p.3 #1 · Zeiss Loxia 50mm f:2.0 Vs Sony-Zeiss FE 55 f:1.8
Tariq Gibran wrote:
Given that Ron Scheffler tested the ZM version of this lens which performed similarly between the A7r and Leica M9 (he says edges slightly better on M9 until F4), I wonder if there is any optical advantage with the Loxia version of this lens? A side by side with the two - ZM vs Loxia Planar 50 - would be interesting.
The MTF's are fairly similar except that the stopped down MTF for the Loxia is at 5.6 and for the ZM at F4. The Loxia at 5.6 appears to show better corner detail (presumably since we don't get MTF at 5.6 with the ZM) but also shows a zone B dip that is not there in the ZM.
In my own tests with the A7, I found a substantial difference. I might have a better copy of the planar, as it performs extremely well on both M9 and M(240) across the frame. The A7 was at least 3 stops behind in the borders/corners and never catches up even stopped down.
p.3 #2 · Zeiss Loxia 50mm f:2.0 Vs Sony-Zeiss FE 55 f:1.8
edwardkaraa wrote:
In my own tests with the A7, I found a substantial difference. I might have a better copy of the planar, as it performs extremely well on both M9 and M(240) across the frame. The A7 was at least 3 stops behind in the borders/corners and never catches up even stopped down.
Interesting. Did you take a look at Ron's test on the A7r? The A7r result is downsized to match the M9 but even so, F4 looks great at the edges and F5.6 even better. F5.6 appears to actually be optimum. I don't see how the M9 or M240 could improve on that result. Before F4 though, I could imagine the Leica being better (as Ron also states). Maybe the A7r has an advantage over the A7, I don't know?
p.3 #3 · Zeiss Loxia 50mm f:2.0 Vs Sony-Zeiss FE 55 f:1.8
Tariq Gibran wrote:
Interesting. Did you take a look at Ron's test on the A7r? The A7r result is downsized to match the M9 but even so, F4 looks great at the edges and F5.6 even better. F5.6 appears to actually be optimum. I don't see how the M9 or M240 could improve on that result. Before F4 though, I could imagine the Leica being better (as Ron also states). Maybe the A7r has an advantage over the A7, I don't know?
Sorry Tariq, I wrote my post based on memory, without looking again at the test. I would say the M9 result is quite good but I think my lens performs a bit better. But the surprise is the A7r, I think it is much better than the A7 with this lens, because my tests showed very weak corners and borders up to f/4 and acceptable from f/5.6 upwards.
p.3 #4 · Zeiss Loxia 50mm f:2.0 Vs Sony-Zeiss FE 55 f:1.8
wfrank wrote:
The miraculous small Distagons seem unfortunately to be a no-go on A7 current generation. But who knows what the future holds with regards to cover glass thickness at al. Time for a coffee on Söder Kosmos?, long time now :-)
Yup, it's about time. I am free next week from tuesday to friday so text me about a suitable day and time. I have a few errands to run on Söder and plenty of time for a fika.
p.3 #6 · Zeiss Loxia 50mm f:2.0 Vs Sony-Zeiss FE 55 f:1.8
Jochen, fact is, Loxia 50 is a "low-premium-priced" Planar. No way are you going to get a super-smooth bokeh. Does it get better? Yes. Great? No.
Oct 06, 2014 at 03:24 AM
Guest
Guest
p.3 #7 · Zeiss Loxia 50mm f:2.0 Vs Sony-Zeiss FE 55 f:1.8
wfrank wrote:
The miraculous small Distagons seem unfortunately to be a no-go on A7 current generation. But who knows what the future holds with regards to cover glass thickness at al. Time for a coffee on Söder Kosmos?, long time now :-)
I highly doubt that Sony will reduce the sensor stack thickness because it'll affect wide-open performance of native lenses designed for thick glass filter.
p.3 #8 · Zeiss Loxia 50mm f:2.0 Vs Sony-Zeiss FE 55 f:1.8
maratus wrote:
I highly doubt that Sony will reduce the sensor stack thickness because it'll affect wide-open performance of native lenses designed for thick glass filter.
I highly doubt that Sony, or any other manufacturer, have ever designed a lens with the thickness of the sensor stack in mind.
I understand that lens manufacturers design lenses for digital by playing on the light rays hitting the corners, making them less oblique and more perpendicular.
But for them to induce over correction of optical aberrations so that they compensate for the stack thickness is way far fetched imho. Canon for instance has DSLR sensor stack between 1 to 3 mm. For which thickness did they design the lenses?
The advantage of the 1st option, is that it works equally well on thin and thick glass covers.
Oct 06, 2014 at 04:23 AM
Guest
Guest
p.3 #9 · Zeiss Loxia 50mm f:2.0 Vs Sony-Zeiss FE 55 f:1.8
edwardkaraa wrote:
I highly doubt that Sony, or any other manufacturer, have ever designed a lens with the thickness of the sensor stack in mind.
I understand that lens manufacturers design lenses for digital by playing on the light rays hitting the corners, making them less oblique and more perpendicular.
But for them to induce over correction of optical aberrations so that they compensate for the stack thickness is way far fetched imho. Canon for instance has DSLR sensor stack between 1 to 3 mm. For which thickness did they design the lenses?
The advantage of the 1st option, is that it works equally well on thin and thick glass covers....Show more →
This test proves how Otus 55mm f1.4 with Booster (effectively 39mm f1.0) MTF degrades without optical stack due to spherical aberrations.
But I think you're right because even 55/1.8 isn't fast enough to show the difference. But on the other hand, 3mm in A7r is a lot even compared to modern DSLRs, let alone Leica (1mm) and film (0mm). Therefore an optimisation for FE lenses should've been made. I wouldn't be surprised if FE 55m 1.8 isn't as sharp wide open on M240 sensor!
p.3 #10 · Zeiss Loxia 50mm f:2.0 Vs Sony-Zeiss FE 55 f:1.8
maratus wrote:
This test proves how Otus 55mm f1.4 with Booster (effectively 39mm f1.0) MTF degrades without optical stack due to spherical aberrations.
But I think you're right because even 55/1.8 isn't fast enough to show the difference. But on the other hand, 3mm in A7r is a lot even compared to modern DSLRs, let alone Leica (1mm) and film (0mm). Therefore an optimisation for FE lenses should've been made. I wouldn't be surprised if FE 55m 1.8 isn't as sharp wide open on M240 sensor!
I have seen this before but I don't really trust the shown results, because a huge DSLR lens like the Otus must have a very distant exit pupil and sensor glass shouldn't affect it in any significant way.
p.3 #11 · Zeiss Loxia 50mm f:2.0 Vs Sony-Zeiss FE 55 f:1.8
Here are a couple of shots which will give you more examples of the bokeh. Please don't give me a hard time over their total lack of artistic value. From memory, the bar shot is f:2.8, and the scooter shot is defintiely wide open, with focus on the speedometer.
p.3 #12 · Zeiss Loxia 50mm f:2.0 Vs Sony-Zeiss FE 55 f:1.8
edwardkaraa wrote:
I highly doubt that Sony, or any other manufacturer, have ever designed a lens with the thickness of the sensor stack in mind.
I highly doubt Sony or anyone else completely omit the filter stack from their lens design simulations. I read an interviewwith Brian Caldwell where he revealed the speedbooster was designed with the filter stack in mind.
edwardkaraa wrote:
for them to induce over correction of optical aberrations so that they compensate for the stack thickness is way far fetched imho.
I agree it sounds far fetched, but that's exactly what Zeiss themselves claim to have done with the Loxia line:
edwardkaraa wrote:
Canon for instance has DSLR sensor stack between 1 to 3 mm. For which thickness did they design the lenses?
Optical thickness is different to physical thickness, but even if the optical thickness was between 1 and 3mm, I would imagine they would design the lenses for 2mm for best average case performance.
edwardkaraa wrote:
a huge DSLR lens like the Otus must have a very distant exit pupil and sensor glass shouldn't affect it in any significant way.
There's nothing that says every (D)SLR lens has to have a very long exit pupil, the CY 45mm f/2.8 has a shorter exit pupil than the rangefinder Voigtlander 35mm f/1.2, many SLR lenses have exit pupils of around 50mm which is well inside the danger zone. Both the Samyang and Sigma 85mm lenses are only 62mm. Also it's not only entrance pupil distance that matters, the diameter is important too, which means wide aperture lenses like the Otus will suffer more.
p.3 #13 · Zeiss Loxia 50mm f:2.0 Vs Sony-Zeiss FE 55 f:1.8
Ok, but the speedbooster isn't a lens. And if Sony really did design its lenses for the filter stack, they must have done a very poor job as their performance in zone C is really bad. Even the Loxia are very incremental improvement over the ZM on a thick glass cover. You would want to believe Zeiss marketing material, but I am skeptical. As for your last point, exit pupil distance is what determines the angle of incidence on the sensor.
p.3 #14 · Zeiss Loxia 50mm f:2.0 Vs Sony-Zeiss FE 55 f:1.8
philber wrote:
Jochen, fact is, Loxia 50 is a "low-premium-priced" Planar. No way are you going to get a super-smooth bokeh. Does it get better? Yes. Great? No.
Ok thanks Philippe. I still have the 1.4 planar ZF.2 and the bokeh of that one gets really good at F2.8. Never any harshness or double edges.
p.3 #15 · Zeiss Loxia 50mm f:2.0 Vs Sony-Zeiss FE 55 f:1.8
edwardkaraa wrote:
Ok, but the speedbooster isn't a lens. And if Sony really did design its lenses for the filter stack, they must have done a very poor job as their performance in zone C is really bad.
Which Sony lenses are you talking about here? Many lenses are soft in the corners for a variety of reasons, not limited to filter stack astigmatism.
edwardkaraa wrote:
As for your last point, exit pupil distance is what determines the angle of incidence on the sensor.
Do you mean exit pupil diameter has no influence? I don't understand how that can be the case.
p.3 #16 · Zeiss Loxia 50mm f:2.0 Vs Sony-Zeiss FE 55 f:1.8
That argument of the lens design taking into account the stack thickness reminds me of those curved field slide projector lenses ( i.e. the Colorplan CF series) which, like the name implies were designed to give a curved focus zone, thus compensating for the curvature that the heat of the lamp causes on non glass mounted slides.
Obvious the slide's curvature is never the same because it depends on many factors ( time of exposure to the lamp's light, room temperature, efficiency of the cooling system, different emulsion bases ). Evidently CF lenses were designed with an average curve in mind and in practical terms I can say, by own experience, that they are useless. The best way to project a slide is a perfectly corrected lens and a flat slide ( glass mount or a very efficient cooling system)
Same with lenses and sensor plates. Unless you design a different lens version for each different sensor stack plate (which is absurd ), if you want to avoid any trouble, take the filter stack away. That means of course a new sensor technology that doesn't needs any filter at all. We're not there yet and I suspect that the manufacturers aren't interested in the least on this for commercial reasons. I also agree with Edward on his scepticism about the marketing literature. In that aspect Zeiss is no different from Sony, Canon or Fuji. Why should they be ?.
p.3 #17 · Zeiss Loxia 50mm f:2.0 Vs Sony-Zeiss FE 55 f:1.8
artur5 wrote:
I also agree with Edward on his scepticism about the marketing literature. In that aspect Zeiss is no different from Sony, Canon or Fuji. Why should they be ?.
Zeiss have left the door open with the specific wording of the above marketing material as well since it states "...the lens design can take into account the influence of the plate...". Note it does not definitively state Loxia has actually done so.
p.3 #18 · Zeiss Loxia 50mm f:2.0 Vs Sony-Zeiss FE 55 f:1.8
Tariq Gibran wrote:
Zeiss have left the door open with the specific wording of the above marketing material as well since it states "...the lens design can take into account the influence of the plate...". Note it does not definitively state Loxia has actually done so.
It's possible but that would be an incredibly misleading, cynical move. From the article:
I was lucky and I had a chance to talk to Mr. Hubert Nasse – Staff scientist for Zeiss Laboratory/Application Technology.
...He led me with big patience through few more charts – showing how Loxia lenses were corrected for the difference in thickness of protective sensor glass, and how significantly this correction affected image quality.
That could, of course all be lies, but isn't it possible that Zeiss did apply correction for the filter stack (splitting the difference between different thicknesses). It seems this would be easier than taking an existing lens design and making it telecentric somehow!
p.3 #19 · Zeiss Loxia 50mm f:2.0 Vs Sony-Zeiss FE 55 f:1.8
Matt Grum wrote:
It's possible but that would be an incredibly misleading, cynical move. From the article:
That could, of course all be lies, but isn't it possible that Zeiss did apply correction for the filter stack (splitting the difference between different thicknesses). It seems this would be easier than taking an existing lens design and making it telecentric somehow!
Yeah, I don't know. I was just pointing out the potential due to the vague and non-affirmative wording (and knowing a bit about marketing). Because of that - and particularly if this chart was the one used - it would be easy for someone in a discussion to interpret things a certain way (jump to conclusions) without there necessarily being an all out lie perpetrated. But then, we do know the Loxia Biogon does perform better so...
p.3 #20 · Zeiss Loxia 50mm f:2.0 Vs Sony-Zeiss FE 55 f:1.8
edwardkaraa wrote:
exit pupil distance is what determines the angle of incidence on the sensor.
Afaik the size and form of the last glass element influences the angle much more than the small differences in the exit pupil distance. Forexample in Zony 55/1.8 the last lens element if very close to the sensor (closer than some poor perfoming rf lenses) but it's almost as large as the front element and fairly flat from the backside. I imagine this kind of lens design requires much less optimizing for the thick sensor stack.
