot to prove on a camera many are having difficulties with adapting. As far as your observation about the 7s, very interesting. Somehow I find it hard to believe Sony changed the stack, I just think those big fat pixels are more forgiving.
They clearly changed something about the stack IMHO. Fat pixels are more forgiving, but won't eliminate the horrible sensor reflections of the A7 for example.
edwardkaraa wrote:
This is fantastic news to me, Jochen! Thanks for sharing! It means that Sony finally realized their mistake and also means the thick cover was not really intentional. I would imagine like Douglas that Zeiss designed the new lenses for the new sensor stack. It would also mean that future models will have a similar stack. The bad news is maybe for A7/R owners but let's wait and see.
Yeah, it'll be interesting to see how it goes. It's so difficult to predict, because we're talking about a combination of sensor stack thickness, the distance from the sensor, the types of materials used in the stack, the micro lenses, pixels, etc. Who knows? The thing that is a concern is that we don't seem to get anything consistent in that regard from Sony.
douglasf13 wrote:
Wow, that is an interesting point. These Loxia lenses are aimed at both stills and video, so maybe Zeiss used the A7s as their point of reference, and they realized the performance of the ZM 35 and 50 was fine on that camera. If so, it's going to be a compatibility mess.
It's also worth noting that the A7 and A7r do not react the same (A7 more susceptible to sensor reflection) so there is likely something different even between the sensor stack in those cameras.
Jochenb wrote:
Edward, I agree about the sensor stack. However, I don't think the native lenses really suffer from it. I also see the negative effects you describe, but I blame the lens designs. When I put on an adapted lens that I also used on a DSLR I don't really see a difference in the peripheral area. What I do notice is exaggerated field curvature. This is a result of that sensor stack. With the A7 stack you also get sensor reflections as a "bonus".
Jochenb wrote:
I'm going to quote myself, because I discovered something great.
Last week I added an A7S to my A7. The A7S clearly has a better sensor stack.
1) The extreme field curvature is gone! My "affected" lenses act normal again, just like they did on a DSLR. It's not that I get fooled by the lower resolution that masks the issue. Planar subjects are tack sharp across the frame again. With the A7 and the same lens only the center is in focus (when focusing on the center that is). The edges and corners are out of focus. Resizing them to the resolution of the A7s also confirms it.
2) The sensor flare/reflections are gone.
I'm sorry for slightly going offtopic, but some might be interested. ...Show more → douglasf13 wrote:
Wow, that is an interesting point. These Loxia lenses are aimed at both stills and video, so maybe Zeiss used the A7s as their point of reference, and they realized the performance of the ZM 35 and 50 was fine on that camera. If so, it's going to be a compatibility mess.
I'm still kind of in the camp of 'believe it when I see it for myself' if I ever have a chance of trying an a7S... but that said, given the later development time of the a7S, and the immediate realization from early tests (for which I suppose I'm partly responsible) that many lenses will fare poorly, that Sony took note of the intense, if niche, interest in adapting small, rangefinder lenses. It could be conceivable that they'll do something with a future 'a7 Mark II' to improve its compatibility and maybe we're seeing that with the a7S?
I would expect this to be an 'upgrade feature/benefit' to entice some a7/a7R owners to the new camera. I don't know about others, but the way these cameras come across, they seem virtually disposable. Use them for immediate needs, move on when the next best thing comes along.
Tariq Gibran wrote:
Here is another comparison of the Biogon Loxia 35/2 and the ZM 35/2. The ZM diagram was taken from the Zeiss Ikon leaflet PDF.
Just playing complete armchair quarterback but perhaps the compression of the front optical group (the entire thing is compressed it appears) in the Loxia results in greater coverage (expansion of image circle) of the Biogon? Might that push the central sharp area outwards enough but possibly at the expense of a little sharpness?
My feeling is we'll get the only definitive 3rd party analysis when Roger at Lensrentals compares the Loxia line against its ZM equivalents. Hopefully he'll do this because it seems to be too good an opportunity to pass up.
I forgot to mention that I'm not using adapted rangefinder lenses. Only Zeiss ZF and native lenses. The 35/2.8 FE also shows quite strong field curvature on the A7. Not anymore on the A7s, so that might be good news for rangefinder lenses too.
I understand that people will find it hard to believe, so all I can say is: try the A7s.
Jochenb wrote:
I forgot to mention that I'm not using adapted rangefinder lenses. Only Zeiss ZF and native lenses. The 35/2.8 FE also shows quite strong field curvature on the A7. Not anymore on the A7s, so that might be good news for rangefinder lenses too.
I understand that people will find it hard to believe, so all I can say is: try the A7s.
I don't find it hard to believe, I just think your assumption about a change in the sensor stack is off. Remember, the NEX 5n/r/t and 6 performed better with a much wider variety of lenses than the 7, and the only difference was the resolution.
Jochenb wrote:
I forgot to mention that I'm not using adapted rangefinder lenses. Only Zeiss ZF and native lenses. The 35/2.8 FE also shows quite strong field curvature on the A7. Not anymore on the A7s, so that might be good news for rangefinder lenses too.
I understand that people will find it hard to believe, so all I can say is: try the A7s.
Have used both the A7s and A7r, I would say my ZM 15 didn't see any improvement in term of smearing due to field curvature. The heavy vignette and color cast, on the hand, are completely gone. I think only the A7r and A7s have gapless microlens, so their stack might be more similar (than the A7).
Regarding the NEX cameras: that pertained mostly to color shift, which many didn't initially differentiate from edge smearing. Color shift is very much affected by pixel density. Comparisons of the NEX vs. the GXR revealed that the GXR had noticeably better edge detail rendering, likely due to the lack of AA filter and therefore thinner sensor stack.
Anyway, I also wasn't that impressed with the FE35/2.8 on the a7 and a7R for infinity type scenes, finding edge sharpness to be relatively poor. Interesting to hear it's noticeably better with the a7S. We'll see if it translates to rangefinder lenses, which has been the observation of some, such as Jim Kasson.
ken.vs.ryu wrote:
How can the sensor cover thickness be different if the focus point stays the same for all a7 cameras?
I'd guess the sensor position is optimized in each camera model to take this into account. I believe it was a reason why the Nikon D800E used the self-cancelling AA filter stack rather than a true non-AA design because it simplified camera production. Both sensors had the same thickness and therefore the E didn't require specific optimization for focus, making it more economical to produce.
freaklikeme wrote:
I don't find it hard to believe, I just think your assumption about a change in the sensor stack is off. Remember, the NEX 5n/r/t and 6 performed better with a much wider variety of lenses than the 7, and the only difference was the resolution.
---------------------------------------------
A change in resolution doesn't eliminate sensor reflections. Since both the A7 and A7s have an AA filter there must have changed something else about the stack.
hiepphotog wrote:
Have used both the A7s and A7r, I would say my ZM 15 didn't see any improvement in term of smearing due to field curvature. The heavy vignette and color cast, on the hand, are completely gone. I think only the A7r and A7s have gapless microlens, so their stack might be more similar (than the A7).
The ZM15 is ofcourse an extreme example. That's why I added that I haven't tried rangefinder lenses. Btw, doesn't the smearing with that lens come from light that hits the sensor at an angle that's too steep? If It's field curvature... the corners should get completely sharp if you focus on them.
I'm going to say it again.. I'm fairly certain that Zeiss is not designing a lens based on a sensor that'll be updated 6-12 months from now.
1) They will not design a lens for one camera, to have it not jive well with the next generation of sensor.
2) Sony has probably already given Zeiss a completely different sensor to work with, something we're not privy to, if in fact Zeiss even needs said sensor to design the lens.
3) Zeiss is probably optimising their lenses for a set value, that may result in better images on one camera then another, but won't be optimised for a specific camera that's currently on the market.
If this were another RX1, I could see them making a lens based on camera (like the rumoured curved sensor fixed lens camera). This isn't that, this is a lens that's being released for A7 mount, but would/could come out for Fuji or Nikon, or anyone else wanting to jump into the FF mirror less game. It's not a Sony brand Zeiss lens.. It's a Zeiss lens that happens to have a Sony mount.
freaklikeme wrote:
I don't find it hard to believe, I just think your assumption about a change in the sensor stack is off. Remember, the NEX 5n/r/t and 6 performed better with a much wider variety of lenses than the 7, and the only difference was the resolution.
I saw a teardown of the Nex-5n and Nex-7 sensor, and I believe that they did have different sensor stacks. Maybe it was Vivek over on getdpi that showed this? Having a hard time remembering.
Jochenb wrote:
The ZM15 is ofcourse an extreme example. That's why I added that I haven't tried rangefinder lenses. Btw, doesn't the smearing with that lens come from light that hits the sensor at an angle that's too steep? If It's field curvature... the corners should get completely sharp of your focus on them.
Smearing would be caused by refraction of the light by the sensor stack in a manner not accounted for in the lens's design. Based on a Zeiss white paper, the refraction caused by the sensor stack primarily affects tangential MTF values, causing field curvature in that direction, while not affecting the sagittal direction nearly as much.
As you can see from the ZM15's MTF, sagittal (solid) lines and tangential lines already diverge quite a bit. With a thick sensor stack this will get worse and the greater field curvature in the tangential direction can never be focused at the same point as on in the sagittal direction, resulting in bad smearing.
with all this talk about short flange distance being a potential obstacle to high IQ (especially in the corners of wide angle lenses) -
I'm starting to hope there is a high-end A-mount announced soon without any mirror that gives us the best of the A7 without the problems. Prior to testing out the A7r, my real hope was for Nikon to release a D800E without a mirror. I've lost hope that Nikon will ever do that, even though it would be great for their sales.
My primary interest in mirrorless is hybrid AF (on sensor PDAF to quickly zero in on the subject and contrast detect to finish it perfectly). This also gets around focus shift issues. And an EVF will help me finally get a MF lens.
Frankly, all this talk about the Loxia 35 biogon and potential trade offs, etc - has me wishing for my Sigma 35/1.4 again. I know some think it is a bit harsh and lacks the soul that Zeiss has, but I thought it was perfectly clean and easy to use.
It strikes me as more than a bit odd that neither Sony nor Zeiss seem forthcoming about anything wider than a 35mm lens and I'm starting to get concerned that there could be issues in developing a reasonably priced high IQ WA for the FE-mount. Of course, these concerns could be put to rest at any moment - but given the interest and concern, I'd think Sony would want to calm those concerns sooner than later.
That's why I added that I haven't tried rangefinder lenses. Btw, doesn't the smearing with that lens come from light that hits the sensor at an angle that's too steep? If It's field curvature... the corners should get completely sharp if you focus on them.