A Look At The New Sony FE 100mm f/2.8 STF GM OSS With Smooth Trans Focus

The new Sony FE 100mm f/2.8 STF GM OSS lens is one of Sony’s latest lens offerings for their full frame mirrorless cameras. It’s another lens in Sony’s pro-grade G Master series, with built-in stabilization and SSM autofocus system. It’s designed to be used as a portrait lens, and it has what Sony calls a Smooth Trans Focus system.

So what is this Smooth Trans Focus that Sony has been talking about, and is it just a gimmick? Well, for the former, it’s essentially an apodization filter that works together with the 11 blade rounded diaphragm to produce very smooth bokeh with soft-edged out of focus highlights. So what’s an apodization filter? Good question. In the case of this Sony lens, it’s a concave neutral gray tinted lens element near the aperture that produces a smooth transition to zero intensity out on the edge of the frame. In short, think of it as a radial neutral density filter that’s clear in the center, and darker on the edges – the opposite of a conventional center filter that helps combat vignetting. And like vignetting, the apodization filter will get darker as it gets closer to the edge.

Since this filter is limiting the amount of light coming through the lens, this apodization filter cuts up to two stops of light – much like an ND filter would. So while the lens itself is titled as a f/2.8 lens, the markings on the barrel actually start at f/5.6. For clarity, you will still get the depth of field qualities of a 100mm f/2.8; you’ll actually have to meter the exposure value at f/5.6 to ensure the correct exposure.

So why would you want this strange filter in your lens? Well, the results are that it gives you incredible smooth transitions within your bokeh, by knocking out all of the distracting hard edges in the background. It does this by having the out of focus elements create disks instead of your standard points on the sensor plane because they’re out of focus. The filter helps taper off the exposure of those disks, fading them out on the edges. The result is silky smooth bokeh, instead of having distracting elements that you might typically find within bokeh.

Simply, what this lens is: a portrait lens that gives very smooth, natural-looking bokeh.

What this lens isn’t: a lens ideal for low light shooting, or for getting swirly, cat eye, or soap bubble type bokeh

So what does this smooth bokeh look like? Here’s a sample we set up in the studio:

Sony 100mm f/2.8 at f/5.6

This was at maximum aperture, with the focus on the center of the orchid on the left. The beads hanging in the background just kind of fade to mush. We also did some comparison shots with the Canon 100mm f/2, Zeiss Milvus 100mm f/2 Macro, and Nikon 105mm f/2. DC:

Canon 100mm f/2 at f/2

Zeiss Milvus 100mm at f/2

Nikon 105mm DC at f/2 (No DC)

These were all wide open, so they’re a stop faster, but we shot them like that to maximize the bokeh effect to illustrate the difference to the Sony. As you can see, all three of these lenses have hard edges to the out of focus highlights. And that’s the difference.

Just for fun, we also shot test charts with these four lenses, for a quick resolution comparison.


Sony 100mm Focus Chart

Canon 100mm f/2 Focus Chart

Zeiss Milvus 100mm Focus Chart

Nikon 105mm DC Focus Chart

These are single samples that all passed our standard inspection, so don’t put too much weight on these charts. That said, the chart shots show some things we already know, that the Zeiss is nice, the Canon is pretty good, and the Nikon is an old design, and it shows. And the new Sony FE 100mm f/2.8 STF GM OSS looks really good. $1500 good? Maybe, just maybe. If you shoot tons of portrait work, it could certainly be an asset in your bag. If you don’t, it could be a fun weekend rental to try something new.

Side note about the Nikon 105 DC: the defocus control doesn’t soften the edges of the bokeh like an apodization filter. It alters the bokeh by shifting some internal elements forward or backward. The images below show what that looks like in practice.


So what do you think? Do you think this unique design has a future in the photography world? 

Author: Joey Miller

I’m Joey. I love cameras, especially old film cameras, and I can’t remember the last day I didn’t take a photo. Digital cameras are great, and they keep me employed, but I also still like processing my own film. I’m stuck somewhere in the middle. I shoot every single day, no matter what.

Posted in Equipment
  • Jonathan

    In decent light or better, and especially when combined with off-camera lighting, and when used on a body with EYE AF, I’d have to imagine that this lens would create some incredible portraits. Let’s consider…
    More DOF than shooting wide open with a faster lens so more of your models face is in focus
    More background blur providing more separation
    More accurate AF with EYE AF
    Seems like a recipe for some WOW images.
    Now, one shouldn’t just obliterate every background, but there’s no denying that special images can be created with the right background blurred out.

  • Nikolai Vassiliev

    Dear Rupert,
    actually i do not simplify lens characteristics to only three parameters but this three can be easily measured and compared, pity but matter of fact of such ‘3 images tests’ (with all my respect to LensRentals Blog).
    So, STF lens (i have a pretty brief experience with Minolta version) is a pretty different beast and i suppose (for better understanding) to perceive it for first approximation (as physicists says) its not bad to count ‘blur amount’, DOF and actual transmission as a base.
    If we left behind amateurs obsession with crazy little DOF like 1.4 or 1.8 100-135-200mm lens (of course needed in some rare cases) i think for portrait (and macro!) work it can be pretty interesting to get more DOF without less smooth background blur.
    My previous comment was about it – does APO filter provide more acute nearly-in-focus areas like increased DOF?

  • Rupert

    CoC in practical reality is a dot (filled in circle). In STF/APD lens the dot fades toward the edges.
    Anybody not calling it a circle would be being excessively pedantic.

    Diffraction is generally only noticeable at small apertures where the STF/APD has no effect, so I don’t expect any significant change there.

  • Rupert

    I often mutter at the screen that there is no need for people on forums to treat each other as poorly as they do. Today I find myself in the position of needing to apologise for my terse reply to your comment yesterday.

    In my original comment I was calling out Joey for simplifying his description to the point of inaccuracy. I should have known that someone on the LR forums would hold me to the same standard. 🙂

    It is interesting to me in the comment by Nikolai (which appears before yours even though it was posted later) that he has simplified the lens down to 3 performance metrics – blur amount / focus plane / light transmission – this, as we know, is technically incomplete. However I feel that for many photographers it might be sufficient. (definitely better than my attempt to use 3 adverbs to indicate that f4 was an incomplete answer)

    The reality is that if we wish to be completely accurate in describing the performance of apodisation filter optical designs we will probably need to be introducing new terminology. As you suggest, incorporating a [differential] intensity function might be one way to go. Optical designers may already have these in use.

    In conversational situations the performance of lenses is usually reduced to general descriptive terms (eg: smooth, creamy, fussy or distracting bokeh) I expect that we will eventually fall into using something similar for apodisation lenses should they become more common.

    Which brings me back to the very beginning. . .

    Joey wrote: “you will still get the depth of field qualities of a 100mm f/2.8” I believe that this indicates to most people “no change” when in fact there is potential for considerable change.

    My use of “[3 adjectives] f4” as you pointed out is technically incorrect. I had intended that it at least indicate “some change”.
    In general use (gross simplification), most people will perceive an extension of the zone of acceptably sharp focus.

    Perhaps Nikolai’s “blur amount / focus plane / light transmission” while still an incomplete description, might be a better choice for conversational use. The full Monty, in my opinion is too much for conversational use.

  • CekariYH

    “As you can see, all three of these lenses have hard edges to the out of focus highlights. And that’s the difference.”
    No, not att all, it’s not the only difference, all three non-Sony has oval OOF circles in the off center area compared to the Sony one.
    You should redo the the Sony STF ones with with shots from T5.6, T6.3, T7.1 and T8 to find out that the circles will still be round and nice compared to the other three.

  • Dave

    But strictly speaking DoF is defined in terms of areas of acceptable focus, which is defined in terms of the measured circle of confusion. Now based on the image samples I’ve seen, the apodisation lens never hits zero transmission. Hence for very bright points of light you can still clearly see the aperture cutoff. So the aperture remains the same and the diameter of your projected point of light will be the same with or without apodisation (my bad, i should have said defining your CoC not measuring it earlier. You could incorporate an intensity function and get different results is my point). It’s interesting as for less intense sources of light then the camera system will have the effect due to the ADC in your system, so you end up with the somewhat counterintuitive idea that DoF will be a function of brightness of the object you are recording and DR of your camera. To say it is approximately f4 is as misleading as saying it’s f2.8 or f5.6.

  • Nikolai Vassiliev

    Fujifilm and Laowa (only other manufactures with such lens – except for Minolta) has relatively low density apodisation elements and effect is significantly less visible.

    For DOF matters – blur amount is 2.8, focus plane is more like F4, light transmission is 5.6 for this 100mm STF lens.
    Interesting, is a circle of confusion is still a circle for STF lens? Does it suffer from diffraction effects in the same amount or we have a stop or two more usable?

  • Rupert

    Because the apodization filter darkens toward the edges of the glass, the effective aperture is smaller (actual f2.8 but meter for f5.6) this changes the depth of field regardless of the metric used to measure or define it.

    thus, as I posted earlier:
    you will get a depth of field SOMEWHAT CLOSER to what you MIGHT EXPECT from APPROXIMATELY f4.0 with extra smooth bokeh.

    Measure your circle of confusion in whatever manner you wish.
    As long as you use a consistent metric in your measurement, you still will NOT get the depth of field qualities of a 100mm f/2.8 from this lens.

    This reality is counter to what Joey stated.

  • Dave

    Being nitpicky myself… but define depth of field. You could argue it both ways depending on how you are measuring your circle of confusion.

  • jswendell

    Wow, this was a perfect bokeh torture test/comparison to show this lens at maximum advantage. The other three shots look hideous by comparison. I didn’t think I cared about this lens. Now I’m wondering how I can live without it.

  • Rupert

    This lens is a nice option for Sony users who want the smoother out of focus zone, however you should note that apodization filters generally increase the depth of field as effectively they make for a smaller aperture.

    Thus the lens will tend to NOT give the depth of field expected from a 100mm f2.8 as you stated, but in reality a depth of field somewhat closer to what you might expect from approximately f4.0 with extra smooth bokeh.

    The best way to see this effect is to use the same optical design with and without an apodization filter. Conveniently, Fujifilm have given us the XF 56mm f1.2 APD (which has an apodization filter) and the XF 56mm f1.2 (same optical design without the APD filter).

    I know I’m nit picking, but people come to Lens Rentals blog for accurate info in among a sea of loose thinking elsewhere on the net. Thus we hold you to a higher standard.

  • brnpttmn

    I’d love a Sony-Zeiss 105/2.5 Sonnar.

  • Excuse me sir, but you are a cyborg as your vision senses is enhanced by device 🙂

  • denneboom

    i wear glasses, i have -4.5 and -4.75. without them everying far away is unsharp for me.
    at night, car lights look like bokeh. (at day the effect is far less)

  • l_d_allan

    off topic?
    I’d much prefer that Sony would introduce an affordable portrait lens in the 100mm to 135mm focal length at about f/2 and about the same price as the $600 85mm f/1.8. To this owner of the FE55-f/1.8, the 85mm focal length is too close to the 55mm.

  • J L Williams

    “Natural-looking bokeh”? Human vision doesn’t experience bokeh effects at all, of course — it’s purely a photographic artifact, so ALL bokeh is unnatural. Smoother or less obtrusive? I’ll go along with that.

  • denneboom

    Smooth bokeh is nice to have, but i dont think everyone in the Photography world will change to STF lenses soon. When everything is Perfect out of focus pictures will look more like a composite.

    The standard bokeh (or should i say “non perfect”) can add character to a image.
    I even shoot with a minilta 50f2 on my sony a7, even though it has only 6 apperture blades. the image from that lens is special to me, thats why i like it. (and its decent sharp too)

Follow on Feedly