Sony FE 16-35mm f/4 ZA OSS Vario-Tessar T* vs Sony SEL 10-18mm f/4,0 OSS
We were among first reviewers to post full res samples from Sony FE 16-35mm f/4 ZA OSS Vario-Tessar T* lens announced just before Photokina 2014.
Unfortunately we weren’t able to get testing sample of the lens soon enough after release and thus we lost interest to make a full review, but since we finally got one from the local Sony PR agency (Bison&Rose Prague), we decided to make at least very brief comparison between it and another Sony WA zoom made for APS-C sensor size – Sony 10-18mm f/4,0 OSS SEL which we reviewed some time ago – http://www.verybiglobo.com/sony-sel-1018-f4-oss-sony-nex-7-rolling-review-part-i/
Before FE 16-35 was announced, many users have been relying on its smaller brother – SEL 10-18 even on Sony full frame A7 camera bodies, because it covered larger sensor from 14-18mm or so about, without significant vignetting. That seems to be good reason to compare those two lenses at least in the overlapping range.
Another reason for this comparison is existence of two systems – Sony A7 with FE 16-35 vs Sony A6000 with SEL 10-18, which in theory should give us same resolution at comparable field of view.
In other words – is there a good reason to pay more and carry larger combo (A7 + FE 16-35) when there is smaller and cheaper alternative (A6000 with SEL 10-18) which at least for landscape or travel purpose can serve equally well. Or not?
Because we had FE 16-35 together with other Sony equipment for only 14 days, we didn’t have enough time to think over this comparison well and in reality we got only two and a half days to collect some test data. Anyway, even if this comparable review is going to be bellow Verybiglobo.com standards, it still can give you few answers…
|Lens configuration (group/element)||10 / 12||8 / 10|
|35 mm-equivalent focal length (APS-C) (mm)||24 – 52.5||15 – 27|
|Angle of view (APS-C)||83° – 44°||109° – 76°|
|Angle of view (35mm full frame )||107° – 63°|
|No. of aperture blade||7 (circular aperture)||7 (circular aperture)|
|Min. aperture (F)||22||22|
|Max. magnification ratio (x)||0.19||0.1|
|Min. focus (m)||0.28||0.25|
|Filter dia. (mm)||72||62|
|Hood shape / mount||petal / bayonet||petal / bayonet|
|Dimensions: Dia. x L (mm)||78 x 98.5||70 x 63.5|
|Weight: (approx.) (g)||518||225|
|Provided accessories hood (ALC-SH134), case||hood (ALC-SH134), case||hood (ALC-SH123)|
Comparing both lenses, we can follow similar design pattern – lightweight metal body, precise assembling, modern design. Both lenses are nice instruments in hand and they both communicate premium quality. Being twice as heavy, Sony FE 1635Z feels more reliable for heavy use in comparison to SEL 1018 that is rather fragile, but that probably hasn’t nothing to do with a real life use.
It is interesting to note that SEL 1018 balances nicely on Sony NEX 7 and A6000, while the same can be said about SEL 1635Z and Sony A7/MII/R/S, but not when lenses are used in opposite combination. SEL 1018 on Sony A7 cameras is too light and system is slightly back weighted, while SEL 1635Z is front weighted on APS-C bodies. This makes perfect sense, but it is worth considering, if you are planning to use either lens on its non native sensor size body.
SEL 1018 on FF? Yes, you can use this APS-C lens on Sony A7 cameras with some limitation, because it renders large projection circle, and many Sony A7 early adopters didn’t really have better option for UWA zoom till new FE 1635 wasn’t introduced. However it was always heavily compromised solution, no matter how some company promoters (read pro photographers) tried to convince you otherwise. If you are interested, you can check projection of SEL 1018 on Sony A7 in our old review – http://www.verybiglobo.com/sony-alpha-a7-ilce-7-and-sony-alpha-a7r-ilce-7r-vs-rest-of-theworld-part-1/ (Scroll down to the middle of the article)
MTF (theoretical, from Sony site)
Using Imatest based tests (computing MTF and aberrations from captured chart) for UWA and WA lenses is always tricky and I am not sure how much relevant obtained data are for a regular users.
Anyway, here are official data from Sony (those are theoretical, computer generated and real life figures would be quite different).
We can’t directly compare those charts, because of different inputs, (only 10 lines pairs/mm are somehow comparable, and focal lengths are different anyway), but we can see the tendency characteristics of both lenses.
a) Wider end is overall better than longer end (10 and 16mm are slightly sharper than 18 and 35mm FL setting)
b) SEL 1635 seems to have somewhat less CA wide open, especially at 16mm
c) Center performance doesn’t improve by much when lenses are stopped down (it is already excellent wide open), and corner performance does improve, leaving weaker only extreme corners.
All those readings are rather typical for this type of modern lenses and in my honest opinion, we are at the point when UWA zooms are perfectly usable for professional work, no matter which one you choose. In comparison to my old Canon EF 16-35 f/2.8 L II zoom, both of Sony lenses are more evenly sharp across the frame and sharper in the center, which is great improvement.
Distortion and vignetting:
UWA and WA lenses have usually more issues with distortion at the first place, followed by vignetting, flare resistance, spherical and chromatic aberrations. In most shooting scenarios that I have experience in my career, those attributes were much more important than absolute sharpness (smeared corners could represent a problem though).
When it comes to distortion and vignetting, we have to consider that most modern lenses are designed with in camera (or external post processing) software corrections in the mind. Sony apply some of those corrections during file processing in camera, as we have been writing here for long time already, so it is hard to speak about optical qualities of the lenses when it comes to those values. Many people rightfully argue, that it doesn’t make sense to test those lenses without software corrections (some of them can’t be disengaged anyway), while others argue, that every single software correction results in loss of resolution.
Without lens correction applied, Sony SEL 1635Z has pronounced barrel distortion at 16mm of above 3% in our measurement, while with correction set to on, distortion stays bellow 1% (around 0,7%) which is great result. Sony SEL 1018 shows lower distortion uncorrected of around 2%, and with correction applied it comes to similar values as its bigger brother (0.8%). In both cases, we can notice slight loss of resolution in the middle part of the image, but really nothing that should affect your images in any way, until some very specific purpose.
At longer end, both lenses shows light pincushion distortion but it stays, even uncorrected, in a very reasonable range of around 1%.
Bellow is a set of the images where you can see uncorrected and corrected version next to each other, just use the slider in the middle to swap between images.
Sony FE 1635 at 16mm
Sony FE 1635 at 20mm
Sony FE 1635 at 24mm
Sony FE 1635 at 28mm
and Finally – Sony FE 1635 at 35mm
We can follow how pronounced barrel distortion change toward pincushion distortion with increasing focal length. Corrected files looks good and the only remaining question is – how much we lose of the original resolution with software correction.
Simply said – basically nothing for most users.
Only in the very critical applications we might notice IQ loss on the pixel level.
(This is valid for low ISO shots, where dynamic range is very wide and tone transitions are smooth. Going toward high ISO settings we might notice some artifacts in certain circumstances though.)
While every software developer would apply different algorithm in the lens profile, they are never linear, so resolution loss might not be visible in the extreme corners but in the zone before them e.g.
Nevertheless bellow are two 100% crops of extreme corner and mid frame to give you better idea – how much (distortion) software correction cost in terms of IQ at 100 ISO. We stroked red circles to map areas where differences are most visible.
Similar pattern can be seen with Sony SEL 1018 lens.
Sony SEL 1018 at 10mm
Sony SEL 1018 at 12mm
Sony SEL 1018 at 14mm
Sony SEL 1018 at 16mm
And finally – Sony SEL 1018 at 18mm
Toward longer zoom ends (18 and 35mm), both lenses improves in terms of distortion, ending up bellow 1% (approx.) of pincushion distortion after correction. At 14mm (SEL 1018) and 24mm (SEL 1635Z) we measured lowest distortion without corrections.
Vignetting is one of the issues with both lenses that we are not really happy about. Both lenses, without software correction, shows very strong vignetting at wide ends – SEL 1635Z around 3 EV, and SEL 1018 just slightly less. With correction enabled, they both comes to more reasonable -2 EV and while this is more typical for UWA zoom lenses, we already have applied some software correction to match the results. This mean that if you apply further corrections to get rid of remaining vignetting, noise will be increased in corrected parts. Not exactly right, but for better interpretation imagine that your sky in the landscape shot taken at 100 ISO, will have after correction noise at the corners corresponding to 800 ISO! This can affect resulting image quality, depending on the particular shot. (Even more if you are planning to stitch images together, in panorama mode e.g.)
On top of rather strong vignetting, Sony E mount cameras are dealing with color shifts in the corners from the day one. Part of the problem is in the sensor topping construction, but main reason remains very short flange distance, forcing outer parts of the sensor to collect light under very steep angles. Sony is applying color shift corrections in their ARW (lossy RAW format) files and we can’t really influence it. Depending on the situation, this color correction can further reduce IQ in the corners of the image, especially on the wide end.
Probably for the reasons listed above, Sony produce lenses with wider field of view than officially nominated (and wider projection circle too) and crop them within processing pipeline in order to deal with short flange distance. This might be a reason why SEL 1018 covers even full frame from around 14mm.
While distortion improves toward longer end of the zoom, with vignetting is not that good, because even at longer ends, it is still hefty at above 1% after correction.
Good news is that by stopping aperture down to f/5.6 and smaller, vignetting (and color shift) significantly improves and at f/8 it comes down to around 1EV uncorrected, which is more than acceptable. Considering intended use of those lenses, vignetting become less of the problem then it seemed initially.
Real life comparison:
As we wrote at the beginning, we didn’t have SEL 1635Z long enough to collect all data for full review. We focused rather on direct comparison in few situations between two camera systems – A6000 with SEL 1018 and A7MII with SEL 1635.
The first scenario is the favorite – wall shot.
Yes, we know how stupid those shots are, but because we are not presenting you here our usual ISO chart comparison, we needed to find something where we will be able to compare relative sharpness across the frame. Because of the differences in DOF we also tried to look for the rather flat target, but on the other hand not too flat to be deadly boring.
So it’s wall… At least, we tried to find nice legacy wall 🙂
Following image shows selected zones (A,B,C,D) from which we further present 100% crops.
For this test we used ARW files corrected in Lightroom CC with Adobe lens profile.
Uncorrected files will show more distortion and vignetting but because those lenses can’t be used on any other system than Sony E (so far), it won’t make much sense to compare them uncorrected.
Here are respective 100% crop comparisons, at different focal lengths.
Sony A7 + FE 1635 at 16mm vs Sony A6000 + SEL 1018 at 10 (15mm FF equivalent)
In the last row, we added Sony A7 + SEl 1018 at 16mm. As we wrote in the introduction, this combination was used by many owners of A7 bodies prior to FE 16-35 release, because it was only wide option with AF and image stabilization at the time.
Looking at above comparison, it can be seen that Sony A7 + FE 1635 has slight advantage over A6000 + SEL 1018, at widest aperture but not as much as some might have been expected. This is not because FE 1635 is not good, but because SEL 1018 is exceptional little lens.
Last option however shows significant decrease of image quality, especially toward corners but stopped down to f/8 and smaller, it actually come close to other two alternatives.
Sony A7 + FE 1635 at 20mm vs Sony A6000 + SEL 1018 at 13 (approx. 20mm FF equivalent)
This time we included also one of the best wide-angle primes, recently available on the market – famous Zeiss Distagon 21/2.8 (in ZE – Canon mount with Metabones Smart adapter).
While in the center of the frame, Zeiss has slight edge over other two lenses wide open (Zeiss at f/4), toward edges it clearly stays behind them, with FE 1635 ZA being most evenly sharp across the frame.
(Of course, Zeiss will perform much better on its native body, so this test is certainly not representative in terms of its optical qualities).
At f/8 Zeiss pulls apart, but zooms keep their own and in a real life scenario, it would be hard to see any differences (in the resolution).
Sony A7 + FE 1635 at 24mm vs Sony A6000 + SEL 1018 at 16 (approx. 24mm FF equivalent)
In this comparison we included also A6000 with Sony FE 1635 (at 16mm which gives effectively 24mm FF equivalent on APS-C), and this combination is very good. For some, focal range of 24-50mm is the most used one, and FE 1635 on APS-C gives great IQ in that range. (Better in corners than Sony SEL 1670Z or FE 2470Z)
This combination can be also attractive for those who are considering movement to FF in the near future.
At this focal length all above presented systems deliver crisp, great quality images all the way to the extreme corners and differences between them are field irrelevant (in terms of resolution).
Sony A7 + FE 1635 at 28mm vs Sony A6000 + SEL 1018 at 18 (approx. 27mm FF equivalent)
This was last focal length where we could directly compare both systems (FF and APS-C). We added also Sony FE 28-70 f/3.5-5.6 kit lens at its widest focal length.
18mm is near strongest FL for FE 1635 and certainly weakest for SEL 1018, but differences are still rather small.
FE 2870 holds its own in the mid frame, showing declining image quality toward edges and corners at f/4 but improving a lot already at f/5.6. It never gets clarity of other two lenses in the extreme corners, but otherwise it shows its surprising rendering qualities (considering that it is cheap kit lens).
Sony A7 + FE 1635 at 35mm vs Sony A6000 + SEL 1018 at 18 (approx. 27mm FF equivalent)
SEL 1018 can’t zoom to 35mm, so we included FE 2870 again and added Sony FE 3528 ZA in the mix.
While Sony FE 35mm f/2.8 ZA is sharpest of the three, even at this focal length (that is clearly weakest for FE 1635), it is quite hard to notice significant differences, except at the extreme corner. FE 28-70mm shows very strong performance at 35mm too, standing just close third in comparison with other two much more expensive options.
We didn’t have time to provide controlled test of flare resistance, but we can say that FE 1635 has very effective coatings, especially when used on other than Sony A7 camera (there are problems with sensor reflection with original A7).
Bellows are just two shots with Sony A7MII and FE 1635 in a direct back-lit situation, showing no flare and very little contrast loss. (results for Sony SEL 1018 were presented in our original review of that lens)
AF and OSS
AF was very fast with both lenses, but we experienced some problems with accuracy in low light conditions. (Especially with A6000 and SEL 1018). Lenses themselves are silent, fast and snappy and the AF performance is more dependent on camera body capabilities.
In our experience we would rate image stabilization effectiveness at around 3 EV in average. We didn’t notice any further improvement of hybrid 5 axis stabilization (A7MII) vs standard in lens optical stabilization. We do recommend to switch stabilization off, when you use lenses on tripod.
Looking at all those results, we can only confirm what we already know from our long-term real life experience with Sony SEL 1018 f/4 OSS. It is outstanding little WA zoom, such a great addition to Sony E mount APS-C program.
On the other hand, Sony FE 1635 f/4 ZA OSS, is equally good and in most areas even slightly better WA zoom but most significant difference between the two are their different sensor size belonging.
On one hand you have full frame system of the exceptional image quality, but you have to pay for it in terms of purchase price, size and weight. On the other hand you have smaller and cheaper APS-C alternative that comes very close in resulting IQ, so close that sometimes we have a problem to tell them apart.
Difference in IQ will start to be more visible with increasing ISO, while even at base ISO, you might notice slightly wider dynamic range of the full frame system.
We also feels that Sony – Zeiss FE 1635 lens, has somewhat puncher colors and slightly better flare resistance. (could be also character of the sensor though).
How tinny differences we are talking about are best represented by following pairs of the images that we took on intention to see for ourselves. We took them all hand-held, trying to simply simulate real life situation of potential user of those systems, testing so not only resolution, but also OSS, auto focusing accuracy, colors, contrast etc. We did post process images to certain extent, trying to get best for both systems, same as we will do with real photos.
We won’t tell you which image was taken with which camera + lens combination. If you’ll be able to tell them apart, you will know which system you really need, if not… well… GAS will make that decision for you.
You can go to our forum section to share your thought and guesses if you like and to discuss this article. We might there also post once a key to used systems.
Here is the forum link: http://www.verybiglobo.com/forums/topic/sony-fe-16-35mm-f4-za-oss-vs-sony-sel-10-18mm-f4-oss/
You can find the correct answer which camera + lens took which image (bellow) on the related forum (link above). There is also link to two full size images on the Flickr for pixel peepers.
(Only A7 + FE 1635 and A6000 + SEL 1018 were used for following images)
Sample gallery of Sony A7 and FE 1635/4 ZA OSS
Please help support this page and upcoming reviews and buy through Amazon affiliate links, with no extra cost for you:
Sony 16-35mm Vario-Tessar T FE F4 ZA OSS E-Mount Lens
Sony SEL1018 10-18mm Wide-Angle Zoom Lens
Sony a7 Full-Frame Mirrorless Digital Camera – Body Only
Sony ILCE7M2/B ILCE7 Mii ILCE7M2/B Mark II Alpha a7II Camera (Body) w/ Sony 55mm F1.8 Sonnar T* FE ZA Full Frame…
Sony a7R Full-Frame Mirrorless Digital Camera – Body Only
Sony Alpha a6000 Mirrorless Digital Camera with 16-50mm Power Zoom Lens
To help this page survive, your donation will be highly appreciated.
Very useful stuff as usual. I don’t have plans to go FF, because lightweight is one of my primary concerns, but you’re confirming I’m not losing much – possibly almost nothing – staying with the a6000 + SEL1018.
Unless you need high ISO, there really isn’t much need to switch to FF.
Are you forgetting DOF and DR?
I am not forgetting it, but there are quite fast lenses that you can use on APS-C to get to shallower DOF, in most cases – shallower than I ever need. Regarding DR, you are right of course, but DR of A6000 is very good too and I have more problems with Sony lossy compression than with the DR difference.
However you are right that both aspects are FF advantages, while APS-C has an advantage of longer reach (with highest resolution) and smaller system in general (smaller body, smaller lenses).
I am using both system equally…
Just give me an IBIS NEX-7 alike+ my Sigma 60+30+ wider but better than the 19 and I will not ask for more…
Thanks Victor, very good review with lots of full size samples, as always…. My findings have been confirmed, SEL1018 is an outstanding uwa lens, shocking me everyday i use it…how little, lightweigh but captures awesome quality images…I had both Zeiss 1635 and SEL1018, but couldn’t sell 1018, since it’s a gem to keep it forever.
On the other hand, 1635ZA has also exceptional sharpness across the frame, between 16-30 it acts like a prime…great sharpness across the frame and also extremely flare resistance…
correcting…samples are not full size 🙂
Thanks Hakan for your comment. Sample images are 2600px on the longer edge, which should be enough to see meaningfull difference if there is some.
The reason why we didn’t post original size images is in the web page optimalization (for the mobile devices) and the rented server space in general (it cost us money of course).
Because you can see some of controlled and rather comparable shots and respective 100% crops earlier in this short review, the purpose of the comparable real life samples was to give my readers the idea – what they can expect from system A or system B in average conditions and situations. We believe here in Verybiglobo.com that 2600px wide is fair enough, but we do not force any conclusion based on that.
I’ve sent you two messages through your site ”contact us” page, asking some more info about A7R+FE1635,,,Could you please check them?
I will try to check it, but if there is chance to move your question here, or even better in the forum section, where you can start new thread e.g. in the Q&A section – http://www.verybiglobo.com/forums/forum/qa-forum/, I will be able to respond much faster. Due to the large number of emails, I am not able to respond promptly on all of them. Comment and especially forum sections are much easier to keep updated. Thank you for understanding,
Many thanks for the reply, i initially thought that ”contact us” would be quicker to reach you. Next time, i’ll open a thread!
Also in terms of blurring out the background, the Sony f4.0 APC and Fullframe lenses seem to deliver the same bokeh, can you confirm? The images are near identical in character also.
Hi Alec, you are right in both conclusions. I am still using SEL 10-18/4 since it is really good and very compact lens.
Amazing review ! Thanks for taking the time to do this . Wish I had found this way sooner switching back and forth between the a7ii and a6300 trying to get a good and reasonably light set up for hiking and shooting . Think I will go all in on the a6300 and 10-18 oss. Cheers !
Thanks, Bjonar, you won’t be disappointed IMHO.