Loxia 2.0/35 vs. FE Sonnar 2.8/35

[thomas]

Though there are already threads about the Loxia 35 this is mostly about the comparison of the Loxia 2.0/35 to the FE Sonnar 2.8/35.
So I thought I open a new thread…

Actually I do like the FE Sonnar 2.8/35 IQ-wise. But I don’t really like handling it manually at all.
I am also finding the light falloff simply much too high.
This is why I was interested in the Loxia 35.

My first impressions in short…:
- the Loxia is so much better to handle manually that any further testing seems to be superfluous. (Still it had to prove it’s also good IQ-wise. And it did…)
- the Loxia shows less light falloff
- one of the main differences related to IQ is the Loxia shows very low distortion whereas the Sonnar has a complex distortion that is pretty heavy at the outer edges of the image frame.
However, if distortion correction is required with the Loxia the lens profile for the ZM 2.0/35 Biogon in Alpa Lens Corrector seems to work pretty good!

So…. some images (actually quite some…).
All shot with an A7R at ISO100. All loaded in Capture One with Standard-settings. Though my Standard is adjusted to set luminance Noise Reduction to zero (color NR to 12) and I’ve quickly adjusted sharpness to 380|0.5|07 for the Loxia … which I’ve also applied to the Sonnar.
I’ve also made LCCs and applied them to correct light falloff in both the lenses 100%.
Other than that no further adjustments in C1.

The light is pretty harsh on the first scene. It is also changing over the course of the test shots… so this will of course have an impact on the comparison.
Also - as always - there might be a slight mismatch in focus between the 2 lenses which may affect a side by side comparison on a pixel level.

These are all PNG-screenshots from the Capture One interface (so not taken from processed files) converted to JPEG. I still think they will do to get a good impression.

1- light falloff at f2.8 - Loxia on the left/Sonnar on the right (on the entire series of this motif).
As you may notice the EXIF of the Loxia shows “50mm” in C1. I’ll investigate why…
I’ve adjusted the exposure of the LCCs so that the brightest values in the very frame center match (sufficiently).
This is without in-camera correction of light falloff (which is applied to the RAW … unlike the other 2 “lens compensations” you can turn on or off in the A7R).

The light falloff of the Sonnar is drastically higher at f2.8:

2- field of view. The Loxia shows a slightly wider field of view:

The following crops are taken from this scene … The building is around 25 meters away.
Make sure to view them at 100% (depending on your browser for instance right-click->show graphic)…


3- Distortion. Upper right corner of the scene, 50% magnification.
As you can see the camera is not perfectly leveled. Doesn’t matter here.
As you can also see the distortion of the Sonnar is pretty heavy in the corners compared to the Loxia:

4- Loxia vs. Sonnar at f2.8, center
maybe there’s a slight advantage for the Sonnar but if so it’s a wash:

5- Loxia vs. Sonnar at f2.8, edge
look pretty much the same to me:

6- Loxia vs. Sonnar at f2.8, corner
clear advantage in favor of the Loxia (mostly due to the distortion and light falloff of the Sonnar):

Let’s stay with the very corner…

7- f4 … Loxia still better:

8- f5.6 … pretty much leveled. But still there is the distortion of the Sonnar… :

9- f8 - same here, though the very corners of the Sonnar are very well usable at f8 even after distortion correction.
So at f8 I think the differences are pretty insignificant:

... tbc. in a minute ...

 

[thomas]

...


Very wide scene (around 250 meters distance to the building):

In the following crops the Loxia is on the right, the Sonnar on the left.

11- Sonnar vs. Loxia at f2.8, center
Again the Sonnar is maybe a tad better. But again the difference is negligible:

12- Sonnar vs. Loxia at f2.8, midfield
The Loxia clearly outperforms the Sonnar by a large margin (it even shows some moiré - at ~250 meters distance at f2.8!):

13- Sonnar vs. Loxia at f2.8, edge
in this comparison the Loxia is ahead at the edge of the frame at f2.8. Again: could just as well be caused by slight mismatch in focusing. Still very close:

Up to here as far as the the comparison of the Loxia vs. the Sonnar goes.


Finally the Loxia wide open, so at f2.0 … crops from the first scene.

14- Loxia at f2.0, center
Given the distance (~25meters) the resolution of the lens leaves nothing to be desired at f2.0.
Really remarkable… IMHO:

15- Loxia at f2.0, edge
the edge is far from unusable either. I would say it looks pretty usable for moderately sized prints.
Then again… who would shoot architecture or landscape wide open when edge to edge (or even corner to corner) sharpness is required?

16- Loxia at f2.0, corner
The very corner of course shows blur wide open. Still not too bad … but certainly not usable for such a motif:

I am really impressed by the Loxia and will definitely keep it. The manual handling is exactly as it should be. Only the aperture ring turns the wrong way around -> http://www.getdpi.com/forum/636135-post87.html
Given its extremely high built quality I also think the price is adequate.
Seems to be a good copy, too. The lens is completely parallel to the sensor plane … there are not the slightest signs of decentering or so. All 4 corners show the same sharpness at the respective apertures.
My Sonnar will go on sale (also a very good copy… my first sample was slightly decentered and showed only 3 sharp corners … this copy is perfectly aligned).

 

[Guy Mancuso]

Thomas awesome test and as predicted from doing my test against the 16-35 it looks like the same results are apparent here as well. My takeaway like you have proved the Loxia across the frame is just about untouchable for a mid wide as most 35mm lenses have field curvature and even if this does it's so very slight . Again same with vignetting and here you prove how much better it is over the Sonner which is known to be heavy. As far as build quality, feel, manual focusing on that alone it's just lovely as can be, can't wait to get the 50 myself. Price maybe 200 more than I think is fair but what else is new. I'm thrilled with mine so far. I wish they had a 21 and 90 today as I would jump all over them.

 

[Guy Mancuso]

As far as size and weight for this camera system which is a big deal to a lot of folks. IMHO it's actually not to small nor to big as it fits the bodies perfectly. This series if they had 4 or 5 of them would rival anything out there for a great travel system

 

[iiiNelson]

Nice tests and it's pretty clear to me that the Loxia is superior to the Sonnar... but I expected that much.

 

[thomas]

most 35mm lenses have field curvature and even if this does it's so very slight .

excellent and very important point - forgot to mention it!!
The Sonnar shows quite some field curvature. This is why it is a bit tricky to focus (you have to focus it "deep" for good edge performance with motifs focused @ infinity ... but of course not too deep... otherwise the image center will go blurry; oh... and forget AF if you are really nitpicking about focusing).
The Loxia shows almost no field curvature! This is why it is much easier to focus (apart from the manual focus ring). It's really focusing the center and forget about the rest (actually as it should be) ...
It still requires careful focusing... especially at wide apertures. It also shows a very mild focus shift... so when focusing wide open and stopping down afterwards you should re-check focusing. But if you intend to shoot at f5.6 or f8 you can very well focus at f2.8 ... the remaining focus shift is really too low to be critical. Focusing with f2.0 and stopping down afterwards is a bit more tricky. But this is of course an unavoidable "property" of fast lenses...

Thank you!

 

[Vivek]

I was under the impression that it is not field curvature but uneven (mustache like) distortions that plague the 35/2.8 Zony lens.

 

[thomas]

As far as size and weight for this camera system which is a big deal to a lot of folks.

Yes, in fact for me!
I always have one or two fast lenses... but mostly (due to the motifs I am shooting) I could very well live with f2.8 or even f3.5 lenses.
I've seen some really good samples of the new Sony 2.0/28. Man, I would pay twice or three the times the price for a well corrected (would love an APO) 28mm with f3.5 or even f4. Really!
Consequently the "Loxia" philosophy is really to my taste and I am definitely going to test the 50mm (though I have to admit that I simply LOVE the ZM 2.0/50 Planar on the A7R - returned 2 copies of the FE 1.8/55 and stayed with the good old Planar... it's so great to handle and damn sharp ... though it has to be stopped down for edge to edge sharpness ... of course).

 

[thomas]

I was under the impression that it is not field curvature but uneven (mustache like) distortions that plague the 35/2.8 Zony lens.

both field curvature and distortion. Personally I am under the impression the field curvature is more critical ...

 

[CharlesK]

Yes, in fact for me!
I always have one or two fast lenses... but mostly (due to the motifs I am shooting) I could very well live with f2.8 or even f3.5 lenses.
I've seen some really good samples of the new Sony 2.0/28. Man, I would pay twice or three the times the price for a well corrected (would love an APO) 28mm with f3.5 or even f4. Really!
Consequently the "Loxia" philosophy is really to my taste and I am definitely going to test the 50mm (though I have to admit that I simply LOVE the ZM 2.0/50 Planar on the A7R - returned 2 copies of the FE 1.8/55 and stayed with the good old Planar... it's so great to handle and damn sharp ... though it has to be stopped down for edge to edge sharpness ... of course).

Thank you Thomas for the excellent test comparison results. There is much discussion as to which is better and why, and it great to see the actual results. I have and love the Loxia 35 and will keep the FE 35/2.8, as my wife loves to use this lens for its size and AF.

The Loxia 50 is a real gem of a lens and truly worthy of having it as part of the Sony lineup. Better still it is about 300 cheaper than the Loxia 35

 

[Jorgen Udvang]

Great test, and while not being an A7 owner, these are just the kind of lenses that would tempt me, making the system look very much like my ancient OM... if they launch a wider and a longer lens. And a 135? I'm sure Zeiss understand that. Hopefully, they sell enough of these lenses to make it worthwhile extending the selection.

 

[ashwinrao1]

Great test, and makes me feel futher validated for trading in my 35 and 55FE's for the 35 and 50 loxias. IQ is great, MF use is perfectly implemented, and build is up to par with high end glass....

Like guy, give me a 21 and a 90, and and a7rII to pair with my a7s...and I think I'd be done ...(cough cough famous last words)

 

[turtle]

There is no such thing as 'THE 35mm FE Sonnar' because of the worst sample variation I have ever encountered. Its more a 'society of related lenses'

Good copies are stunning, but it took me three samples to get one. I suspect this also explains why we get such huge variations in comparisons. In Reid's testing, the Sonnar was markedly superior to the Loxia at the edges and corners the frame at wider apertures and I suspect from the samples you're showing that your copy is a little weak. Looks more glowy at the edges than mine.

Out of the three I had, the first one was slightly softer on the left side, but in overall terms it hit peak at f4 and was only a hair behind at f2.8. The second was very soft on the right side and the third is superb all round... but.... keeps improving to f8, whereas the first copy, superb at f4, actually dropped off at the edges after f4. Honestly, you'd think they were different lenses.

Zeiss optical QC is pretty good from my experience with the ZMs, so I expect the inconsistencies in the 35mm Sonnar vs Loxia to be due to the Sonnars and it is very possible that only a minority of samples actually fulfil the design's potential to the max. If I were buying again now, I would buy the Loxia, simply so I would not have to keep testing and putting lenses back in the mail. Ridiculous. But at least I now have a 110g lens that is very good indeed.

This link may be useful for some of you: What do ‘good copy’ Sony FE Lens files look like?

 

[cam]

You all make me want to loan my lenses out to somebody who can test them properly -- I am just no good at this type of thing! I've honestly no clue as to whether or not I have a "good" copy of the FE 35/2.8, though I was assured by the seller that I do.

The biggest temptation, for me, is the manual focus for a wide angle. Set it at f/8 and forget (or tweak as needed). That's how I use my R-X1 in a lot of instances, hmmmmmmmm...

 

[biglouis]

You all make me want to loan my lenses out to somebody who can test them properly -- I am just no good at this type of thing! I've honestly no clue as to whether or not I have a "good" copy of the FE 35/2.8, though I was assured by the seller that I do.

The biggest temptation, for me, is the manual focus for a wide angle. Set it at f/8 and forget (or tweak as needed). That's how I use my R-X1 in a lot of instances, hmmmmmmmm...

Made me 'lol' because that is pretty much what I do with my lenses of 35mm or wider.

LouisB

 

[Slingers]

What a fantastic test. Thank you.

 

[thomas]

Good copies are stunning, but it took me three samples to get one. I suspect this also explains why we get such huge variations in comparisons. In Reid's testing, the Sonnar was markedly superior to the Loxia at the edges and corners the frame at wider apertures and I suspect from the samples you're showing that your copy is a little weak. Looks more glowy at the edges than mine.

True! Sample variation seems to be a serious issue with the FE Sonnar 2.8/35 (as well with the FE 1.8/55). I have heard of samples that are really good at the edges right from wide open. But these "spectacular" samples seem to be the rare exception. I think my copy is in the good midfield (on par with many image-samples I've seen). The edges are not so great at f2.8 but do get really sharp at f4. Only the very corners stay blurry until f5.6. The alignment of my copy is accurate (parallel focus plane / the 4 corners look the same on fully frontal subjects) which is why I kept it.
Here's a 100% crop of my Sonnar 35 from the above scene at f4 taken from the very edge. I wouldn't say it looks "weak" ...:

Too, light falloff and especially distortion are certainly not sample-depended but "by design" of the lens. In this regard the Loxia 35 clearly outperforms the Sonnar 35... regardless of the "sharpness"-capabilities of the respective samples.

 

[iiiNelson]

There is no such thing as 'THE 35mm FE Sonnar' because of the worst sample variation I have ever encountered. Its more a 'society of related lenses'

Good copies are stunning, but it took me three samples to get one. I suspect this also explains why we get such huge variations in comparisons. In Reid's testing, the Sonnar was markedly superior to the Loxia at the edges and corners the frame at wider apertures and I suspect from the samples you're showing that your copy is a little weak. Looks more glowy at the edges than mine.

Out of the three I had, the first one was slightly softer on the left side, but in overall terms it hit peak at f4 and was only a hair behind at f2.8. The second was very soft on the right side and the third is superb all round... but.... keeps improving to f8, whereas the first copy, superb at f4, actually dropped off at the edges after f4. Honestly, you'd think they were different lenses.

Zeiss optical QC is pretty good from my experience with the ZMs, so I expect the inconsistencies in the 35mm Sonnar vs Loxia to be due to the Sonnars and it is very possible that only a minority of samples actually fulfil the design's potential to the max. If I were buying again now, I would buy the Loxia, simply so I would not have to keep testing and putting lenses back in the mail. Ridiculous. But at least I now have a 110g lens that is very good indeed.

This link may be useful for some of you: What do ‘good copy’ Sony FE Lens files look like?

I think that's the problem. The "good copies" ARE the exception. If only 20-30% of the samples out there are "good copies" then I wouldn't call it a great lens when the vast majority of the samples out there don't perform as expected, reviewed, or advertised for whatever reason. Maybe Sean Reid got a bad copy of the Loxia or maybe he got an exceptional copy of the 35FE Sonnar. I can't say. What I can say was that my copy was a decent copy at least and I really didn't find it all that impressive compared to what I consider great 35mm lenses that I've used in the past to include the Sigma Art, Leica Summicron version 4/5 ASPH, or the CV Nokton II. Just wasn't in the same league as those lenses. Doesn't mean it was horrible and maybe I would've kept it for $500 but not $800.

It's the same issue I had with the 24-70FE it's a good lens at $6-800 but at $1200 not so much. Maybe it's my standards that are skewed but thankfully I can mount whatever on these bodies.

 

[thomas]

The Loxia 50 is a real gem of a lens and truly worthy of having it as part of the Sony lineup.

just ordered one
time to sell some stuff I am definitely not going to use anymore...

 

[philip_pj]

Might have been early manufacture blues. You hear and see many different opinions so a buyer might have to either grit the teeth and try several or stay away from this one, for now at least.

See here for a performance plot of the FE35/2.8:
Sony Lens: Primes - Sony FE 35mm f/2.8 ZA Carl Zeiss Sonnar T* SEL35F28Z (Tested) - SLRgear.com!

- open the a7r blur index chart and run the left slider up and down the apertures; I think it shows exactly what is going on with poor copies. Its not normal for lenses to suffer far worse corners on stop down from wide open! Good on them for publishing that one too.