p.12 #1 · A thinner sensor stack may be possible after all!
edwardkaraa wrote:
It is no secret that the 35 biogon has a resolution drop in the extreme corners, but really the affected area is so tiny I would consider it negligible. I seem to have an excellent copy of the biogon, and it has even been finetuned at oberkochen, but I have to say, like with all traditional designs, it is very important at what distance to focus in order to maximize sharpness across the frame.
The MTF for the ZM from Zeiss only shows us the behavior stopped down to F4 where the corners even there look fairly respectable (they certainly don't take a dive into severe blurryland). I would have expected them to improve at 5.6 and certainly look great by F8. Frankly, in this instance, I don't think the poor extreme corners are due to the lens itself but rather the interaction with the lens and the M9 sensor (since those MTF's are based on film, not digital).
p.12 #2 · A thinner sensor stack may be possible after all!
If the 25 Biogon samples look good, I'm sending my camera in for the conversion. Question for those of you who use AF lenses, especially the 55 FE, is there any deleterious effect in terms of AF performance? I'm going to assume that IQ (hopefully) is not diminished as a result of the mod.
p.12 #3 · A thinner sensor stack may be possible after all!
Tariq Gibran wrote:
The MTF for the ZM from Zeiss only shows us the behavior stopped down to F4 where the corners even there look fairly respectable (they certainly don't take a dive into severe blurryland). I would have expected them to improve at 5.6 and certainly look great by F8. Frankly, in this instance, I don't think the poor extreme corners are due to the lens itself but rather the interaction with the lens and the M9 sensor (since those MTF's are based on film, not digital).
As far as I understand, the MTF bench is purely optical and does not use any recording medium, film or digital, but I could be wrong on that. Anyhow, my copy of the biogon behaves exactly as the MTF let you believe, and that is on both M9 and M240. The somewhat weak corners are there at wide apertures but go away by f/4. I don't exclude sample variations, or simply different RF calibrations.
p.12 #4 · A thinner sensor stack may be possible after all!
edwardkaraa wrote:
As far as I understand, the MTF bench is purely optical and does not use any recording medium, film or digital, but I could be wrong on that. Anyhow, my copy of the biogon behaves exactly as the MTF let you believe, and that is on both M9 and M240. The somewhat weak corners are there at wide apertures but go away by f/4. I don't exclude sample variations, or simply different RF calibrations.
The MTF's that Zeiss have given us most recently for the lenses that are specifically designed for digital (the Loxia's and the more recent ZM 35/1.4) are measured with the respective cover glass thickness taken into account (perhaps they place the glass in the optical path). I have not see them but apparently Zeiss even supplied Lloyd with two different MTF's for that ZM 35/1.4 - one for the Leica M digitals and one which showed how the lens performed on the Sony A7's. The older ZM's that predate the Leica digital M's were originally designed for the Zeiss Ikon (and film) and do not take into account any cover glass thickness. These MTF measurements will not always apply (particularly for the wides of course) for how a specific lens will perform on digital sensors. That's my understanding anyway.
p.12 #5 · A thinner sensor stack may be possible after all!
edwardkaraa wrote:
It is no secret that the 35 biogon has a resolution drop in the extreme corners, but really the affected area is so tiny I would consider it negligible. I seem to have an excellent copy of the biogon, and it has even been finetuned at oberkochen, but I have to say, like with all traditional designs, it is very important at what distance to focus in order to maximize sharpness across the frame.
I agree Edward. I was surprised at my sample above when I first saw it, but the edges are very good. Overall sharpness is wonderfully even as the graphs would show.
I think Ron's analysis is a good one. The 35 cron will give some mid frame dips in return for the corners.
uscmatt99 wrote:
If the 25 Biogon samples look good, I'm sending my camera in for the conversion. Question for those of you who use AF lenses, especially the 55 FE, is there any deleterious effect in terms of AF performance? I'm going to assume that IQ (hopefully) is not diminished as a result of the mod.
My AF tests the camera seemed normal (with LAEA4). I asked about this, and they seemed to believe it would not hurt AF. The FE 35 and 55 will loose a tad of performance, but the Loxia 35/2 should be better. Pure speculation on my part. I would guess all the Loxia wides would be better.
p.12 #6 · A thinner sensor stack may be possible after all!
uhoh7 wrote:
I agree Edward. I was surprised at my sample above when I first saw it, but the edges are very good. Overall sharpness is wonderfully even as the graphs would show.
I think Ron's analysis is a good one. The 35 cron will give some mid frame dips in return for the corners.
My AF tests the camera seemed normal (with LAEA4). I asked about this, and they seemed to believe it would not hurt AF. The FE 35 and 55 will loose a tad of performance, but the Loxia 35/2 should be better. Pure speculation on my part....Show more →
Respectfully, that's a Meatball pitch. given the close distance. The shot you previously posted at infinity tells the tale of how the lens performs on the M9 at distance into the corners (unless there is something specifically going on with your copy or rangefinder)- and the MTF's are based on infinity performance as far as I'm aware.
p.12 #7 · A thinner sensor stack may be possible after all!
Thanks Charlie regarding the FE lenses. The 55 is so good that a little performance loss isn't a big deal as a trade-off for useable 25mm and 35mm rangefinder lenses. I was crestfallen when I saw what happened to my Biogons.
By the way, great work you've presented on the forum. I'm sure there are many other lurkers out there like me who check into the Leica and Sony image threads daily!
Feb 06, 2015 at 01:07 PM
Steve Spencer Offline • • • • • • • Upload & Sell: On
p.12 #8 · A thinner sensor stack may be possible after all!
uhoh7 wrote:
I agree Edward. I was surprised at my sample above when I first saw it, but the edges are very good. Overall sharpness is wonderfully even as the graphs would show.
I think Ron's analysis is a good one. The 35 cron will give some mid frame dips in return for the corners.
My AF tests the camera seemed normal (with LAEA4). I asked about this, and they seemed to believe it would not hurt AF. The FE 35 and 55 will loose a tad of performance, but the Loxia 35/2 should be better. Pure speculation on my part. I would guess all the Loxia wides would be better....Show more →
I doubt very much the Loxia will be better with the thinner cover glass--they were designed for the thicker cover glass. Much like the FE lenses, I would expect the Loxia will lose a tad of performance as well. We will need tests at infinity with all the FE mount lenses to know for sure, though.
p.12 #9 · A thinner sensor stack may be possible after all!
Steve Spencer wrote:
I doubt very much the Loxia will be better with the thinner cover glass--they were designed for the thicker cover glass. Much like the FE lenses, I would expect the Loxia will lose a tad of performance as well. We will need tests at infinity with all the FE mount lenses to know for sure, though.
I've read just the opposite, the reason why the FE35/2.8 basically outperforms the Loxia on the A7 is that perhaps Zeiss was not ready to fully commit to such extremes. My reasoning is: they went a aways in the thick CG direction, away from the film plane, and that just might put those lenses right in the sweet spot for the Kolari thickness.
For landscape photography I don't mind trading 2/35 ZM Biogon's very slight softening of extreme corners and edges for its uniform performance and low distortions across 90% of the frame.
Back to Kolari mod; Biogon 2/35 ZM can now produce easily satisfying results on A7R
Steve Spencer Offline • • • • • • • Upload & Sell: On
p.12 #11 · A thinner sensor stack may be possible after all!
uhoh7 wrote:
I've read just the opposite, the reason why the FE35/2.8 basically outperforms the Loxia on the A7 is that perhaps Zeiss was not ready to fully commit to such extremes. My reasoning is: they went a aways in the thick CG direction, away from the film plane, and that just might put those lenses right in the sweet spot for the Kolari thickness.
But I suspect we will see sometime soon.
Where have you read that? Everything that I have seen from Zeiss suggests that they were designed specifically for the cover glass of the A7 series cameras.
p.12 #12 · A thinner sensor stack may be possible after all!
Great stuff Michael, Charlie and Jack
I'm still awaiting for my A7rM to arrive back in Australia. Everything I have seen with different M lenses, the modification is excellent. I'm thankful to Michael and Kolarivision for pursuing this mod. I had approached several labs in Australia not long after the A7r came out, but the cost for small runs was too high here.
I really don't feel there will be a downside other that the possibility of increased dust and maybe in 18 months to replace the "thin filter" with an upgraded version. At least the modification can be done with a week. Compare this to the sensor replacement with the M9 taking an unacceptable time frame IMO. Though kudos to Leica for honouring their commitment to the M9.
Regardless of the Sony's about to be announced, having a "thin filter" mod A7r will be great to have for the many problematic M lenses. I suspect the new ZM 35/1.4 will also perform much better with the A7rM.
p.12 #13 · A thinner sensor stack may be possible after all!
Charles, I think the new ZM will look great on the modded camera. The big question IMO is whether Kolarivision will be able to apply these mods to the IBIS-capable sensors, which likely will be inevitable in many/most future releases.
p.12 #14 · A thinner sensor stack may be possible after all!
Ron. I suspect if the modification becomes a lot more popular there will be ways to adapt to the IBIS sensors. It still early in the evolution, and I am sure there will be other labs that will also offer this service.
p.12 #15 · A thinner sensor stack may be possible after all!
Steve Spencer wrote:
Where have you read that? Everything that I have seen from Zeiss suggests that they were designed specifically for the cover glass of the A7 series cameras.
I don't think there is any evidence, but I do believe the same. The performance in the corners of the FE35 and especially the FE24-70 is just not good enough for Zeiss, and I suspect that they knew that Sony had used a cheap, thick sensor glass covering in their initial models, to keep prices down and increase the chance of success for the system as a whole. I think that they then designed lenses which performed acceptably on such glass, but which will shine, or at least do much better, when Sony releases higher-speced models.
Feb 07, 2015 at 05:15 AM
Steve Spencer Offline • • • • • • • Upload & Sell: On
p.12 #16 · A thinner sensor stack may be possible after all!
carstenw wrote:
I don't think there is any evidence, but I do believe the same. The performance in the corners of the FE35 and especially the FE24-70 is just not good enough for Zeiss, and I suspect that they knew that Sony had used a cheap, thick sensor glass covering in their initial models, to keep prices down and increase the chance of success for the system as a whole. I think that they then designed lenses which performed acceptably on such glass, but which will shine, or at least do much better, when Sony releases higher-speced models.
But the FE 35 and FE 24-70 are Sony/Zeiss likely designed by Sony. And if you look at the ZM 35 f/2 on the modified cameras it looks a lot like the Loxia 35 on the unmodified cameras, so I don't think the Loxia lenses will do better on the modified cameras and I certainly don't think the Sony/Zeiss FE lenses will do better and there is a little evidence there. In Roger's imatest results, which are close up the FE lenses did just a tad worse and this deterioration should get a bit worse at infinity.
p.12 #18 · A thinner sensor stack may be possible after all!
Thick cover glass is not inherently bad if the system (lens+stack+sensor) is designed as a system. Look at Micro 4/3 where the standard cover stack is 4mm thick. Yet fine results are achieved by lenses designed for that system. There are many ways to make lenses perform better with a thick cover glass, including minimizing the ray angle and, more importantly, the variation in ray angle seen by the light rays that combine to form an image point (why retrofocal SLR lenses are generally less affected) but also by including the glass in the optical path design. That was the point made in Roger and Brian's experience with the M4/3 speed booster and the Zeiss Otus. Brian designed for the M4/3 stack and performance was compromised by removing it from the path.
As to this modificarion, it will improve the performance of any lens designed for film, period. How much will vary. But it will improve. M lenses. R lenses, etc.
As to whether it will impair properly-designed native lenses for the A7 series - it should. It is changing the optical path away from the design of the system. Assuming the native lens was truly designed with the glass in mind - and I have to think the FE and (likely) Loxia lenses are. Expecting it to improve them is assuming they were not designed for the A7 series in the first place. Unlikely.
I will be interested to see how the ZM 35mm 1.4 performs on this. I suspect it was designed for the Leica cover stack, and we are really close to that here. So it could be a real winner, and this also might explain the excellent performance with some of the more-modern Leica glass. But, again, all glass designed for film (no cover stack) will improve. This is what we are seeing and the only question is 'how much' on any particular lens.
p.12 #19 · A thinner sensor stack may be possible after all!
Steve Spencer wrote:
But the FE 35 and FE 24-70 are Sony/Zeiss likely designed by Sony. And if you look at the ZM 35 f/2 on the modified cameras it looks a lot like the Loxia 35 on the unmodified cameras, so I don't think the Loxia lenses will do better on the modified cameras and I certainly don't think the Sony/Zeiss FE lenses will do better and there is a little evidence there. In Roger's imatest results, which are close up the FE lenses did just a tad worse and this deterioration should get a bit worse at infinity.
Of course the ZM 35/2 is designed for film and the best you will see from it digitally is on the M9, not A7.mod. The Loxia 35/2 is certainly not up to those standards on any stock A7 series camera, and most seem to feel it is bested by the FE 35/2.8. I don't think expectations have been met with the 35 loxia for many potential buyers.
Why? How could this be? As Carsten notes, the 2470 has also been a real disappointment, though it's not terrible of course, but not what we'd hoped for and certainly not for the price.
So, Steve you and I can guess about who thought what, and what was designed for what, but those are things we can never know.
However, I will bet you a six pack of nice beer, loser funded via paypal, not to exceed 10USD, that the Loxia 35 is better on the A7.mod or A7r.mod. Also I suspect the 2470 will be better too, but let's stick to the Loxia.
Better will simply mean edges and corners at some apertures. without terrible loss in the center.
I do agree with you and Michael that the FE35 and 55 will be a bit worse. But I agree with Carsten that the Zeiss 2470 and Loxia were not "properly designed" for the thick cover.
Michael, I would add that a thick cover for M43 is no problem because you don't have near the angles going with that tiny sensor. In fact ALT FF glass seems fine on the M43, abeit, cropped.
Feb 07, 2015 at 05:57 PM
Steve Spencer Offline • • • • • • • Upload & Sell: On
p.12 #20 · A thinner sensor stack may be possible after all!
uhoh7 wrote:
Of course the ZM 35/2 is designed for film and the best you will see from it digitally is on the M9, not A7.mod. The Loxia 35/2 is certainly not up to those standards on any stock A7 series camera, and most seem to feel it is bested by the FE 35/2.8. I don't think expectations have been met with the 35 loxia for many potential buyers.
Why? How could this be? As Carsten notes, the 2470 has also been a real disappointment, though it's not terrible of course, but not what we'd hoped for and certainly not for the price.
So, Steve you and I can guess about who thought what, and what was designed for what, but those are things we can never know.
However, I will bet you a six pack of nice beer, loser funded via paypal, not to exceed 10USD, that the Loxia 35 is better on the A7.mod or A7r.mod. Also I suspect the 2470 will be better too, but let's stick to the Loxia.
Better will simply mean edges and corners at some apertures. without terrible loss in the center.
I do agree with you and Michael that the FE35 and 55 will be a bit worse. But I agree with Carsten that the Zeiss 2470 and Loxia were not "properly designed" for the thick cover.
Michael, I would add that a thick cover for M43 is no problem because you don't have near the angles going with that tiny sensor. In fact ALT FF glass seems fine on the M43, abeit, cropped.
...Show more →
That's a great bet, but we better make it 15 dollars. We take our beer seriously here in Canada
Of course we don't know for sure how it will turn out, but in my view Zeiss screwed up if the Loxia don't work best with the stock cover glass.
And I never liked FF glass on m4/3rds that much. It just didn't look that great, but many people do.
given the close distance. The shot you previously posted at infinity tells the tale of how the lens performs on the M9 at distance into the corners (unless there is something specifically going on with your copy or rangefinder)- and the MTF's are based on infinity performance as far as I'm aware.