This 28mm is the cheapest prime lens for the FE system by quite a bit. It has a surprisingly fast f/2 aperture for its price too and although it's not that small (what fast FE lens is?) it is quite light. In the handfull of tests that I've seen for this lens it comes out pretty well for visible colour, but let's see what kind of skillz it has with infra-red...
● Mount: Sony FE
● Adapts to DSLR: No
● Adapts to (other) ILC: No
● Focal Length (FF): 28mm
● Field of view (FF): 75°
● Lens Elements: 9
● Lens Groups: 8
● Aspherical Elements: 3 (1 Advanced + 2 ED)
● Aperture range (f stop): 2.0 - 22
● F-stop increments: 1/3
● Aperture Blades: 9 (curved)
● Aperture Ring: No
● Autofocus: Yes (silent)
● Manual Focus Ring: By Wire
● Min. Focus Distance (cm): 29
● Max Reproduction Ratio: 1:7.7
● Internal Focus: Yes
● DoF Scale: No
● IR offset markings: No
● Body Material: Metal
● Filter Thread Material: Plastic
● Filter thread: 49mm
● Static Filter Rotation: Yes
● Dimensions (mm): 64 x 60
● Weight (g): 200
● Dust / Moisture Sealed: Yes
● Mount Seal: No
● Manufacture Country: China
● Manufacturing Span: 2014 >
● Price New: £240 - 400
● Price Second Hand: £220-300
● Accessories (Included): Hood
Quick IR Performance Scores
● Pure IR HotSpot (Lack of): 5/10
● Colour IR HotSpot (Lack of): 5/10
● IR Sharpness (Middle): 6/10
● IR Sharpness (Edges @ f/8): 4/10
● IR Sharpness (Edges @ f/2): 2/10
IR Quality / Hot SpotHere's a new set of images, a test for IR hot spot that I'm trying out (see more about it here). Basically what you're looking for here is a clean dark sky, with no bright spots.
This lens does vignette fairly heavily wide open and that can look like a large hotspot, but most of this is not. You'll find the hot spot more obvious here closer to the bottom of the set. This 850nm IR filter shows a patternef spot starting to get anmoying at f/8 here, but if a clean sky is important to you then even f/4 is still an issue. To some extent the hot spot is always there, it just gets larger, more blurred and subtle as the aperture opens up. Since this is the worst case scenario here f/8 is almost always fine under normal shooting conditions and mostly so is f/11. Thus I would say this lens is generally acceptable in relation to the hot spot issue. If you're someone who shoots pure infra-red at f/16 - f/22 often then I'd recommend you think carefully before buying this lens.
Here is a video to show the hot spot performance at f/22. This helps show how related to the bright areas of the scene they are. Unlike a standard circular spot these are not very easy to remove and compensate for.
This next set was shot using a Hoya R25A filter. On a full spectrum camera this is very similar to a 590nm conversion.
Although the sky wasn't quite so clear when I took this set with the R25A filter you can still see how much cleaner the colour infra-red is compared to the pure 850nm B&W samples. This could be partially because of less IR light coming through the filter, but also because of the reduced contrast in the scene. All the noticeable hotspot here is contained in the blue channel, not that it's much comfort if you're trying to process a colour IR image.
Check out these next couple of samples to see how some particularly high contrast scenes can be problematic, even at wider apertures. These are both shot at f/8, which is much more distressing. The bright spot here shows up very clearly on the tree trunk. I'm showing two samples here to show how the spot moves and confirm that it's not a fluke light on the tree itself.
Lens FlareHere are some samples at different wavelengths to show how it copes with unavoidable lens flare in each. These are all shot at f/8. As you can see the more infra-red light that comes through the filter the more obvious the flare becomes. It's actually very good in colour and even when external filters are used like this. In IR however it looks downright horrible. I wouldn't advise shooting into the sun when taking infra-red photography on this lens.
Kolari Vision Hot Mirror (Colour)
Hoya R25A (Red / 590nm)
B+W 093 (850nm IR)
SharpnessWhen I did a test between colour, UV and IR filters recently I got a lot of good comparison samples for this lens at different wavelengths. The difference in sharpness between the colour shots and all infra-red ones was quite noticeable. It was obvious overall, but in the corners it was especially apparent. All the shots were taken on a tripod and with accurate autofocus so I knew it was a fair test. I was unhappy with this lenses IR shots before this test and after looking closely at these samples I could now see exactly why. Click the image below to see the comparisons for yourself. It even shows up on these compressed samples at less than 2 mega-pixels, thus they certainly don't stand up scrutiny at 24mp (nearly fourteen times the size).
CompetitionLike most Sony lenses there isn't much in the way of competition for autofocus lenses (yet), but there is more here than with the 50mm FOV. There is now a 25mm f/2 in the form of the new Zeiss Batis, which is a high performance lens and has the added size, weight and cost associated with the brand. You could also get a Canon EF 28mm f/1.8 with an AF adapter. It might be a little faster than the Sony lens, but focusing will be sluggish on the first four A7 bodies. A potentially larger issue is the size and price by comparison (not including the cost of the adapter). I actually paid £100 less for the Sony 28mm lens (with a cashback deal) than the Canon usually costs.
If you're willing to forgo autofocus and buy second hand primes there are many options, but it will be hard to save money while keeping decent image quality and the relatively fast f/2 aperture. Although some old 28mm f/2 lenses can be picked up from companies like Vivitar or Sigma for a little less money, they tend to be rather lacking in corner quality. Models from Pentax, Canon and Nikon are interesting for their image quality, but they cost as much or more than the autofocus Sony. Although the adapters for them can be cheap they make the lenses quite a bit bigger and heavier too. Mostly front heavy & oddly balanced compared to the Sony 28mm.
The Leica M mount Voigtlander Ultron (with adapter) will be a better match for the size and performance of the Sony lens (albeit it a bit heavier), however it will cost more, have an odd filter size and still be without autofocus. On the flip side it does look cool on the A7 bodies (or retro, or hipster, depending on your opinion).
ConclusionsI wish there were more lenses like this for the Sony A7 series FE mount - Primes with faster apertures, at a more reasonable price (and quality). If there were I'm sure that the Sony A7 series would be even more popular than it is already. Perhaps Sony are hoping that 3rd party lens makers will fill these gaps later, but that was a risky strategy when they started out. Now that they've done so well there's a much bigger chance of that happening. Thus I feel much better recommending the system to people on a modest budget these days, but it certainly wasn't like that, even a year ago. As it stand right now, this Sony 28mm f/2 feels like the odd one out in an exclusive, high-end catalog. The FE lens collection is growing fast though, so this could change with the next announcement. I'm certainly crossing my fingers for more lenses like this soon.
In visible colour this lenses is a brilliant addition to the FE system and a game-changer for value primes, but... for infra-red I'm not at all convinced. Its IR quality just doesn't match that of its colour performance. It's like a totally different lens in infra-red and that's a real shame because this could have been a very nice walk-around landscape lens indeed. The hotspot that it produces isn't even the main problem. It's not the most extreme hot spot I've seen, nor does it show up often during normal shooting, it can just be problematic at times. The main issue here is sharpness (and to a lesser extent contrast). Infra-red photography can be passable, but it just never seems to shine. Clarity is generally low. Corners start to get better when stopping down heavily, but not long before diffraction becomes a problem. Now I'm making it sound plain horrible here and that's quite not fair.
Perhaps I've been spoiled by the insane performance of the Zeiss FE 55mm, but I do feel that if you're someone who zooms in on your photos to revel in their beauty in any way you won't be happy with this lens for infra-red.
General Pros and ConsPros (rated out of 10 for how awesome they are)
(9) - Accurate and silent auto focus
(7) - Good construction (metal... mostly) and great balance with the camera
(6) - A great value prime lens for Sony FE at last (although not really for IR)
(4) - Included lens hood is nice
Cons (rated out of 10 for how annoying they are)
(9) - Infra-red is noticeably less sharp and especially in the corners
(8) - Nasty looking lens flare in IR
(6) - Hot Spot can be problematic
(5) - Focus by wire speed is inconsistent, lack of feedback make it difficult to use
(5) - No Aperture ring
(3) - Plastic filter thread (forgivable at this price though)
(1) - Vignette could be better
Overall Score (Colour IR): 4/10
Overall Score (pure IR): 4/10
Below Average IR Performance
Below Average IR Performance
Here are some b&w infra-red images, taken with the B+W 093 filter (on the full spectrum Sony A7). This filter has a 50% transmission at 850nm (the amount of visible light that passes through this filter is nominal).
This last example shows the real-world colours (before WB / channel manipulation), exposure and tones of the hot mirror, 850nm and Hoya R25A filters together. The overlapping of the first two show a near black image. This illustrates what happens when an IR filter in front of a normal (non-converted) camera. The latter two show that a low wavelength conversion can be switched to a higher one by using external filters.
Bonus - Multi-Spectrum Sample
This lens seemed to like UV a bit better than it did IR. It was more sensitive than the Zeiss 55mm f/1.8 even though it wasn't quite as fast. Here is a test shot of it using the Baader-U UV filter showing some sunscreen. The shoulder stripe had just been applied here, but the sunscreen on the face was applied 3 hours before and been washed just before taking this image. Showing that it does stand up to water somewhat, like it is supposed to.