fbpx
Lenses and Optics

Just the Lenses: Tamron 15-30mm f/2.8

Published February 24, 2015

Our Just the Lenses posts are optical tests where we compare various lenses on the optical bench. Unlike DxO or Imatest test results, no cameras are involved, eliminating one of the major variables. It’s particularly useful when we’re looking at third-party lenses that can be used on various cameras. It’s hard to extrapolate the results of a test made using a third-party lens on a Canon 5D Mk III when you are trying to determine how it might compare to one shot on a Nikon D800, for example. Testing on the optical bench gives a direct comparison between lenses without any other variables.

We’ve been quite interested in the new Tamron 15-30mm f/2.8 Vi DC USD, and since we already had optical bench results for two very similar brand name lenses (the Canon 16-35mm f/2.8 L Mk II and the Nikon 14-24mm f/2.8 AF-S) we thought it would make a good candidate for a Just the Lenses test.

 

Roger Cicala, Lensrentals.com, 2015

Tale of the Tape

The Tamron 15-30mm f/2.8 feels heavy and solidly built. Like the Nikon, its hood is built-in, which has advantages and disadvantages. Since it’s always on, it’s always providing protection, and it’s built more solidly than a simple hood. But if you lose the lens cap you can’t just buy a temporary replacement at the local camera store; you have to buy the special cap that fits over the hood. Not to mention if you break one of the petals it requires a trip back to factory service to be replaced.

Like the lens itself, the focus and zoom rings feel very solid. There’s no loose sloppiness in this lens, unlike some other third-party wide angles. But some may find the resistance a bit higher than they’re used to; there certainly is more resistance when zooming or focusing than with either the Nikon or Canon wide-angle zooms. Some will like it better. Some won’t.

Other factoids about the three lenses are listed below.

  Nikon 14-24 f/2.8 Tamron 15-30 f/2.8 Canon 16-35 f/2.8
Price$1,797$1,199$1,699
Weight (lb)2.22.41.4
Length (In.)5.25.74.4
Filter threadnonenone82mm
Aperture blades997

In size and weight, the Tamron is much closer to the Nikon than the smaller Canon lens. Like the Nikon, it can’t take screw-in filters. It’s significantly less expensive than either of the brand-name lenses, although I wouldn’t call it bargain priced. My opinion going into the test, therefore, was the Tamron would have to provide optical performance at least as good as the Canon lens (considered the weaker of the two manufacturer’s wide zooms) to be competitive.

Optical Bench Results

All lenses were tested on a Trioptics Imagemaster Vertical MTF Station. Five copies of each lens were tested and the results averaged unless noted otherwise. Each copy was tested in four rotations with 0 degrees being standard camera orientation, the cut being made from the left edge to the right edge of the field of view. The additional cuts were made after 45, 90, and 135 degrees of rotation on the bench.

Testing each copy at 4 rotation angles gives us an excellent way to tell how even each side and corner is.

 

Results at the Wide End

The graphs below represent the average of 20 MTF readings for each lens (5 copies of each lens, with 4 MTF readings each) at the widest end of the zoom range, comparing the Tamron to the Nikon and Canon zooms. Note that the readings are actually at slightly different focal lengths: 14mm for the Nikon, 15mm for the Tamron, and 16mm for the Canon.

If you don’t speak MTF, don’t worry. It’s not hard. Higher on the vertical axis is better. Dotted and solid lines of the same color close together are better (far apart is astigmatism). The horizontal axis goes from the center of the lens at “0” to the edges of the lens at “-20” and “+20”. Lower lp/mm (black and red lines) have association with strong contrast, while higher lp/mm are associated with ability to resolve fine detail.

 

Comparison of Tamron 15-30 and Canon 16-35 Mk II MTF at widest focal length.

 

Looking in the center 1/3 of the image (0 to 6mm from center) the Tamron doesn’t resolve quite as well as the Canon, although it is still very good and very close. In the middle 1/3 of the image (7 to 15mm from center) the Tamron is actually a bit better. In the outer 1/3 of the image, though the Canon is superior. There’s not a clear-cut winner here. I won’t call it a tie because they are different. A landscape shooter for whom the corners are critical might prefer the Canon. A wedding shooter needing the middle of the image for group shots might like the Tamron better.

 

Comparison of Tamron 15-30 and Nikon 14-24 at widest focal length.

 

Compared to the Nikon 14-24 f/2.8 AF-S, the Tamron is just a tiny bit behind in center resolution, and about equal or a little better in the central 1/3 of the image. The Nikon is much better in the edges and corners in the sagittal plane, which it’s famous for, and roughly the same in the tangential plane. Technically, the Nikon has more astigmatism in the corners, but that’s because of the increased sagittal resolution, the tangential resolution is about equal in the two lenses.

Overall at the wide end the Tamron does well. It’s perhaps not quite as good as the two manufacturer’s zooms, but very close.

Results at the Long End

There’s a greater difference in focal length here, with the Nikon at 24mm, the Tamron at 30mm, and the Canon at 35mm. So some might say we’re comparing apples and oranges. On the other hand, the extreme ends of the focal range are what we like to test since it’s where any weaknesses usually show up.

 

Comparison of Tamron 15-30 and Canon 16-35 Mk II MTF at longest focal length.

 

At the long end, it’s apparent that the Canon is just a bit better than the Tamron both in the center and off-axis.

 

Comparison of Tamron 15-30 and Nikon 14-24mm MTF at longest focal length.

 

Again, the comparison shows that the Tamron just isn’t quite as good as the Nikon at the long end.

23mm Comparison

Let’s remember, though, that the long end of the three lenses is quite different. For that reason I thought it worth comparing them all at about the same focal length, so we ran another set at 23mm to give a direct comparison at that focal length. This might give a slight disadvantage to the Nikon, since this is near its extreme focal range. On the other hand, the Nikon lens designers got a slight advantage because they didn’t design the lens to zoom as far as the others. So I’ll call it all even.

And I’m glad we did check, because the information is worthwhile, demonstrating a different philosophy of lens design.

Comparison of Tamron 15-30 and Canon 16-35 Mk II MTF at 23mm.

 

The Tamron is at its best here in the middle of the zoom range, with a better MTF curve than at either extreme. The Canon, on the other hand, appears to be at its weakest here; not as good as it is at either extreme and not as good as the Tamron at 23mm. (Remember, this is the average of 20 MTF readings from 5 copies, and there were no bad copies.)

 

Comparison of Tamron 15-30 and Nikon 14-24 MTF at 23mm.

 

Compared to the Nikon, the Tamron is holding its own when shot at 23mm with an MTF curve that’s as good as the Nikon’s, and perhaps a bit better since it has less astigmatism. That’s pretty remarkable.

Field Curvatures

Field curvatures give us a lot of nice information, explaining sometimes why corners are softer (because they are not in the plane of focus of the center of the lens used for the MTF graphs), and demonstrating visually the lens’ astigmatism. I’m just showing the curves for all three lenses at 23mm. The curves are similar in shape and direction at both ends of the zoom. In the case of the Canon lens, though, they are much less severe at areas other than the mid-point of the zoom.

 

Looking at the graphs above (all done at 24mm) we can see, for example, part of the reason for Nikon’s superb corner and edge performance in the sagittal plane. The field curvature has a little “mustache” shape, but really stays in the same plane of focus as the center all the way out to the edges. The tangential field has a pretty wicked “U” curve, which also explains some of the astigmatism we see on the Nikon graphs.

The Tamron’s field curvature at this focal length is much like the Nikon’s in the sagittal plane, and has much less curve in the tangential plane. Again, this shows the ‘sweet spot’ of the Tamron lens. At the wider end there is more astigmatism and at the longer end some loss of resolution.

The Canon’s field curvature is more severe than the others although the sagittal and tangential curves are similar in shape through most of the field. (Please forgive the broader “depth of field setting” on the Canon curves. I accidentally moved a slider when generating these and they take so long to do that I decided to go with it as it was rather than redoing it. Tying up a $200,000 machine for an hour when we need it for other things isn’t always possible.)

Summary

Like so many tests, this one just emphasizes the fact that different lenses are different.

At the wide end of the zoom range, the optical differences are complex. In the middle 2/3 of the field the Tamron holds its own or is a bit better, but it’s not quite as good in the corners. At the long end (30-35mm) the Canon lens is better than the Tamron. At 23mm, though, the Tamron is as good as the legendary Nikon 14-24, and clearly better than the Canon 16-35. And, of course, the Tamron comes at a significantly lower price than either of the manufacturer’s lenses.

So which one do you want? As is usually the case, it’s an individual decision. If you have a good 24-70 zoom already and plan on using this lens from the wide end to 24mm, then the Tamron is right up there with the other two at 24mm and just a bit behind at the wide end. For Canon shooters who want to their wide zoom at focal lengths longer than 24mm, the Canon is better and goes further.  The Nikon doesn’t go longer than 24mm, so the Tamron gives some extra reach to Nikon shooters, although at slightly lower resolution.

Choices are nice, aren’t they? But deciding can be complicated.

Roger Cicala and Aaron Closz

Lensrentals.com

February, 2015

 

Author: Roger Cicala

I’m Roger and I am the founder of Lensrentals.com. 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 Lenses and Optics
  • Roger Cicala

    Bruce, from a sharpness standpoint I’m comfortable that it would. My tests don’t check for distortion and that would be critical, so I’d definitely check some review sites concerning that.

    Roger

  • Bruce Wilson

    Hi Roger,
    Many thanks for the review with lenses only.I don’t want to what the Tamron is like on Canon when I wold be mounting it on a Nikon D 810. I like the VC and reduced flare compaired with the Nikon. Would the Tamron have high enough image quality to use architecture and landscapes for magazines or, would I be better off with the Nikon?
    I would appreciate your thoughts.

  • Roger Cicala

    My apologies, Stefan. I thought you were referring to the optical bench article.

    Roger

  • Roger Cicala: My comments was for Niels comparison. I have no objections to yours. All seems to be in order. Cant either comment the others about if the MRT-curves are correctly named.

  • Stephan,

    DXO lens modules: 14-24 was the only supported lens. Since DXO hasn’t added a module for the Zeiss 15mm, it’s safe they never will. Tamron is very new, and odds are a module will be added … but when?

    DXO PP: Std. DXO corrections, minus any ClearView processing. IF I did this part right, I would stick w. that procedure, except set the exposure to ‘as shot’. Or maybe I would tweak the sliders to get the best picture that I know how, since that’s what I would be doing in real shooting conditions.

    Lighting conditions & Time of day: Shot w. one, then another, and then the other, hitting my key composition pts & a few other. BETTER WAY: Rent one D7100 per lens, minus my camera. Also rent (or buy) an A/S-compatible QR plate, so I could swap the cameras out to the tripod. (This minimizes composition variation problems, as a bonus.) But even on a cloud-less day, this is not perfect, because a few minutes can cause enough of a shadow shift.

    Composition: My biggest regret, which was also thwarted by the Tamron not starting at 15mm. Better than 1 camera/lens mentioned above –> Use one of the candidate lenses, and get at least one pic for each test composition, then upload to the mobile device of your choice (e.g. iPad, iPhone, etc., etc.). This will give you something to match when testing the other lenses. Why is this important? Once you pick a lens to buy, you will be looking for a good copy. You will NOT be getting the lighting conditions that you had before, but at least you can match the composition.

    Preconceptions at the start: A) Nikon 14-24 would suffer a quick death because of flare issues. Reality: Issue was over-hyped. Flare bad, but within my tolerance. B) Zeiss 15mm would be sharpness king, but would lose because I want the flexibility of a zoom. C) Had a bad copy of the Tamron the previous week, so this one would be edged out barely in image IQ by the Zeiss, but would still best against flare.

    Bottom Line: No the Tamron-Zeiss-Nikon test wasn’t scientific, but got in enough test shots to be confident enough for a purchase.

    Since I did my testing, bought the Nikon & pitted against my walk-around lens; Nikon 17-55 f2.8. Other than looking at center areas, cannot believe hot pitiful the 17-55 looks.

  • Roger Cicala

    Stefan, I never do this, but really, you should read the article before you comment.

  • Thanks for sharing. I don’t Think your tests really are scientificall. First, you Lightning conditions are different since the sun sometimes shine and sometimes don’t. Secondly you converted using DxO which do you automated funcions to render it and as you Point out, Nikon has been applied lens corrections. I have used DxO and it uses alot of algorithms to enhnance sharpness where it lacks sharpness and do this in different degrees all over the image. Likewise CA, distorsion and vinjetting.
    You should have shot when the lighting consitions were the same at all time otherwise its hard to test them head to head as you did. And the RAW converter should have been totally reseted in all parameters.

    /Stefan

  • Robin Perkins

    Any testing or opinions on the Tamron’s “Coma” Big factor for my nightscapes.

  • Hi Stephan,

    http://graphitepaddle.smugmug.com/Photography/Super-wide-lens-testing/

    Except for the 1st two compositions, the others had focus points close up. Favoring the Nikon was based on 100% of points in the upper third of the composition. Guess what? Looking at other areas of the composition, it was closer or had the Zeiss 15mm, or Tamron coming out on top. Focusing was always manual. (When I switched the Tamron & Nikon to AF, did see much of a difference.)

    OT: Flare results went from being close (but clearly favoring the Tamron), to the Nikon being bad, but not as bad as the reviews on it lead me to believe.

    Bottom Line: Not disparaging Roger’s testing, but think the ultimate test for any buyer is to test these lenses on THEIR camera, on compositions THEY expect to see.

  • Niels

    Hi Stephan: Will post either Fri or Sat night. Of note, didn’t look at center area of the lens; when breaking the composition into a 3×3, I was looking at the outer squares & at f11. Can post the f2.8, f5.6, f8, & f16 results if interested.

  • Niels: Could you post the comparison images? So far you are the only one I’ve seen this kind of “negative” Review.

  • Niels

    I was one of those on the Tamron 15-30 band wagon, until I rented it + Nikon 14-24, Zeiss 15, and Tokina 11-20; then took pics. The Nikon, which I was biased against because of: A) Action of the zoom & MF rings, B) User reviews of flare issues, and C) Lensrentals Repair Data 201-13.

    Emphasis on apertures f8 & f11, focal length 15mm. Zeiss & Tamron could close to the Nikon sharpness, sometimes, but the 14-24 was always on top. Flare control was comparable among these 3 … Tamron significantly better than the other 2, with the Tokina being pathetic (like the 11-16 & 12-28).

    1st Zeiss lens that I didn’t like, out of 3.

    Tamron was fantastic w. flare, nice fluid MF & zoom rings, but doesn’t hold a candle to the Nikon in image quality. Also didn’t like the 15mm being closer to 16mm; not much of FOV gain from someone w. 17-55.

    Bad copy of the Tamron? Then it happened twice: Rented it last week, and the Lensrentals receipt shows 2 different S/N’s.

  • John H

    It’s funny and ironic how the inter-webs work. It’s been the gospel truth for the longest time that ultra-wides are highly utilitarian, but far from what some would call “prime lens” perfect. Yet many still post as if $1200 and VR should translate to perfection.

    If nothing else, your report shows this to be yet another “more than good enough” lens for most casual shooters, and not at all bad for the discriminating pixel-peeper. The VR alone makes it a no-brainer for many. I happen to be one of the lucky few with a perfectly suitable Tokina 16-28 and the lens isn’t decentering right in front of my eyes. I use it when needed for travel documentary and it’s sweet. Paid $500 for it barely used, with the original box and materials.

    Were someone to pick this or any of these UW lenses up and just go shoot with them (rather than study charts) they’d be happy. Ignorance IS bliss.

  • Eric Tung

    Hello, Roger. I have looked through the test and I found maybe the MTF result of Nikon 14-24F2.8 at 14mm F2.8 is wrong? Because the mtf chart is totally different from another test you did in “Canon Wide-Angle Zoom Comparison”
    http://wordpress.lensrentals.com/2014/07/canon-wide-angle-zoom-comparison
    when you compare canon 16-35F4 at 16mm vs Nikon 14-24F2.8 at 14mm.

  • Lars

    Thank you for all your wonderful comparisons!

    Would you please check the mtf curves for the Nikon 14-24? They seem to differ from the ones presented at the end of the Canon wide angle zoom comparison.

    Lars

  • Andrew

    Ah I see well thanks for the follow up. Can’t imagine the amount of time it takes to conduct these tests, but I really appreciate that you do them!

    Looking forward to seeing further findings 🙂

  • Roger Cicala

    Andrew and others, I see what you are saying. I spent some time today trying to track this down and it may be that our database has a mislabeled lens: we pulled the comparisons from data gathered before I was out sick and while I was gone it was refiled. It’s possible during the refiling it was mislabelled. We’re going to re run a set of 16-35 f/2.8 lenses to make certain but weren’t able to do that today. I have to leave for WPPI Saturday and want to do this recheck myself, so for now consider that the 16-35 results at 16 and 35 might be wrong. The 23mm I know are correct because I ran them for this test.

    Roger

  • Andrew

    Amazing test very very thorough! It appears there may be some mistake though… please see your previous test of the Canon 16-35mm f/4L vs Canon 16-35 f/2.8L II.

    http://wordpress.lensrentals.com/2014/07/canon-wide-angle-zoom-comparison

    The Canon f/2.8L II test charts from this article look identical to the Canon f/4L IS charts from the Canon Wide Angle Zoom Comparison article.

    And the Canon f/2.8L II @ f/4 test charts from that previous article look much worse than the Canon f/2.8L II @ f/2.8 in this article.

    Please correct me if I’m wrong… but it seems like perhaps the charts in this article are actually the Canon f/4L IS?

  • I’m more interested in distortion figures than sharpness, it is more relevant to my line of work. Do you not test for that? Thanks,

  • David

    > I’ve seen tens of thousands of images from various copies of the Canon 16-35. In short, they all sucked.

    I’ll construe this narrowly to mean you aggregated the images of terrible photographers. If you’re talking about the lens, then I’m impressed to see a troll on this site.

  • My thoughts exactly Matthew! Having a couple of other more than capable lenses for low-light and astro I see no need for this heavy and imho relatively expensive lens. Others that shoot a lot of stars might think otherwise, but hey they already own at least one astro-dedicated lens right?

  • All in all, I must say I’m kinda disappointed. Not because (as most others would argue) “if Tamron wants to charge $1200 for this lens, it’d better be flawless!” …but instead, because they thought they could get away with making a freaking 2.5 lb lens. Simply put, if you want to make a lens HEAVIER than either the Nikon 14-24 or the Tokina 16-28, it had better be flawlessly sharp.

    Different priorities, different priorities.

    If coma and vignetting are lower than on the Nikon 14-24, though, I’d consider it for astro-landscape photography. It would require simultaneously owning a crop-sensor kit with something lighter like the new Tokina 11-20, though, for lighter weight travel work…

  • @Matt, in short, YES. The Canon 16-35 will eventually become de-centered, unless you baby the crap out of it. And even then, no promises. I don’t have the same volume access to lenses as Roger does, however I work in post-production so I’ve seen tens of thousands of images from various copies of the Canon 16-35. In short, they all sucked.

    The Tokina, from the images I’ve seen, probably has an equal or greater chance of being DOA, (decentered on arrival) …or getting decentered with moderate use.

    Personally, as someone who spends days on end blasting down dirt roads in the desert, needless to say NEITHER lens is acceptable lol. The 14-24 has been the ONLY lens I can trust to stay sharp, if you get a sharp copy, which is also far more likely compared to the others.

  • Roger Cicala

    Bartosz, the Canon versions are Mk II, I’ve added that in the text. I didn’t think about it because we haven’t had Mk I 16-35s for many years.

  • Mike

    * my samples

  • Mike

    Thanks again, Roger, for the great service and valuable information you provide to the photographic community. I rented one of these from you and took some sample shots this past weekend. I don’t own anything wider than 24mm so it’s a type of photography that I’m not really that familiar with. I was thoroughly impressed with the quality, handling, and operation of this lens. The fact that it compares favorably in image quality to lenses costing 40-50% more makes me think it is going to be very well received. Add to that the fact that it has image stabilization and, in my opinion, you have a real winner. It seems that many people feel that IS (VC) in a wide angle is not that useful. My testing was very limited but the benefit when hand holding at 15mm and 1/10 second was remarkable. See images 99 and 100 of may samples at http://mijz.com/Tamron_15-30. All of these images were taken with my Canon 6D.

  • I guess there is few mistakes in this comparison:
    1) MTF chart @ wide and tele looks like from Canon 16-35 L IS f/4 not 16-35 L f/2.8 but chart @23mm is probably from Canon 16-35L f/2.8
    2) Is it Canon version II (photo) or mark I (text)?
    3) Are you sure that Nikkor is better then Tamron @ wide end? If I read these two charts with no mistake Tamron results looks better.

  • Kuba
  • Pawel

    Canon 16-35/2.8 II is very useful for street or reporter shootings. It offers an excellent center quality full opened and delivers a very good performance for wide settings (16-25mm) from f/4 onward. The amount of vignetting at f/2.8 can be critical for some purposes. One has to stop it down to f/5.6 for landscapes, when a corner to corner image quality is needed, but most landscapes are took at f/5.6 – f/8 anyway. Considering that, the field quality is far more better than the technical performance.

    Personally I found Nikon 14-24 too heavy for a fine and effective work and too prone to flare (no autofocus on 5D III and 6D of course). The image quality is perfect at 21mm, but at 14mm I would highly recommend Canon 14/2.8 II which is the unsung hero here, so well controlled and as good as it gets, but also overpriced.

    Zeiss 21/2.8 Ze is a stellar performer at f/2.8 and onward, except the amount of vignietting at f/2.8 – f/4. The lateral CAs are extremely low, so one can see the image resolution even more fine than it is. The distortion is well controlled but not so easy to correct with PhotoShop due to the secondary mustache distortion curve. The lens is perfect for a low light landscape photography and for a quiet, reflective or contemplative style of photography.

  • Cat

    Does it mean, all three are worse then Canon 4,0/16-35 mm L?
    The Tamron and the Nikon with their curved glass are heavy load and very sensible for dust
    and difference between 15 and 16 mm is not the world.
    And are there so many night shooters that need f:2,8?

Follow on Feedly