Sigma's Intriguing New f/1.8 Zoom for APS-C

[fike]

http://www.dpreview.com/news/2013/04/18/Sigma-announces-worlds-first-F1-8-constant-aperture-zoom-lens

Based on the aperture spec alone, this seems really interesting.  If IQ turns out to be good, this lens could be a serious new option for serious APS-C shooters. 

[RFPhotography]

HA!  I was just about to post that.  Too bad it's just for cropped frame cameras.  Although I suspect the design/build logistics and cost to create this for a full frame camera played into it.

[BJL]

As to "why not for 35mm format": size, weight and cost! Upsizing this design to something like 27-52mm [EDIT: corrected from 24mm] constant f/1.8 would involve scaling everything up by 1.5x linear, so about tripling the volume and weight of the lens elements. Instead, this lens is likely designed starting from an existing constant f/2.8 zoom for 35mm format and adding at the back converging elements in the form of a 1.5x focal reducer, which reduces focal lengths, image size and aperture ratios (f-stops) by that factor of 1.5, while keeping the angular Field Of View the same over the smaller image size. This is roughly what Olympus did in the design of its two constant f/2 zoom lenses for Four Thirds; 14-35 and 35-100.

It is interesting that the zoom range is only 2x: I suppose that there are optical quality and size problems with a wider zoom range at such low f-stops, just as constant f/2.8 zooms are limited to narrower zoom ranges than slower zooms, like even f/2.8-4. That is why if I crave speed (and shallow DOF that goes with it), my zoom lens design sweet slot is f/2.8-4 or slower where 4x and 5x zoom ranges are available, and I would rather increase speed by using lens designs of those f-stop ranges upsized to larger focal lengths with a larger format sensor --- if the sensor is still affordable.

Since this lens is doing much the same as adding a 0.66x focal reducer to a 35mm format 27-52mm f/2.8 zoom, it is similar to what is done by the new Metabones Speed Booster, which is a 0.71x focal reducer, but that only works on "mirror-free" cameras like Sony NEX, Fujifilm X and m4/3. So for the mirror-free amongst us, an alternative is:
35mm format 24-70 f/2.8 + Metabones Speed Booster = 17-50mm f/2 covering "APS-C" frame size.

And these could probably be combined:
Sigma 18-35mm f/1.8 + Metabones Speed Booster in m4/3 mount = 13-25mm f/1.3 covering 4/3" frame size.
But I wonder how bad the aberrations get in such an extreme optical combination.

[Vladimirovich]

As to "why not for 35mm format": size, weight and cost! Upsizing this design to something like 24-52mm, constant f/1.8 would involve scaling everything up by 1.5x linear, so about tripling the volume and weight of the lens elements.

first of all their APS-C design is 27-52 FF FOV, not 24 and that matters a lot @ wide end.

then Sigma had excellent 24-60/2.8 FF (24-60 - do not mix w/ their other designs), so going from 24 to 27 and from 60 to 52 even w/ going from 2.8 to 1.8 shall be quite compact.

here is that oldie

[BernardLanguillier]

If the MTF are representative of the actual level of perfo of these lenses, it would seem that Canon and Nikon are starting to get their ass kicked by Sigma big time...

Cheers,
Bernard

[BJL]

First, thanks for catching my slip on minimum focal length: corrected now

... even w/ going from 2.8 to 1.8 shall be quite compact.

Going from 2.8 to 1.8 with the same design is at very least scaling up lengths by a factor of about 1.5, so doubling entrance pupil area and the areas of lens elements. But in fact it is not so easy to scale down aperture ratio: for one thing, aberration control gets far harder as the f-stop gets lower. Can you explain a simple way to increase entrance pupil area by 2.25x while maintaining compactness and controlling aberrations well?

To put it another way: if compact constant f/1.8 zoom lenses for 35mm format are viable, why did none of Canon, Nikon, etc. do it in the film era, when usable ISO speeds were so much lower than with modern sensors, and there was if anything more demand for bright lenses than today?

[Vladimirovich]

First, thanks for catching my slip on minimum focal length: corrected nowGoing from 2.8 to 1.8 with the same design is at very least scaling up lengths by a factor of about 1.5

no need to repeat what was something that nobody was arguing with... look @ FF 24-60/2.8 above, imagine more compact FF 28-52/2.8, multiply by 1.5... that is ok.

PS: EOL retail price of FF 24-60/2.8 was $212 ($212, Cameta Camera) and you can google for images from that lens, there were very good.

, so doubling entrance pupil area and the areas of lens elements. But in fact it is not so easy to scale down aperture ratio: for one thing, aberration control gets far harder as the f-stop gets lower. Can you explain a simple way to increase entrance pupil area by 2.25x while maintaining compactness and controlling aberrations well?

To put it another way: if compact constant f/1.8 zoom lenses for 35mm format are viable, why did none of Canon, Nikon, etc. do it in the film era, when usable ISO speeds were so much lower than with modern sensors, and there was if anything more demand for bright lenses than today?

and can you explain why  Canon, Nikon, etc. did not create that lens that Sigma did for APS-C ?

[BJL]

... look @ FF 24-60/2.8 above, imagine more compact FF 28-52/2.8, multiply by 1.5...

... and can you explain why  Canon, Nikon, etc. did not create that lens that Sigma did for APS-C ?

To repeat, you do not get a f/1.8 design from a f/2.8 design by simply scaling up some dimensions. The only simple way to convert from a f/2.8 design to f/1.8 is to include a 1.5x reducer at the back, which increases bulk significantly.  I believe that Olympus did something similar in designing its pair of f/2 zooms for Four Thirds format.

The only design conversions that can just use a 1.5x linear scaling are between lenses of equal minimum f-stop and similar FOV range.

As to why Canon and Nikon have not done it: maybe because they are focusing their high end products on 35mm format and getting more photographers to step up to 35mm format models like the D6 and D600, so they are letting lens development for EF-S and DX format languish. In recent years, most of the more ambitious new lens models for formats smaller than 35mm have come from companies who are not making 35mm format bodies, and so are more committed to the new smaller formats: Olympus and Panasonic [two f/2.8 zooms and lots of primes], Fujifilm [mid-speed zooms and lots of primes] and now Sigma.

Note that this lens is not much different from what Olympus did some years ago with its 14-35 f/2 for Four Thirds: the minimum f-stop is only 10% different, while Olympus offered a somewhat wider 2.5x zoom range. Sigma's unusual choice of f/1.8, almost exactly 1.5x lower than f/2.8, is a clue that the design is done with a built-in 1.5x focal reducer behind a f/2.8 design for 35mm format.

[Vladimirovich]

To repeat, you do not get a f/1.8 design from a f/2.8 design by simply scaling up some dimensions.

we are not talking about internal design, we are talking about possible dimensions

The only design conversions that can just use a 1.5x linear scaling are between lenses of equal minimum f-stop and similar FOV range.

OK, add 1.5x for that Sigma design to scale... still quite compact.

[Hulyss]

Here we go. Lets talk about facts, then.

http://lcap.tistory.com/entry/Sigma-ART-18-35mm-f18-Preview

[Vieri Bottazzini]

Here we go. Lets talk about facts, then.

http://lcap.tistory.com/entry/Sigma-ART-18-35mm-f18-Preview

Image quality seems pretty good indeed 

To me, besides being a pretty revolutionary lens in itself - it breaks the constant-zoom f2.8 barrier, or f2 for micro 4/3 - bringing finally APS-C Standard zooms to the level of FF as far as DOF control and low-light shooting goes (or better, if one wants to compare things at equal ISO), this lens indicates that Sigma thinks APS-C is there to stay despite all the efforts made by Nikon and Canon to sabotage it and try to convince us to upgrade to FF (mucho bucks for them!).

I wrote my impressions on this lens ON MY BLOG for those of you interested.

Best,

Vieri

[Quentin]

Looks great.  Sigma, kicking sand in the face of Zeiss, Leica and the rest.  Who would have thought?

[stevesanacore]

I've had a few Sigma zoom lenses in the past and although they were innovative designs, they were very poor optically. I'm skeptical but eager to try one and see for myself. I just don't see how Sigma can make a lens at the quality of Canon or Nikon and charge less. Usually top quality means higher prices. How do they justify it? What corners do they cut?

[Malina DZ]

Great kit lens for SD1M. Other brands have better performing primes for serious non-Sigma APC-S shooters. I'm concerned with an effect of heavy vignetting @ f1.8 and multiple aspherical elements on out of focus highlights, especially since it's an Art category lens.

[Quentin]

I've had a few Sigma zoom lenses in the past and although they were innovative designs, they were very poor optically. I'm skeptical but eager to try one and see for myself. I just don't see how Sigma can make a lens at the quality of Canon or Nikon and charge less. Usually top quality means higher prices. How do they justify it? What corners do they cut?

That's the past.  The new Art range is very different.  Their new 35mm f1.4 is optically superb edge to edge, and better than the brand equivalents sold at a higher price.  I know, I use it on a D800E and its the sharpest lens I have on that camera.  I am hopeful they will carry that quality and performance in to Art range zooms.

[BernardLanguillier]

Other brands have better performing primes for serious non-Sigma APC-S shooters. I'm concerned with an effect of heavy vignetting @ f1.8 and multiple aspherical elements on out of focus highlights, especially since it's an Art category lens.

I have to agree with Quentin here. The DPxm series lenses and the new 35mm f1.4 are clear proofs that Sigma engineers now are able to design and manufacture extremely high quality lenses while keeping price very low.

Cheers,
Bernard

[scooby70]

I have to agree with Quentin here. The DPxm series lenses and the new 35mm f1.4 are clear proofs that Sigma engineers now are able to design and manufacture extremely high quality lenses while keeping price very low.

Cheers,
Bernard

Personally I think they began bringing out good lenses long before the recent 35mm f1.4. The ones I own myself that can IMVHO stand comparison with other mass market lenses are the 50mm f1.4, the 85mm f1.4 (even better than the 50mm, IMVHO) and the 150mm f2.8 macro. I also have the 12-24mm and whilst edge performance isn't world class it is at least a full frame 12-24mm.

[fike]

In the past, I have avoided third party lenses like the sigmas, but I got a 15mm f/2.8 fisheye and love it. On APS-C it is excellent and resists flare VERY well.