Sony FE 85mm f/1.4 G Master Lens MTF and Variance

Published April 11, 2016

fe 85 G

We are pretty excited about the new Sony G Master Lenses for FE mount cameras. Our excitement may be a little different than yours because we see the Sony FE lenses as an evolving new product. There were some things we really liked and some things we really didn’t like about the first FE mount efforts, but we’ve been seeing signs of progress. We were hopeful that the new Master lenses were going to take another step up in optical quality.

The first one we took to the lab for testing was the FE  85mm f/1.4 G Master lens. Early reports from photographers indicate that Sony has done well with this one. We don’t always review lenses, but we do test them for MTF and variation, and for us, reducing copy-to-copy variation was one of the things we really hoped this lens would do.

A Note About Sony MTF Testing

I discussed in our post about the 70-200mm f/4 lenses that our testing algorithms and presentations are changing as we improve things and try to make them more scientific. If you missed that there are two major points.

First, the variation algorithms have changed, both to make the charts easier to see (we show a 1 Standard Deviation range, rather than the 1.5 S. D. we used to) and to eliminate the Consistency Number. We found that the Consistency Number was too blunt of a tool; it showed only one limited part of variation. There were cases when there were two lenses had the same consistency score, but one was quite a bit worse than the other in ways the score didn’t show.

Second, I want to continue to point out that the MTF bench is not designed to test lenses that require power to maintain focus position, which FE lenses do. We’ve worked around that by making an electrically live mount, but the electronics block some of the test points at 20mm from the center (the right side of the graph). For that reason, the measurements at the edge have fewer measured points than the other points tested. Take them with a mild grain of salt.

Finally, one note about the Sony 85 G Master lens in this test. We found that it performed best with 2mm of optically glass placed in the pathway, simulating the cover glass of a camera sensor. Most 85mm lenses we’ve tested do not improve with glass. This is of NO significance to persons shooting with the lens, but I mention it for completeness sake. We check every lens we test to make sure we’re presenting the best MTF we can obtain. (If you’re interested in how much difference it makes, at the end of the article I presented the MTF charts of the same group of lenses tested with, and without, the glass. It can be very significant.)

MTF Tests

I’ve compared the MTF results for the Sony FE G Master with four other well-known 85mm f/1.4 lenses: The Zeiss ZE 85 f/1.4, Nikon 85mm f/1.4 AF-S, and Zeiss Otus 85mm f/1.4.

OlafOpticalTesting, 2016

OlafOpticalTesting, 2016

OlafOpticalTesting, 2016


It should be pretty apparent that the Sony lens belongs in this company, at least as far as MTF goes. The ZE is a bit sharper in the center but the Sony is better toward the edges. The Sony and Nikon are actually quite similar. The Otus, of course, smokes all the others in the central part of the image, but at that price I would expect it to. You can split hairs all you want about which is better at what, but these are all really good lenses. The tiny differences between them would certainly be lost in the other variations of cameras, lighting, and technique.

And yes, I know you all wish I had the Zeiss Milvus 85mm f/1.4 MTF charts to compare it to. I wish I could ever get 10 of the in stock at one time so I could test them. So far no joy, I’m afraid.

Copy-to-Copy Variation

We ran our new algorithms on the same lenses we tested above to give you some frame-of-reference for comparison.


OlafOpticalTesting, 2016

OlafOpticalTesting, 2016


Let me get one thing out of the way really early. Those of you who compare will find the graphs for the other three lenses are just a bit different than the original variance graphs we’ve used, because, as I said, the algorithms are different and a bit more sensitive. But in our previous system, the Zeiss ZE Planar lens was acceptable and the Nikon and Zeiss Otus lenses very good as far as the copy-to-copy variation.

The Sony FE 85mm f/1.4 G Master lens compares very well with these other top lenses. There’s none of the severe copy-to-copy variation we saw with the Sony FE 35mm f/1.4 lens. Just to give you more of a comparison, below are the variation graphs for the Sony FE 35mm f/1.4 and Rokinon 85mm f/1.4 lenses, both of which have significant copy-to-copy variation. When you compare the G-Master to these lenses, there’s no question Sony has done a really good job improving variation on this new lens.


OlafOpticalTesting, 2016

OlafOpticalTesting, 2016

Because I know some fanboys are going to ask, yes we’ve retested Sony 35mm f/1.4 lenses recently, and no the copy variation hasn’t changed. If you thought it might you need to read this to understand why it’s unlikely to happen. I know you really, really want it to be so. Really, really wanting something to be so, unfortunately, doesn’t work all that often.


Of course, the proof of a lens is in the pictures it takes, not in this simple testing. But the test results can be useful and these are certainly reassuring. The Sony FE 85mm f/1.4 G Master lens is pricey, but not much different than the price of other top-quality 85mm f/1.4 lenses. It delivers the image quality you would expect for the price and copy-to-copy variation is reassuringly low. Sony has definitely taken strides to fix the variation problem that we saw in some of the earlier FE lenses. Give them credit, and also give them some understanding. They are turning out lots of new lenses in a new mount with a bunch of new technology. There are always going to be some growing pains when a company does that.


Roger Cicala, Aaron Closz, and Brandon Dube

April, 2016


Addendum: The difference between glass and no glass.

I talk a lot about how some lenses require a certain amount of glass between the back of the lens and the image sensor. All digital cameras have some. Certain lenses are rather insensitive to this and don’t care if there is glass present or not. Others are tuned for a specific amount of glass between lens and sensor. Wide-aperture, wide-angle lenses are most likely to be sensitive, but some longer focal length lenses can be, too, as was the case here.

I’m just putting the MTF charts with and without optical glass in the path to show you how much of a difference that made with this lens. If we had presented the results without glass the Sony lens wouldn’t have looked very good at all.


OlafOpticalTesting, 2016

When we talk about the glass effect in photographs we tend to concentrate on problems in the corners, which definitely do occur and are noticeable. But notice also that center sharpness and astigmatism/lateral color are also affected significantly. (Without doing specific tests, it’s difficult to tell whether the sagittal-tangential separation on the MTF chart is from astigmatism or lateral color aberration.)

Author: Roger Cicala

I’m Roger and I am the founder of Hailed as one of the optic nerds here, I enjoy shooting collimated light through 30X microscope objectives in my spare time. When I do take real pictures I like using something different: a Medium format, or Pentax K1, or a Sony RX1R.

Posted in Equipment
  • spaceman

    Great article!

    Are the comparative mtf graphs averaged over apertures? In that case I’d like to see the mtf graphs wide open.

    According to DXOMark the Otus is incredible wide open. The Nikkor slightly less incredible, but the Zeiss ZF.2 seems appalling wide open ( compared to the other two, when it comes to resolution.

    I would love to see how the Sony compares wide open as that is how I would shoot it very often, even some landscape shots wide open if the sharpness in the focus plane is very high!

  • I’m afraid we haven’t checked the Zeiss 135, but chances are high it is not sensitive to stack thickness.

  • NightPhotographer

    Thank you again for spending your time providing us with information that we cannot find anywhere else. I have a question. Does this glass between the back of the lens and sensor that you are talking about resemble the sensor stacking? If this is the case, it seems that the difference that stacking makes is huge. It’s for a while that I want to get an A7XX for my Zeiss 135mm APO but due to the fact the Nikon’s sensor stacking is defenitley different than Sony’s, I am afraid that mounting this lens on a Sony camera is not a wise decision . Can you please tell me if you have tested the Zeiss 135mm APO ZF.2 on an A7XX? Is this lens sensitive to stacking?


  • Henry, we don’t test on-camera anymore. Someone else will have to try that, and it’s a good point.

  • Henry Z

    Thank you for your work!

    By reading your posts about the cover glass thickness I believe it would become a more and more critical variable while we actually use the lens feasibly.

    Therefore, would you please test the 85GM optical performance with Kolarivision A7R2?

    I’m asking it because the 85GM possibly is the best AF85 lens that ever launched, which makes Sony becoming quite advanced on portrait photography… Meanwhile, the capability with legacy lens should ideally makes (kolarvision-modified) Sony more superior on landscape photography.

    — Therefore, It is quite attractive for me to know if the Kolarivision A7R2 is good on both wide-angle legacy lens (like Zeiss D15/2.8 ZM) and native G Master lens (like the 85/1.4 GM)…I would even pay for the test result for your professionality.

  • Thanks for the detailed explanation as always Brandon!

  • JGro, lenses at 400 and up we still have to use test charts and Imatest.

  • JGro

    Too bad. 🙁 Being spoiled by the level of sophistication of your testing, I find it hard to make any decision for lens purchases without a lensrentals test. 😉
    Since you use your testing equipment not only to provide MTF and variation tests for us readers, but also to QA your rental fleet, I wonder how you handle QA for long lenses. Do you have other means of assessing such a lens, which does not give that amount of data but is good enough for QA?

  • joel richards

    Can’t wait to see it when you get sufficient data.

  • Brandon Dube

    The charts from version 2 of the software and on have the number of samples printed on them. The milvus data (which Roger forgot we have, oops) is from 10 samples. 3 Batis lenses have been tested.

  • Brandon Dube

    Three Batis.

  • Brandon Dube

    No. Going to 400mm is already hell from vibrations, and the target in the MTF bench is not small enough for a 600mm lens.

  • Lee

    As in, this chart was based on 3 copies instead of the usual ~10?

  • speedy fisher

    I didn’t think so. Still interesting to see some numbers for the filter stack though.

  • There really isn’t any way we can tell much about that.

  • Omesh, it is to some degree.

  • Omesh Singh

    Hi Roger, Thanks for the valuable info.

    I noticed the Otus MTF indicates a dip towards the corners. Is this due to field curvature?

  • speedy fisher

    Interesting data on the sensor glass. How much difference do the microlenses on the sensor (below the filter stack) make on top of this, since they do vary between camera models? I’m guessing that there’s no easy way Olaf can test something like that.

  • alowprofile

    Ok. Thanks Roger.

  • David Braddon-Mitchell

    Three Batis or three Milvus?

  • JGro

    Hi Roger,

    I know you and Olaf have a busy schedule, but is there any chance we will ever see MTF and variation data for the three 150-600 zooms?



  • Hakkaan, writing it up tonight, should be out tomorrow.

  • It’s definitely noisy, although it’s much quieter with the camera set to Video than Photo. On the other hand, AF is slower in video, which I assume is why it’s quieter.

  • John Dillworth

    It certainly looks like Sony on their best day is now competing with the other major players on their best day. Maybe you can still say Sony does not have the depth of Canon or Nikon but their 3 recent efforts put them in first class

  • alowprofile

    Hey guys. Thanks for testing! Looks good. Wondering if you heard any of the grinding/auto focus noise that everyone is referring to in the comments of this thread?

    Seems bizarre to have a de-clickable aperture ring for quiet video if your auto-focus is making so much noise. Please let us know. Thanks!

  • Hakann Vatansever

    thanks, great job as always..awaiting teardown of this lens and revealing of scratches/grinding noise issues all around.

  • Brandon Dube

    Andrew, consider the cone, fan, or bundle of rays that form the image on-axis (choose your own vocab).

    The rays from the center of the aperture are incident on the coverglass at an angle of about 0. The rays from the edge of the aperture are incident at an angle equal to the image-space numerical aperture of the lens. The NA and F number are related by F/# = 1/(2NA).

    The coverglass is plano on both sides; if you do snell’s law with an input angle of about 0 (so angle of incidence of about 90) and again with an input angle of about 30 degrees or so (for f/1.4), these rays do not go to the same place. The result is a blur (spherical aberration). There is some small axial color added as well, since the coverglass has some dispersion, but it will be very very small.

    Due to vignetting and the tilt of the incident bundle off-axis, the aberrations added are a bit less predictable, but still easily modeled in ray tracing software.

    Expensive microscope objectives, like the one below, have what is essentially a focusing element which adjusts the spherical aberration to accommodate a range of cover/slip glass thicknesses that are used in microscopy. The same feature is found in some of Nikon’s “DC” lenses, though with a less refined implementation.

  • Thanks guys for another test! Any ideas on why the cover glass thickness might affect results in the center? The previous theory was that oblique ray angles intersecting the glass in the corners would cause various issues, but it seems that the rays hitting the center of the glass should be perpendicular. I guess rays arriving from the edges of the lens to the center still have pretty steep angles, so that might affect things? But one would think that it would also cause color shifts in the center if that were the case.

  • Brandon Dube

    Three have already been done, but they rent like hotcakes. Suffice to say it is competitive with the other lenses in this class in all respects.

  • Guys, thank you all SO much for your time doing those tests and sharing your thoughts! One silly question – are you planning to put the Batis 85 though your optical bench any time soon? 🙂

Follow on Feedly