New Tamron 35mm

Tamron have a new 35mm f/1.4 in development. Slightly off-topic, I know, but the F-mount version is coming first (end of June, apparently). DPReview seem to have pre-released the press release, since it's now disappeared again. I assume we can link to it when it reappears.

Just thought I'd mention it because Tamron seem very proud of it, and it'll be interesting to see how it goes up against the Sigma 40mm - possibly interesting enough to make me hesitate on doing the Sigma 35+50mm Art to 40mm trade-in (I'd really rather have the 35mm and 50mm options).

I'd like to see it with VC like the f/1.8 version.

That'd be interesting, but it doesn't sound like it.

The tab I still have open says 14 elements in 10 groups, four LD (low dispersion) and three GM (aspherical) elements, and £929.99 at the end of June. There's a lens diagram which I'm sure we'll all be able to see when the actual embargo on the press release lifts. (The lens itself was announced back in February, I believe, this is just an update on information.)

I do see 35mm as quite a bit wider than 50mm, particularly for group shots. I have the 35/50/85mm Sigma Art collection, but the former two are optically... well, a lot better than the Nikkors (and I have some 50mm Nikkors), but way behind reports of what the 40mm can do. Sigma might want to think about pre-announcing mk2 versions of the 35 and 50mm if they have those in their roadmap. The 35mm and 50mm combo are nice for low light group and individual shots (and the 85mm), so the 40mm certainly compromises my flexibility if it's replacing two nearby focal lengths. Between that and the 105mm f/1.4, I'm being pulled away from my traditional 35/50/85/135 sequence.

I'll be interested to see what this does.

Including VC would make it a killer lens. As it is now, I prefer the f/1.8 Tamron in this FL over the f/1.4 Sigma(although both are excellent). The Sigma is no doubt sharper at f/1.8 and probably about equal to the Tamron when both are wide open, but it's also unforgiving wide open and I'd rather have VC(VR) than an extra half a stop for the types of situations where I use these lenses(although lately I've been grabbing the 35mm f/1.4 AI-s in preference to all, although its character is very different fron either of the others). I'll also add the caveat that I own the Tamron, and have not owned the Sigma.

As I see it, the Sigma is better built and also focuses better. Thom Hogan describes the Tamron f/1.8 as "sliding" when focusing, and as strange as that sounds it's a very apt description in my experience. It's not as "decisive" as a Nikon AF-S or the Sigma, but rather seems to creep up on the focus point.

In my mind, the Sigma is the lens to beat at the FL/aperture, so Tamron needs to bring SOMETHING to the table if they want to compete. That "something" could be either better optics at the same price point, or equivalent optics for less(perhaps with the build quality of the f/1.8). It will be interesting to see what happens.

As I've often said too, it's a bit alarming that Nikkors really are not even under consideration these days for fast normal primes. You'll see that a lot of long-time Nikon shooters are turning to the ART lenses and others rather than the expensive and not as good Nikkors. When I bought my Tamron, I had it(used) along with a used ART and new 35mm f/1.8 Nikkor AF-S. The Nikkor and Sigma were about the same price(taking into account, again, that the Sigma was used and Nikkor new), while the Tamron was the least expensive-I think it's about the same price as the Nikkor new. Ultimately, I was ready to spend the ~$600 for the ART or Nikkor, but the Tamron won because it made the most sense for me. I never thought I'd pass on a Nikkor and consider a Sigma or Tamron instead, but that's where I was.

I´m not that fan of VR, just a lot of additional parts which can break, for short lenses I dont need it, for longer I bring the monopod.

This said, the Sigma Art 35 will be hard to beat.

Including VC would make it a killer lens

Maybe they're going to make a Z-Mount version, in which case lens VR is less important. IBIS can do the heavy lifting.

Mind you, if it did, the combo of lens VR and IBIS in a 35mm lens would make handheld shots of very slow or stationary subjects easier.

When I traded my 35mm and 50mm ART for a 40mm ART, I kept my 24-35mm f2 ART, so as to cover the wide end.

They are all big, heavy lenses, but I can cope with them, so far!

Huh. Still no sign of the announcement on dpreview. The diagram didn't contain anything that looked like a VC unit, I'm afraid.

I don't actively want a heavy lens, but generally it's a minor decision for me compared with the quality of the optics. At 35mm, I'm probably more worried about my subject moving than hand shake, most of the time.

There were very mixed reviews of Tamron's 35mm f/1.8 VC lens, otherwise I might have gone for it. It was also a bit overpriced IMO, for such a modest specification. And I find it hard to fault my AF-S 35mm DX Nikkor with the same aperture at a lower price.

Still, new lenses are always interesting. Let's wait and see what the price/performance is like. I suspect Tamron will have their work cut out to compete with Sigma's well-established Art lens. Plus, if you can do without AF, Samyang's mighty beast of a 35mm f/1.4 delivers lots of performance for not much money.

The announcement did say nice things about the expected bokeh, for what it's worth.

That, in my experience, has always been a weakness of fast 35mms going back to the manual focus days.

Even the old Nikkor is a bit "busy" in this department.

Okay but that 40mm Sigma costs $1150, weighs 2.6 pounds and corrects every pixel like the soulless technocrat you expect from a recent Sigma, so it’s not really a lens that everything has to be compared to.

A quick search at one of the obvious stores in the UK lists the Sigma 40mm at £1099. The RRP for the Tamron is £930. The Sigma has 16 elements in 12 groups, 3 FLD and 3 SLD (anomalous dispersion). The Tamron is 14 elements in 10 groups. For comparison, a 55mm Otus is 12 elements in 10 groups and the Sigma 35mm Art is 13 elements in 11 groups. The Tamron isn't quite as expensive (or complex) as the Sigma 40mm (although it's more expensive than the 35mm), but it's in the same ball park, and I'm not holding my breath for it to be tiny. (The 35mm Summilux is half the weight of the 35mm Sigma, but it's also more than five times the price.)

The Nikkor version, meanwhile, is upwards of £1500, 10 elements in 7 groups, one aspheric, and no anomalous dispersion elements (and, I hadn't realised, only slightly lighter than the 35mm Sigma). Technology has moved on, unlike the pricing, and it shows in some tests.

The Sigma 40mm ART could be got for £750 grey when i switched 6 months ago.

and corrects every pixel like the soulless technocrat you expect from a recent Sigma

Once it's truly sharp, you can change pretty much everything you want in post.

The opposite isn't true.

Retrospective sharpening does tend to sharpen things you don''t want to have sharp edges (like bokeh). Some Sigma lenses seem to have a bit of a bokeh fringe on them, especially in the foreground, and since that's a 3D effect it's quite hard to correct in post; I get that sometimes the effect is a little intrusive. You could argue that imperfect sharpness can give a depth of field that's not a simple plane - there's a region of space of relatively "constant sharpness", which means you don't start seeing bits of the image that are supposed to be "in focus" losing sharpness relative to the actual focal plane if they're slightly separated from it. Whether that's beneficial is questionable, but again it's hard to fix in post-processing, and I wonder how much it factors into discussions about the "transition region". While I'm not sure it's possible to make a lens truly apochromatic (unless symmetry helps?) I do get the impression that lenses tend to correct the focal plane for multiple frequencies without necessarily correcting the cone angles (effective aperture), which means you're still stuck with some colour fringing on out-of-focus regions.

Lens design is complex. I'd kind of like to try to design something with the features I'd like (smooth bokeh, minimal LoCA, controlled coma, no mechanical vignetting, absolute sharpness not so vital) and get it manufactured, but I suspect such things are never going to be cheap.

You could argue that imperfect sharpness can give a depth of field that's not a simple plane - there's a region of space of relatively "constant sharpness", which means you don't start seeing bits of the image that are supposed to be "in focus" losing sharpness relative to the actual focal plane if they're slightly separated from it.

That doesn't have to be argued, it's a recognised side-effect of residual spherical aberration in optical text books.

What happens is that SA prevents image points coming to a single plane, and instead spreads the least-width ray bundle longitudinally along the lens axis. With the effect of artificially increasing depth-of-focus and hence depth-of-field.

A lens that totally corrects SA would therefore appear to give a shallower DoF for a given aperture. While the converse is true for lenses with some residual SA.

This effect also accounts for why large aperture lenses, which almost invariably have some residual SA at full aperture, can be quite difficult to focus accurately. The axial spread of 'best focus' fools both the eye and AF systems. So instead of 'snapping' in and out of focus as the myth of wide aperture lenses would have us believe; they tend to drift from a highest-contrast 'focus' to a smallest diameter point-spread 'focus', and the two planes aren't co-incidental.

:-) Sorry, I meant you could argue that it's beneficial (or not, if you want peak sharpness). Although I should read some more optical text books!

It's not something I've seen software ray-tracers try to emulate, although it shouldn't be too hard. Doing it while retaining optimal sharpness (introducing a nonlinearity into the ray) would be an interesting hack.

Ah, we're back:

Tamron says its new SP 35mm F1.4 is 'its finest lens ever'

"The lens will be available in Nikon F mount first at the end of June, and then for Canon EF at the end of July. It will cost $899 / £929.99" **

That's shameful.. $ : £ still 1:1.3! Therefore $899 should = £691..... WTF!

Don't call us, we'll call you...:(

Not.

PS That's £180 more than grey Sigma 40mm..... it better be good!

20% VAT, Mike. Also, as someone trying to afford a conference ticket in the US, I wish it was 1:1.3. Oh for the happy days of 2:1 ratio when I got my 14-24.

I imagine it'll be less in shops. The question is going to be how well Tamron's "finest lens" stacks up in tests. They've made some good products, but marketing departments do like to maximise how critical everyone is likely to be.

Officially, it IS 1:1.27. as of 07.08 this morning, I just looked........ £1 gets you $1.27.

Now that's a fact but sadly has nothing to do with how much people want to charge for things...!

The link below shows what competition it's up against...and why i did the deal!

Sigma A 40 mm f/1.4 DG HSM review - Image resolution - LensTip.com

Anyone wondering about the Nikon AFS 35mm 1.4 should go here .. briefly.

Nikon Nikkor AF-S 35 mm f/1.4G review - Image resolution - LensTip.com

Yes. (Sorry, being about to buy an expensive thing in dollars, that there was a recent slide down from $1.30 annoys me, even if it's not by much. Best I don't dwell on it, because this isn't the place to talk Brexit.)

Resolution isn't everything (as with the design of the DC lenses). However, it's definitely something, and I'm not going to claim magical capabilities from the Nikkor. To my mind, the 35mm Sigma is substantially weaker (especially at the edges of the frame, which I care about for a wider lens), and the 40mm has the "disadvantage" that it's not a 35mm - some like ~42mm as a normal focal length, but since I have the 14-24 (Nikkor) and the 85mm Art, I prefer the 35mm/50mm pairing to fill the gap. 40mm isn't the end of the world, but unless I wheel out my much older 50mm glass, a single fast(er than f/2.8) lens in the sub-85mm range has a lot to fill if I do the trade-in. I imagine that those who have, say, a 19mm/28mm/58mm/105mm/150mm set rather than 24/35/50/85/135/200 (not that I have a 135 any more) may feel it fits better, but the 58mm Nikkor is a bit specialist. (Actually, now I think about it, I do have a fast 58mm, but it's a Petzval...)

If the Tamron is in the same ball park as the 40mm and better than the 35mm Art, it may have some appeal to those with the same concern - although it won't make the 50mm Sigma Art any better. (Remember when we used to be really impressed at how close to the 55mm Otus the 50mm had got? Being an optical engineer is thankless.) If Sigma are developing updated 35mm and 50mm lenses with the optical quality of the 40mm, they might benefit from getting a roadmap out there before Tamron steal some thunder.

Or the Tamron may turn out to be terrible when tested. :-) We'll have to see.

Resolution isn't everything

Indeed not, but what seems to kill high resolution (esp at the frame edges) is naughty things like Coma, Astigmatism and Chromatic Aberrations, which AKAIK, no-one actually wants, however artistic they want the final image.... although I suppose uncorrected LOCA is used sometimes.

Indeed not, but what seems to kill high resolution (esp at the frame edges) is naughty things like Coma, Astigmatism and Chromatic Aberrations, which AKAIK, no-one actually wants, however artistic they want the final image.... although I suppose uncorrected LOCA is used sometimes.

Absolutely. But things like bokeh rendering and behaviour in the transition zone also have an effect, and arguably can be visible at a larger scale than resolution. (So can LoCA, which is why it annoys me, but most Art lenses aren't too bad on that. I still have an 85mm Sigma shot with significant LoCA in it, so they're not perfect.)

For example, here's the Sigma 40mm chromatic aberration test from lenstip. To my eye, the detail farther from the focal plane than the camera is nicely blurred, but the detail closer to the camera (admittedly this should be rare in most scenes) seems to have an outline, which I find a little distracting. Some other Art lenses have had similar behaviour. For a lens with well-corrected LoCA and very high resolution at the focal plane, I'm a bit confused at the light paths which cause this. I vaguely expect standard LoCA to be caused by different wavelengths having different focal planes (and because the light cone is offset at different frequencies, with the tip of the cone closer to the lens for blue/red than green, you get green outlining of subjects more distant from the camera and purple outlining nearer the camera). There's a secondary effect that even if all the wavelengths focus at the same focal plane, you could have a different effective aperture for different wavelengths, which would give you symmetrical colour fringing either side of the focal plane. Having minimal LoCA but still having asymmetrical behaviour either side of the focal plane (and apparently fairly smooth bokeh) confuses me. But I'm so not an optical physicist. :-)

For at least a couple decades now, British list prices in Pounds have tended to very close to the US list price in Dollars.

A big factor, is, as noted, the VAT.

20% VAT

Another is, I suspect, a somewhat more competitive market in the USA.