Previous page Next page

Sigma 18-35mm F1.8 DC HSM Review

September 2013 | By Andy Westlake, Richard Butler
Buy on Amazon.com From $799.00

Review based on a production Sigma 18-35mm F1.8 DC HSM

Sigma has a long history as a lens maker, having been founded over 50 years ago. In the film era it was best known for relatively inexpensive lenses that undercut the camera makers' own equivalents in terms of price. But this has changed over the part decade or so; while other companies have shifted manufacturing to cheaper locations such as China and Thailand, Sigma has stubbornly refused to move from its factory in Aizu, Japan. This means it can no longer compete in the same way on price alone, and it's therefore switched its focus towards higher-value offerings.

Over the past few years we've seen increasingly ambitious concepts appear from the company's design studios. The original (and recently-replaced) 30mm F1.4 EX DC HSM has long been one of our favourite lenses for APS-C SLRs, and the 50mm F1.4 EX DG HSM grabbed our attention back in 2008 due to its sharpness at large apertures. Most recently the 35mm F1.4 DG HSM impressed us with its exceptional optical quality at a very competitive price. This all bodes well for the company's latest offering - the record-breaking 18-35mm F1.8 DC HSM, which is the first constant F1.8 SLR zoom lens to hit the market.

Sigma's choice of F1.8 as maximum aperture isn't a coincidence; it means that the lens will offer the same control over depth of field as an F2.8 zoom does on full frame. What's more, it will also offer effectively the same light-gathering capability as an F2.8 lens on full frame. By this we mean that it will be able to project an image that's just over twice as bright onto a sensor that's slightly less than half the area, meaning the same total amount of light is used to capture the image. This is important as it's a major determinant of image quality. Essentially it means that APS-C shooters will be able to use lower ISOs when shooting wide open in low light and get similar levels of image noise, substantially negating one of the key advantages of switching to full frame.

As we'd expect at this level, the lens uses an ultrasonic autofocus motor for fast, silent focusing. It's compatible with Sigma's new USB dock which allows you to fine-tune autofocus behaviour in much more detail than the AF microadjust corrections found on SLRs, which should help get the best possible focus accuracy and make the most of the large aperture. It also incorporates several of the thoughtful design touches that we were impressed by on the 35mm F1.4, including an improved AF switch, and a large grip area on the base of the barrel for better handling.

The lens's 27-53mm equivalent focal length range is obviously a little limited, but should still be rather useful for such applications as wedding and events photography. So while it may not quite match the capabilities of a 24-70mm F2.8 on a full frame SLR, for existing APS-C users it should offer something very close. Crucially, at a street price of around $800 / £650 at the time of writing, for existing APS-C shooters it's an awful lot cheaper than buying a 24-70mm F2.8 and a full frame SLR to go with it.

Overall the 18-35mm F1.8 is a really intriguing product, and we applaud Sigma for pushing the boundaries of lens design ahead of the more conservative camera manufacturers. But can an F1.8 zoom really deliver good results? Let's find out.

Headline features

  • 18-35mm focal length (approx 28-50mm equivalent)
  • Extremely fast F1.8 maximum aperture
  • Ring-type ultrasonic focus motor with full-time manual override
  • Initially available in Canon EF, Nikon F and Sigma SA mounts; Pentax K and Sony Alpha to follow

Angle of view

The pictures below illustrate the focal length range from wide to telephoto (on Canon APS-C, 1.6x). The 18-35mm covers a modest 2x zoom range.

18mm (29mm equivalent) 35mm (56mm equivalent)

Sigma 18-35mm F1.8 DC HSM specifications

 Date introduced  April 2013
 Street Price (August 2013)  • $800 (US)
 • £650 (UK)
 • €850 (EU)
 Maximum format size  APS-C
 Focal length  18-35mm
 35mm equivalent focal length (APS-C)  • 27-53mm (1.5x)
 • 29-56mm (Canon 1.6x)
 Diagonal angle of view  76.5° - 44.2°
 Maximum aperture  F1.8
 Minimum aperture  F16
 Lens Construction  • 17 elements in 12 groups
 • 5 SLD glass elements
 • 4 glassmold aspherical elements
 Number of diaphragm blades  9, rounded
 Minimum focus  0.28m / 0.92ft
 Maximum magnification  0.23x
 AF motor type  • Ring-type Ultrasonic Motor
 • Full time manual focus
 Focus method  Internal
 Zoom method  Rotary, internal
 Image stabilization  No
 Filter thread  • 72mm
 • Does not rotate on focus
 Supplied accessories*  • Front and rear caps
 • Lens hood LH780-03
 Weight  810g (28.6 oz)
 Dimensions  78mm diameter x 121mm length
 (3.1 x 4.8 in)
 Lens Mount  Canon EF, Nikon F, Pentax K, Sigma SA, Sony A

* Supplied accessories may differ in each country or area


If you're new to digital photography you may wish to read the Digital Photography Glossary before diving into this article (it may help you understand some of the terms used).

Conclusion / Recommendation / Ratings are based on the opinion of the reviewer, you should read the ENTIRE review before coming to your own conclusions.

Images which can be viewed at a larger size have a small magnifying glass icon in the bottom right corner of the image, clicking on the image will display a larger (typically VGA) image in a new window.

To navigate the review simply use the next / previous page buttons, to jump to a particular section either pick the section from the drop down or select it from the navigation bar at the top.

DPReview calibrate their monitors using Color Vision OptiCal at the (fairly well accepted) PC normal gamma 2.2, this means that on our monitors we can make out the difference between all of the (computer generated) grayscale blocks below. We recommend to make the most of this review you should be able to see the difference (at least) between X,Y and Z and ideally A,B and C.

This article is Copyright 2013 and may NOT in part or in whole be reproduced in any electronic or printed medium without prior permission from the author.

Previous page Next page
266
I own it
464
I want it
28
I had it
Discuss in the forums

Comments

Total comments: 333
12
Eric Ouellet

Why those micro focus adjustment are necessary and would it be true for the new Canon EOS 70D also ?
The focus adjustment should be necessary on any lenses because it is the camera who read its and tell the objective to focus nearer or further.
How this could only affect this lenses ? Is this only a question of incompatibility between brands or could it be corrected by a firmware update ?

2 upvotes
Eric Ouellet

I have the impression that you penalize the lens for a problem of the camera body itself (focus evaluation) ???

3 upvotes
gftphoto

It seemed a bit strange that in the DXO test article not a single Nikkor lens was included in the extensive list of comparison lenses.

1 upvote
Andy Westlake

DxOMark's article is titled "Sigma 18-35mm F1.8 DC HSM A *Canon mount* lens review", which might explain the lack of Nikkor comparisons. Expect those to show up when DxOMark gets a Nikon mount version to test.

Comment edited 2 minutes after posting
7 upvotes
Daniel Bliss

I think DXO also kind of borked their test of the Nikkor 17-55. Their perceptual resolution shows the 17-55 at only 6MP on the D7000, a performance that would leave it well short of 35mm film and no better on the D7000 than on the D200 which has very heavy anti-aliasing and 40 percent less resolution. I can tell you that isn't the case at all. I suppose there was something wrong with DXO's test — perhaps a D7000 AF system out of adjustment or something like that. I can only say that my own experience of the 17-55 leaves me very curious to see how it would perform with a D7100 — in other words, it's in the ballpark with the Sigma.

The other question, of course, is focusing accuracy out of the lab. f1.8 gets into some pretty demanding territory on the AF system so I'll be interested to hear how that pans out for people. The combination of the 17-55 and most Nikon bodies is finicky enough already; the AF on the body has to be PERFECTLY in adjustment to run that lens reliably.

2 upvotes
Andy Westlake

DxOMark's lens testing doesn't rely on autofocus, but uses manual focus instead, carefully checked to ensure that the lens is optimally focused.

A far more likely reason for getting disappointingly low sharpness data would be that they may have tested a 'bad copy' of the lens.

2 upvotes
yabokkie

we are all human but I don't think people at DxOMark are stupid enough to use AF in their tests. MF on live view should be the standard for many years. only the sharpest shots will get processed and the highest number used.

btw, I only see bad copies of DX17-55/2.8, a good example of "good build, no good glass."

Comment edited 3 minutes after posting
3 upvotes
marike6

I agree with Daniel that the DxOMark test of the 17-55 Nikkor seems flawed to say the least. I owned the 17-55 f/2.8 a number of years ago pre-D7000 days, but remember it being bitingly sharp lens with extremely good color/contrast typically found in higher grade Nikkors.

Note that 17-55 f/2.8 performed extremely well in the Photozone and Lenstip resolution tests.

1 upvote
Dazed and Confused

These comments have made me realise that lots of people don't understand what the 'standard' in standard zoom means....

5 upvotes
yabokkie

it could mean different things and can be made clear with one word, standard aperture/grade or zoom range.

no problem unless some one call standard super high grade.

Comment edited 3 times, last edit 2 minutes after posting
0 upvotes
marike6

A "standard zoom" traditionally goes from wide-to-short telephoto like the classic 24-70 f/2.8 FF zooms. The APS-C equivalents are the 17-50 or 17-55 f/2.8 lenses.

The Sigma 18-35 f/1.8 only goes from wide-to-normal (27-50 on Nikon/Sony, or 29-56 on Canon) and therein lies the confusion.

Since it lacks the telephoto focal lengths it would seem to be a more specialized, slightly less versatile lens than a traditional standard zoom.

Comment edited 1 minute after posting
2 upvotes
Shamael

This is crap. Who can define a standard zoom? For that, we need first to fix that standard. If you look at other manufactureres, a fast Zoom with such an aperture is always short in range, the Nikon 17-35/2.8 for example, the Tokina 11-16/2.8 is another one. The faster a zoom gets, the shorter the range. Doing a zoom of a displacement of more than 20 mm with such an aperture is a challenge, and imagine the size you need to do this.

So, if Sigma brings us a 18-35 with 1.8, one can only say, "hats off" it's a performance. Now, if they did a 18-55/1.8, it would be the size of 300/2.8 barrel, and I doubt that anyone would buy it. My only concern is that it is somewhat expensive, but, i will not buy it anyway, I prefer to invest that money in a huge prime with 1.8 or 1.4 aperture.

0 upvotes
yabokkie

it used to be 35-70mm before but makers have been pushing it wide and wider to 28-70 and 24-70 now.

18-35/1.8 is about 28-54/2.8 equiv. so only thing we may complain is the narrower range which makes it more "standard" than standard 24-70/2.8 ones.

2 upvotes
Dazed and Confused

@Shamael

"the Nikon 17-35/2.8 for example, the Tokina 11-16/2.8 is another one."

But that's my point - those aren't standard zooms. They're both wide zooms - the first for Full Frame, the latter for APS-C.

1 upvote
vladimir vanek

hey, would "a premium aperture standard-range lens with standard wide-end and a bit shorter-than-standard tele-end" designation do? :)

0 upvotes
AmateurSnaps

Can
"Physically rather large for a standard zoom"

And whats standard about a f1.8 18-35mm lens.

Nice review and a very impressive achievement by sigma. Shame the likes of Canon can't innovate on anything less than £10,000

4 upvotes
yabokkie

f/2.8 zooms for 35mm format are good reference.

Sigma 24-70/2.8 is sold for less than half of C/N so price is not a special issue for this lens.

0 upvotes
Andy Westlake

Just to tidy up the unexpected controversy over my use of the term 'standard zoom' - this is traditionally used for a zoom that covers the 'normal' focal length range (~28-35mm on APS-C), and so might plausibly used 'as standard' for everyday shooting. The 18-35mm fits within this category.

4 upvotes
rfsIII

When is controversy ever unexpected on this site?

1 upvote
marike6

From the Lenstip review: "accuracy of the autofocus, our assessment is very positive". Any AF errors were less than 4%.

Note they tested it on a Canon 50D.

The Lenstip Con list for the Sigma 18-35 is appropriately short.

Any large aperture lens shot wide open, especially handheld will give occasional focus errors.

The Gold Award is cool but DPR's rather long list of "Cons" for this ground-breaking lens is unfortunate as Sigma truly deserves nothing but Kudos for.

5 upvotes
Andy Westlake

The lens's AF accuracy is great if you sit in a studio and point it at high contrast, well-lit test charts. Much less so if you go out and actually take photos with it in the real world. Personally I think photographers should be more interested in the latter. YMMV.

7 upvotes
Stu 5

Valid point Andy. Lenses do need to be tested under lots of different types of lighting otherwise problems like the one you discovered are well... not discovered. As soon as you test two or more lenses that show the same issue then it starts to turn into a major issue.

1 upvote
marike6

No my mileage is pretty much the same. The point is your review even talks about focus errors "wide-open" in "less than ideal conditions". But judging by the comments, the big takeaway from this review is that the 18-35 has an AF problem.

ALL lenses shot wide open in low-light have a lower accuracy hit rate related to the thin plane of focus. This "problem" is even greater on FF.

3 upvotes
Andy Westlake

@marike6: "ALL lenses shot wide open in low-light have a lower accuracy hit rate related to the thin plane of focus."

Obviously I know that full well (chances are I've shot with more different lenses and systems than almost anyone commenting here), but in my judgement the 18-35mm has an unusually high misfocusing rate when shot wide open in real-world use. This is corroborated by my colleagues who've also used the lens. The limited depth of field is not all that there is to it here - it misfocuses more often, and more obviously, than a Canon EOS 6D + 24-70mm F2.8 shot side-by-side at F2.8.

6 upvotes
HubertChen

Relevance of AF accuracy in bad light

I frequently shoot with 35mm f/2 wide open. And I actually do shoot wide open more in less than ideal lighting, because then you need to. So AF in real life in low light to be accurate is important, as it would be a popular shooting condition for people who would buy such a lens.

Is repeatedly accurate focus in low light 35 mm f/2 in real world even possible?

Until a few months ago, I would have set no. With cameras I shot in the last decade I could not accomplish it. With my own dated DSLR AF was so bad wide open that I always focused manually with split screen. Works 100% with tripod and fixed target. In real life everything moves and I lived for a decade with more misses than hits.

Then I got a new camera and suddenly pretty much all my pictures are accurately focused. Bad light, wide open and moving subjects.

I was stunned. Now I would not accept less with new equipment at this price level. I guess that is what Andy is saying.

2 upvotes
Pitbullo

How the heck can you have limited zoom range as a "con"? If you buy a 18-35mm, then that is what you get.
If you buy a 16mm prime, you cant say that it is a negative that it is too wide. It is what it is!

16 upvotes
Andy Westlake

The lens doesn't exist in isolation. If you compare to other standard zooms that you might buy to put on your camera for everyday use - a 17-50mm F2.8, for instance - it has limited zoom range. If you go out and shoot with it, the limited compositional flexibility this imposes really is quite obvious.

The list of 'Cons' is simply things we find users may need to consider when making a purchasing decision. The whole idea is that you can look at them and decide whether they matter you to you. If not - no problem.

9 upvotes
marike6

But the 18-35 is not a Standard zoom, and there aren't ANY standard zooms with f/1.8 max apertures.

2 upvotes
yabokkie

18-35/1.8 is a standard f/2.8 equiv. zoom for APS-C SLR. Sigma did a great job that it's much better than small aperture 17-55/2.8 zooms from Nikon and Canon.

what's great is not the aperture, but a high quality wide-standard angle zoom for APS-S SLR which is the most difficult mount (that happened to become popular for some technical reasons we had more than a decade ago).

Comment edited 3 minutes after posting
0 upvotes
DanielFjall

Would you ever go and buy a 18-35mm zoom? Probably not. Unless it's a 1.8 throughout the whole range and has some superb optical performance - which makes one wish it stretched a bit further to perhaps 50mm, no? As much as I'd love to have this lens, the limited zoom range is a dealbreaker for me.

2 upvotes
webmiser

I sometimes change my lens when I'm out shooting. I have been known to carry as many as five :o! With my 70D I can do 8-400mm with three lenses and still carry a 100mm macro and 85mm portrait.
My 'usually on' lens is now the 18-135mm STM. It has replaced the 24-105L as it is more versatile and handles movie clips much better.
I use the 18-35mm F1.8 for indoor stills and movies and have found no focussing issues. It is a fantastic lens in my view.
I have been a Sigma fan for a long time. QC issues have been unfortunate, but Sigma can usually fix them and if they can't they will replace the lens. They want to keep their customers. The recent offering from Sigma have been outstanding. I love the 35mm F1.4 which I use on my old 5D to give me amazingly sharp shots wide open.
In closing may I say that the DPR.com reviews seem to be pretty fair and accurate to me. I only read them after I've bought something, though :o)

0 upvotes
garyknrd

Good job in the AF tests IMO. Thanks

2 upvotes
RichRMA

Any still labouring under the illusion you can still use film camera lens circles of confusion as guides to the acceptable focus errors of a lens? Sorry, that ended around the 4-5 megapixel mark, perhaps even less than that. Cameras now are capable of such high resolution that any focus discrepancy is visible.

0 upvotes
Cheezr

E-mount please!

3 upvotes
mosc

You kind of get this already with the alpha mount and an adapter, don't you? It's still a PDAF optimized lens. It's also near as 2 lbs. Buy an SLT to stick it on, sheesh.

Comment edited 41 seconds after posting
0 upvotes
yabokkie

this is the greatest lens that I had been expecting and it came with better quality than expected, but I decided not get it for it doesn't add value for me. would be a no-brainer before 5D, good before 5D2.

still, great job and go ahead (f/1.8 zooms for APS-C mirrorless, too, and f/1.4 zooms for 4/3").

Comment edited 4 minutes after posting
0 upvotes
David Naylor

Wow, this "AF issue" is being blown completely out of proportion. I've had this lens for over a month now and haven't had even the slightest indication that there is anything wrong with the AF. If there is any inconsistency I'd say it is well within the limits of what is normal for any lens using phase-detect focus.

And since the open-loop AF myth has been de-bunked (i.e. PDAF *does* confirm focus after shifting) then a large part (all?) of any inconsistency will be the fault of the AF sensor.

7 upvotes
white shadow

You are quite right. According to "lenstip.com" in their reveiw which is much earlier than DPR, they have not notice any problem with AF except a little at 18mm.

However, they have encountered flares to occur in their test.

Overall, they have a very possitive conclusion.

So, we do not know how DPR reached their conclusion.

Now, we have to wait for "photozone.de" for their review and conclusion.

5 upvotes
David Naylor

In fact, none of the previous reviews I have read have mentioned any AF issues.

2 upvotes
Andy Westlake

We reached our conclusion in much the same way as usual - by shooting hundreds of real-world shots across a range of subjects and under different lighting conditions, using Canon bodies from the EOS 7D to the EOS 100D. As explained in the text, we looked at AF microadjustment (both in-body and via the USB Dock), which improved AF accuracy but didn't fix the shot-to-shot consistency we saw. Finally we got hold of a second copy of the lens to see if it showed the same symptoms which it did. Then the three of us who'd used the lens that most (all highly experienced SLR users) discussed it and decided the 18-35mm's focusing inconsistency was beyond what we'd usually expect.

I'm not sure what more we could do.

Comment edited 2 minutes after posting
8 upvotes
revio

@Andy Westelake:

What about trying a sample for another mount, like for Nikon?
I have owned a Sigma EX 18-50/2,8 for Canon, exhibiting quite "outspoken" AF-issues. Friends have owned Nikon and Sigma w no trouble at all. Actually, I have seldom heard of such issues with Sigma in Nikon or other non-Canon mounts. Surely there have been such issues, but I have a distinct gut feeling that Canon is the mount Sigma have had most trouble with, historically and possibly even today.

Comment edited 2 minutes after posting
4 upvotes
David Naylor

Thanks for the details. I'd really like to see this tested by Lensrentals.com too with a comparison vs a few other lenses. I just find it a little strange that this hasn't been noted anywhere else, and I believe I've read/watched pretty much every review out there.

Obviously, having bought this lens I know my view is probably biased, but I'm really trying to look at this issue objectively.

Btw, did you notice if the AF inconsistency was any different at 18 vs 35mm?

2 upvotes
Andy Westlake

@revio - we can't try the Nikon mount version, as it's not yet available.

@David Naylor - my personal impression (which I haven't systematically tested) is that AF inconsistency is a little worse at 18mm compared to 35mm, but no focal length is immune. For the record, the lens also focused more accurately and consistently in live view than when using the OVF.

Comment edited 44 seconds after posting
1 upvote
davidgp

Very informative review Andy. Thanks for your efforts.

While I've had the 18-35 for less than a week and have had limited opportunity to use it, I have to agree with David Naylor about the AF. I have not had any concerns with it thus far. I was surprised to find it was a concern in the review.

By way of contrast, my 50mmf1.8 is totally untrustworthy re: AF. I do not feel that way about the Sigma f18-35. Yes I know, apples and oranges. I'm just trying to give my sense of the magnitude of the problem: I don't doubt that the reviewers have identified a legitimate issue, but I simply haven't found it to be a limiting factor thus far. It certainly is a limiting factor with my nifty-fifty.

1 upvote
Octane

quote from the review
"... when focusing manually with most Nikon SLRs you have to remember to set the aperture to F1.8 before entering live view, as they're incapable of adjusting the lens's diaphragm during viewing."

Definitely not true for Nikon cameras and lenses in general. Just tried it with my D800 and 85/1.4 and it works just fine adjusting the aperture in live view. It even works with old mechanical link aperture lenses (like the AF DC Nikkor 105/2)

Maybe it doesn't work with Sigma lenses, but that is then their incompatibility.

2 upvotes
CameraLabTester

"they're incapable of adjusting the lens's diaphragm during viewing."

Yes, 3rd party lenses sometimes forget a few nuts and bolts, here and there. but hey... that's what makes 3rd party lenses exciting!

The GOTCHA!

.

1 upvote
Andy Westlake

@Octane - your D800 is, along with the D4 and D3 series, one of the few exceptions to the rule (which is why the text said 'most'). Of course these are all full frame cameras, and unlikely to be used with the 18-35mm. All of Nikon DX format SLRs show this rather unexpected behaviour in live view, where changing the aperture setting doesn't adjust the diaphragm. I've modified the text to make this more clear.

@CameraLabTester - this isn't anything to do with 3rd party lenses, but is inherent to Nikon's implementation of live view in all but its most expensive SLRs.

Comment edited 4 minutes after posting
7 upvotes
Horshack

AF inconsistencies for the different focal lengths across a zoom lens aren't too uncommon, which is why Canon started including the ability to set independent AF adjust values for either end of a zoom lens, with the firmware interpolating between the two for the intermediate focal lengths. The 70D is Canon's first APS-C body that supports this zoom dual-AF adjust facility; it would be interesting to see if that sufficiently resolves the AF inconsistency problem you observed. Perhaps you can revisit this when you have a 70D on hand?

0 upvotes
Andy Westlake

As we explained in the review, Sigma's USB Dock and Optimisation Pro software allows more detailed AF microadjust setting (4 focal lengths x 4 distances) than any SLR body. Even so, while it improved AF accuracy, it didn't fix the inconsistency we saw.

The 70D, however, is quite likely to give the best possible focus accuracy and consistency, as well as speed, when using its Dual Pixel AF in live view.

Comment edited 18 seconds after posting
2 upvotes
Horshack

Thanks Andy. I read the review but the conclusion is incorrect. That nature of a PDAF system limits the utility of a lens-based MA feature. It's actually much more effective to adjust in the body than in the lens. A PDAF acquisition involves multiple phase-detect sensing and lens movement cycles. Altering the magnitude of any single movement by the lens (lens-based MA) is an out-of-band correction for which the PDAF system will readjust for in the next sensing/movement cycle. In other words, it's an alteration that is lost in the feedback loop.

0 upvotes
Andy Westlake

I guess the main argument I'd use against this conclusion ('In other words, it's an alteration that is lost in the feedback loop') is that in-lens MA demonstrably works in practice, and can be set to give consistently accurate AF under favourable conditions. This was very obvious when I set microadjust up using the USB Dock and Optimisation Pro.

How might this work? Presumably, after using in-lens microadjust, the camera's AF system is happy that the lens is correctly focused after the initial focus movement, and doesn't readjust it again.

Overall, my experience is that armchair theory is all very well, but is no substitute for testing things out in practice.

1 upvote
Horshack

Andy, the determination on when the PDAF system has achieved sufficient focus (within DOF marign) is based on the phase detection. The need for AF adjust occurs when the optical path through which phase detection is measured has a misalignment or lens aberration which affects the accuracy of the phase comparison. Focus-element positioning errors (mechanical errors) are self correcting via the multiple PD cycles. But those errors are detected/calibrated against the phase sensing mechanism, which has no self-calibration mechanism and thus requires AF adjust. It may seem like an armchair theory but it's based on my development of DotTune, which you can read about here: http://www.dpreview.com/forums/post/50774257 and here: http://www.dpreview.com/forums/post/50883023

0 upvotes
Andy Westlake

And again - Sigma's in-lens microadjustment demonstrably works *in practice*, and that really is all that matters.

Comment edited 2 minutes after posting
0 upvotes
Horshack

Our discussion is whether in-body AF adjust works better than in-lens, and my assertion remains that in-body is better. I've experimented with Sigma's in-lens AF adjust on the 35mm Art (compared with in-body AF adjust) and my empirical results match the theoretical assertions. You say the Sigma's AF adjust works in practice yet you still report that its AF is inconsistent. Perhaps you're defining "work" as an improvement over the baseline, or perhaps as just observing a noticeable change in AF behavior. Going back to my original post I still believe the 70D's dual-AF adjust for zooms has a good chance of yielding more consistent AF results than what you observed using Sigma's adjustments.

0 upvotes
Andy Westlake

The Sigma's AF is inconsistent under less-than-ideal conditions. However, with a well-lit high-contrast target of the type used for measuring AF accuracy and determining microadjustments, it's much more consistent. Under these conditions, without in-lens AF microadjustments it's consistently (slightly) inaccurate, but with them it's consistently accurate. Therefore, in-lens AFMA works.

However, even after microadjust, the 18-35mm's AF remains inconsistent under less-ideal conditions on every camera we tested it on.

Comment edited 2 times, last edit 2 minutes after posting
0 upvotes
Horshack

Thanks Andy. For reference, an AF adjust value that appears to be correct/accurate but is not can manifest as inconsistent shot-to-shot results, esp in marginal AF conditions. Can't say that's what's causing your issue but just wanted to mention it as a possibility. I have a post describing why here: http://www.dpreview.com/forums/post/50781856.

Also, was the lighting temperature for your low-light AF test the same as your abundant light tests?

Comment edited 2 minutes after posting
0 upvotes
Andy Westlake

I didn't do a low-light AF test. I went out and shot with the lens in real-world use, and that's where its AF inconsistencies became clear. Obviously, camera/lenses need to be able to focus under all types of light.

0 upvotes
Edgar Matias

This lens would be even more interesting on m4/3 giving you 36-70mm-e (moderate wide to moderate telephoto) at f/1.8.

4 upvotes
yabokkie

more interesting because of longer focal length? or
because m4/3" lenses are smaller aperture ones?

you will waste 0.6 stops mounting this on 4/3".

3 upvotes
Edgar Matias

More interesting because 36-70mm is a more practical working range. You have wide, normal, and telephoto, in one fast lens. On APS-C, it only covers wide and normal.

See the reviewers' conclusion that its "relatively narrow zoom range is undeniably a little restrictive."

I suspect they would not have made that comment if it were an m4/3 lens covering 36-70mm-e.

1 upvote
Edgar Matias

'you will waste 0.6 stops mounting this on 4/3".'

Sigma wasted 0.6 stops making it an APS-C lens.

2 upvotes
yabokkie

shameful that Canon cameras are resolution challenged but D7100 got 14.8MP within 4/3" range (it's 13.4MP for 70D).

it's much better to be able to do post-cropping for we can have better framing and less chance to clip-out wanted image when composing too tight or shooting fast moving subjects.

Comment edited 2 times, last edit 2 minutes after posting
0 upvotes
yabokkie

> Sigma wasted 0.6 stops making it an APS-C lens.

we deserve f/2.8 ones for 35mm format, so
we deserve f/1.8 zooms for APS-C, and
we deserve f/1.4 zooms for 4/3" not stopping at f/1.8.

Comment edited 2 times, last edit 1 minute after posting
0 upvotes
AlpCns2

Both my Canon and Nikkor 50mm 1.4's are not consistent focusing in very low light. In fact most lenses are not that good focusing in low light. And none of them have microadjust.

My Sigma 50mm 1.4 and 85mm 1.4 are better performers in low light. Maybe it's also dependent on the body used?

1 upvote
Jonathan Lee

i'm seriously interested in ... a 16-35 f1.8 for FF. :)

1 upvote
Anadrol

Great... but now please make a 16-35 version.
Yeah... people are never happy :p

2 upvotes
SushiEater

Focus issues? How about testing OEM lens in the same under less-than-ideal conditions? Are you kidding? White horse "miles" away with no contrast to focus at all at F1.4, seriously?
How about testing on Nikon instead of Canon? There are so many variables here that it is hard to believe this review.

3 upvotes
Andy Westlake

We can't test on Nikon because the F mount version of the lens isn't yet available. The example shown on the review is illustrative, but simply one of many which were misfocused at F1.8. If you'd been standing beside when I was shooting it (and I'm pretty sure you weren't), you'd know it wasn't 'miles away', but about 3m.

3 upvotes
SushiEater

For this lens even 3m is like miles away if you consider that the eye of the horse is so small and no contrast at all. Might as well focus on the white wall. Plus camera like 700D could easily be fooled to focus on the background with a lot of contrast. So if you have more misfocused examples but with better target we sure would love to see them. And also would love to see Canon 35mm F1.4 in this situation side by side.

Comment edited 2 minutes after posting
0 upvotes
Haider

Just use a split-prism DOH! This way you know exactly what you're focussing on. AF is over-rated.

0 upvotes
Zvonko

I just don't understand why any brand would sell a lens which needs tweaks for focus from the moment you buy it.
Looks like a FAB lens but as others have said, the focus issues suck.

0 upvotes
revio

The way PD-AF as technology works is such that the precision needed to make a lens that always focuses perfectly well at max aperture like at F:1,8 is next to impossible to reach in mass pruduction like most lenses are the results of, so to speak.
Even if not mass produced it would still be next to impossible, simply because of the way such AF-systems are desiggned to work.
That´s why the need for "after manufacture adjustment" have come into existence, and why it is the chosen way of manufacturers to make their products more usable.

The constantly risen resolution (higher pixel count) of modern cameras made this necessary; had all cameras stopped at fex 6 megapixels these things (micro adjustment of AF) would not have came about.

Comment edited 3 minutes after posting
0 upvotes
EKB

This lens is not for me, not because I want a wide-ranging zoom, but because I want a long-ranging one. The 18-35mm range is just not one I use very much.

But if Sigma came out with a 35-70mm f/1.8 ASP-C companion to this lens, then I would WANT it.

4 upvotes
Jun2

That would be great.

0 upvotes
Frenetic Pony

I'd take both, and then a 70-210mm on top just to cover range I could possibly want. But 18mm at APS-C is a bit too narrow still, where's my 16mm-35mm eh?

1 upvote
Frank_BR

For a given aperture the weight (and the cost!) of a lens increases basically with the third power of the focal length. So a 35-70mm f/1.8 lens would weight about 6.5kg and cost about $6400. Still interested? :-)

Comment edited 4 minutes after posting
2 upvotes
kimchiflower

I'm not sure. The Olympus 35-100 F/2 costs around $2500 and weights a little over 1.5kg.

0 upvotes
EKB

If the cost and weight of a lens increases with the third power of the focal length, then this implies that there ought to be a big pile of good, cheap wide-angle lenses - that instead of "Nifty Thrifty 50s" there ought to be "Nifty Thrifty 15s". It implies that a Nikon 50mm f/1.8 prime should go for $600 - or even more for an FX version, given the 35mm f/1.8DX that goes for $200. It implies that either the Nikon 17-55 f/2.8 is way overpriced (especially for a DX lens), or the 24-70 f/2.8 is a serious bargain.

Something is wrong with this theory.

1 upvote
Frank_BR

@kimchiflower
The Olympus 35-100 F/2 was designed for the 4/3format, so it covers an area much smaller than an scaled 18-30 for APS-C. The Olympus designers could use relatively small rear optical elements, mitigating somewhat the effect of the "third- power law". Consider also that if the Olympus lens aperture were F/1.8, instead of F/2, the weight and cost of the lens would be considerably higher. To get an idea of how the weight increases fast with the aperture, compare a 50/1.4 lens with a 50/1.2.

0 upvotes
Frank_BR

@EKB
The "third power law" is a general trend, not an absolute law. If all the linear dimensions of a lens are scaled up by a factor k, the focal length increases by k, the f-number keeps constant, but the volume (and the mass) of the lens increases by k^3. In practice, this "law" is not so horrendous because the masses of the several parts are scaled only partially, or not at all, by the third power, as is the case of the mount. In general, the weight of the front optical elements tends to follow the third power law, but the diameters of the rear elements are related more to the sensor size than to the focal length, so the third power law applies only partially to the rear elements.

0 upvotes
EKB

@Frank_BR
In the 18mm to 70mm range, the general trend is so weak as to hardly be worth calling a trend at all. In fact, can you even name a single pair of Nikon F lenses of 70mm or shorter focal range that fit this third-power trend?

If I had to guess, I'd guess that a 35-70mm f/1.8 APS-C lens would be somewhat heavier and more expensive than the 18-35mm f/1.8. But not by a factor of 8. In fact I'd be suprised if it were more expensive by a factor of more than 2. On the other hand, I would *not* be suprised if a 35-70mm f/1.8 turned out not to weigh or cost *any* more than the 18-36mm f/1.8.

0 upvotes
Frank_BR

Prime lenses for the FF format with focal length below 70mm can have very dissimilar design: retrofocus (35mm or lower), double-gauss (around 50mm) and telephoto (70mm or higher). Of course, direct scaling does not apply for different types of design.

You should compare lenses of similar designs, like the Nikkors 105/2.0 and the 200/2.0, for example. These lenses weight 640g and 2930g, respectively, with a ratio of 4.57. This value is below the "expected" 6.91 (1.9^3), but the discrepancy shouldn't be a big surprise because the weight of certain elements that made up the lens (mount, barrels, motors, etc.) does not follow the "third power law".

0 upvotes
EKB

"Prime lenses for the FF format with focal length below 70mm can have very dissimilar design."

Yes. Exactly so. And the same applies to zoom lenses and to lenses designed for the APS-C format that have focal lengths below 70mm.

"You should compare lenses of similar designs"

Why? Why should I expect a 35-70 f/1.8 to have a similar design to the 18-35 f/1.8?

In fact, I would expect it to have a different design - to *need* to have a different design - in much the same way that, e.g., the Nikon 24-70 f/2.8 has a different design than the 14-24 f/2.8

As it turns out, the 24-70 has about the same cost and weight as the 14-24 - in fact, it's a bit lighter (900 vs 1000g) and a bit less expensive. By analogy, I'd expect a 35-70 f/1.8 to have about the same cost and weight as the 18-35 f/1.8 - not eight times the cost and weight.

0 upvotes
Frank_BR

"I'd expect a 35-70 f/1.8 to have about the same cost and weight as the 18-35 f/1.8 - not eight times the cost and weight".

I wouldn't be so optimistic. Consider first the Nikkor 24-70 F/2.8 lens that weighs 900g. Now consider a hypothetical 24-70 F/1.8 lens. The front elements of this lens would have a 2.5x larger frontal area compared to the 24-70 F/2.8. Usually, an F/1.8 lens has a much more complex design (more elements) and is physically longer than an F/2.8. What would be the weight of the hypothetical lens? Possibly more than 3kg, I guess.

In summary, with due respect to you opinion, I doubt it would be possible to design a 35-70 F/1.8 with the same weight of the Sigma 18-35 F/2.8.

0 upvotes
EKB

1. I'm not considering a hypothetical 24-70 f/1.8 lens, I'm considering a hypothetical 35-70 f/1.8 lens - and one that APS-C rather than FF, to boot.

2. By your logic here, the Sigma 18-35 f/1.8 couldn't possibly exist either:

Consider the Nikon 17-35 f/2.8 that weights 745g. Now consider a hypothetical 18-35mm f/1.8. The front elements of this lens would have a 2.5x larger frontal area compared to the 17-35 f/2.8. Usually, an f/1.8 lens has a much more complex design (more elements) and is physically longer than an f/2.8. What would be the weight of the hypothetical lens? Possibly more than 3kg, according to this logic.

3. Or to put it this way: The old Nikon 20-35 f/2.8 weighed 588g. The old Nikon 35-70 f/2.8 weighed 664g. The Sigma 18-35 f/1.8 weighs 811g. So a Sigma 35-70 f/1.8 should weigh what? Maybe 900-950g? Or maybe less, because the Nikon lacks the weight-adding 2mm at the wide end.

0 upvotes
Frank_BR

"1. I'm not considering a hypothetical 24-70 f/1.8 lens, I'm considering a hypothetical 35-70 f/1.8 lens - and one that APS-C rather than FF, to boot."
-------------------------------------------------------

A hypothetical 35-70 F/1.8 lens for APS-C wouldn't be much lighter than a 24-70 F/1.8 lens, because in this case the overall weight would be given by the upper limit of the focal length, which is 70mm for both lenses.

"2. By your logic here, the Sigma 18-35 f/1.8 couldn't possibly exist either:
Consider the Nikon 17-35 f/2.8 that weighs 745g. Now consider a hypothetical 18-35mm f/1.8. The front elements of this lens would have a 2.5x larger frontal area compared to the 17-35 f/2.8. Usually, an f/1.8 lens has a much more complex design (more elements) and is physically longer than an f/2.8. What would be the weight of the hypothetical lens? Possibly more than 3kg, according to this logic. "
--------------------------------------------------------------------
to be continued…

0 upvotes
Frank_BR

That was not a proper comparison! Even though those lenses have almost the same focal length, their angular coverages are very different for the wide angle end: 104º versus 73º. The difference of coverage makes a profound impact on the lens design. The Nikon is an ultra-wide- angle retrofocus zoom lens of very complex design, whereas the Sigma works in the relatively easy range of 73º to 43º. To have an idea of how difficult is to design a lens for a 104º angular coverage, please take a look at the Fig.29 (*) at:

https://dl.dropboxusercontent.com/u/60804047/British%20Journal%20of%20Photography%20-%20Feb.%201980.tif

Sadly the graph stops at 90º, but it is reasonably to conclude that a 104º angular coverage (Nikon) would demand 3 to 4 times more optical resources than 73º (Sigma).

(*) excerpt from "New Developments and Trends in Photography Optics at Zeiss", by Walter Wöltche, Head of Mathematics Department, Photo Optics Division, Carl Zeiss, Oberkochen

to be continued…

0 upvotes
Frank_BR

"3. Or to put it this way: The old Nikon 20-35 f/2.8 weighed 588g. The old Nikon 35-70 f/2.8 weighed 664g. The Sigma 18-35 f/1.8 weighs 811g. So a Sigma 35-70 f/1.8 should weigh what? Maybe 900-950g? Or maybe less, because the Nikon lacks the weight-adding 2mm at the wide end."
-------------------------------------------------------------------------

The Nikon 20-35 F/2.8 lens is an ultra-wide-angle, so it shouldn't be used in the comparison with the Sigma. However, I think it is valid to compare the Nikon 35-70 F/2.8 with a hypothetical Sigma 35-70 F/1.8. The question is: How heavy the hypothetical 35-70 F/1.8 would be? I don't know for sure, but, again, the Fig. 29 again gives a hint. Although the graph stops at F/2.0 and is specific for a normal lens, the Fig. 29 suggests that the cost of raw optical glass for an F/1.8 lens would be several times more than for an F/2.8 lens. This agrees with my perception that the weight of a 35-70 F/1.8 lens would be several kg.

0 upvotes
EKB

I'm not convinced. I still think that the 20-35mm f/2.8 is a proper comparison. Yes, it's harder to get 104º on an FX sensor than to get 73º on a DX sensor, but it's also harder to get 73º on a DX sensor than to get 73º on an FX sensor. So is getting 73º on a DX sensor closer in diffiulty to getting 104º on an FX sensor or to getting 73º on an FX sensor?

I believe that it's much closer to the former (given the same lens mount and flange distance), but I also think that we'll have to agree to disagree about this.

1 upvote
rallyfan

I don't like hauling lenses around and especially short focal lengths. This would be useful. The AF comments have cast serious doubt though. Maybe an update?

0 upvotes
itsastickup

Its AF problem is with low-light at low contrast targets. It may well be more usable than one might think from the review. Aimed at eyes and faces it may well be just fine for the average wedding photographer.

0 upvotes
Zerg2905

I think this is an impressive lens design...but why the AF issue? WHY? Come on Sigma, this MUST be solved, I have sold my entire set of Sigma lenses for the same reason: AF was horrible! Cheers...!:)

3 upvotes
itsastickup

I've been lead to understand that this is partly because the manufacturers of camera bodies often don't share the electrical connection protocols which sigma and others have to then reverse engineer, so to speak. So essentially it would be Canon being proprietal "Come one Canon, release the details...."

0 upvotes
imsabbel

I always wonder about the whole APS 1.8 vs FF 2.8 equivalency:

Are FF sensors as sensitive as APS-C, per mm^2? Because the equivalency only works if they are.

And I am not sure about that, seeing that APS-C outsells FF by more than an order of magnitude and correspondingly, the processes would be more optimized...

0 upvotes
Karroly

The APS 1.8 vs 2.8 FF equivalence mentioned here is for depth of field ONLY !
About sensor sensitivity, if ISO is a standard scale to characterize it, then a 200 ISO FF sensor as the same sensitivity than a 200 ISO APS-C sensor. If not, then handheld light meters are just useless...

2 upvotes
Anadrol

Yes 200 ISO is 200 ISO no matter the sensor size (even if there are discrepancies between brands), but the noise is different.

FF sensors are about as sensitive per mm² as APS-C sensors because if you look at DXO Mark results, the ISO performance is proportionate to the sensor surface (for same generation sensors of course).

Comment edited 2 minutes after posting
3 upvotes
forpetessake

"it offers a 28-54mm equivalent zoom range, and promises similar depth of field control to an F2.8 zoom on full frame" -- It offers everything the same as 28-54/2.8 lens on full frame, not only FOV and DOF -- why is that difficult for the DPR staff to say the whole truth out loud?
Another thing they should have mentioned. The lens and most APS-C DSLRs it supports are just as large and heavy as an equivalent 24-70/2.8 lens + FF camera. But the latter offers better range, better resolution, etc. That's an important point if somebody considers acquiring this lens.

Comment edited 5 minutes after posting
3 upvotes
D1N0

This is a lens for APS-C shooters. It doesn't matter what the full frame equivalent is. The full frame 24-70 F2.8 lens on aps-c would be Like 36/105 mm F4 on Full frame. But the choice is between lenses for people who have already chosen a sensor format. For these people Full Frame is non existent as long as they are shooting aps-c. But this is probably to hard to get into your thick skull.

Comment edited 42 seconds after posting
5 upvotes
zodiacfml

When a person decides on an APSC lens, one doesn't compare it to an FF equivalent. Sigma also created this on that principle too, there isn't anything like this on APSC, besides this still is f1.8, giving faster speeds.

The constant aperture and the constant high IQ in all focal lengths just costed it too much size and weight though. I'm all for IQ but I don't see myself using this for hours.

1 upvote
rallyfan

Ive not worked with anyone that cares about equivalency ramblings. FF fans: Don't buy the lens. Simple. Kthnxbye.

2 upvotes
Stefan M

Sorry. Can't hear it anymore. There is no such thing as equivalent to APS/FF/m43...whatever. Hence this lens is exclusively for APS it's even more absurd talking about any FF equivalency. A 18-35mm F1.8 lens will always be a 18-35mm F1.8 lens.

I guess a lot of people will enjoy this lens.

3 upvotes
rallyfan

Stefan M is exactly correct.

1 upvote
nawknai

I agree with Stefan M.

In fact, there are so many digital shooters using cameras with an APS-C camera that it would be better to move forward by just stating focal lengths for APS-C sensored cameras.
People are used to APS-C sensors, and the view that an 18 mm lens will give them on such a DSLR. There's no need for so many people to convert everything to full-frame equivalency when the majority of DSLR and mirrorless camera shooter aren't shooting FF.

I know some people will disagree, since the history of 35 mm film goes back decades, and SLRs have existed well before the advent of digital. However, I read that there are currently a far greater number of active DSLR users today than there were SLR in the decade before digital.

We don't state equivalent focal lengths and apertures in terms of medium format, and that used to be "full frame". I don't know why we treat 35 mm as some sort of standard, especially when those people who use the smaller sensor far outweigh FF users.

Comment edited 1 minute after posting
4 upvotes
AlpCns2

Glad to hear I'm not the only one completely sick and tired of this equivalency nonsense. I use APS-C and full-frame 35mm digital. And (real full-frame) medium format as well. I even dare to use film sometimes.

I am odd, perhaps, but I do not walk around all day making equivalency computations and DOF charts, "light gathering" comparisons and whatnot.

I shoot.

0 upvotes
nawknai

"The lens has a 50 mm focal length, or approximately a 35 mm equivalent focal length on an APS-C DSLR."

;)

1 upvote
Petka

At least in fixed lens cameras the lens should be described with picture angles like 80-4 diagonal degrees for a 20x zoom. Then the buyer would know exactly what the zoom range is no matter what the sensor size and the actual focal lengths are. People would get used to this quite fast, I think.

I shoot with APS-C and so called FF cameras and used to shoot also 120 film with several different film gate sizes, so I need not convert focal lengths to any equivalencies. It also seems that many are quite ignorant about the basic optical laws when doing the useless and often misleading/mistaken equivalency conversions.

And yes: 4x5" is the real "full frame", 135 is a miniature camera, originally made to use motion picture film...

Comment edited 2 times, last edit 1 minute after posting
3 upvotes
zodiacfml

I totally disagree and probably the DPR staff too, they don't do that just because 35mm is considered standard or any important. It's only because we need one format to compare other formats. I didn't liked it when I was new to camera reviews but it made sense or understood it later on.

For example, if I am an APSC and a Nikon 1/RX100 user, I would like to have a standard unit to compare lenses of different formats and focal lengts.

2 upvotes
Petka

View angle would be a perfect measure of that. 60 degree lens would be the same on a smartphone, APS-C, 135, 6x7, 4x5, whatever.

0 upvotes
DaddyG

Dubious focusing in low light wide open is nothing new.
Canon's older primes provide inaccurate focus in low contrast situations. (From my experience: EF 85 1.8, EF 50 1.4, EF 50 1.8).

1 upvote
Hendrick B

Would have been nice if they made a full frame version also!

0 upvotes
Sad Joe

'Inconsistent autofocus in difficult conditions even after microadjustment' - for me this is the KILLER - sorry Sigma I won't be buying one....

0 upvotes
Jun2

Canon 70D with new live view AF, should solve the problem. Too bad, the review didn't test that.

4 upvotes
Becksvart

Maybe it's a bit like some instances of the Sigma 85 1,4: different AF depending on whether coming from mfd or infinity, hopeless to micro AF adjust away.

0 upvotes
bugbait

I second this. A in article assessment of the focus consistency addressed via DP-AF in live view with the 70D seems a must. These two products are THE current hot news for APS-C. If the reviewer does do a follow up please PM me with a link to your results.

0 upvotes
Jun2

I used Canon 50mm 1.4 that wide open AF was not consistent, worse than my copy of Sigma 30mm 1.4.

1 upvote
jenbenn

Impressive IQ but unfortunately only a toy rather than a tool. If I shot landscapes I would have much preferred a slower, lighter lens with image stabilization. For the travel, street, documentary and event work that I do the inconsistent AF renders the lens useless for me. I might as well use a slower, smaller, lighter and more inconspicous lens and shoot at higher ISO. After smoothing the noise in pp the result should be the same as a misfocussed image from this lens. In the end it seems the lens will not turn your Aps-c camera into something of a full frame equivalent.(Save for the dof )

Dont get me wrong, I admire the optical achievement. But without reliable Af, this optical splendour is just wasted for the majority of the practical applications this lens was designed for. My agency (alamy) rejects all misfocused images (however slight) and for large gallery prints misfocussed images are not usable anyway. What a shame.

1 upvote
Francis Carver

Where exactly is the "shame" part with this lens, I guess I'm not getting that part?

1 upvote
jenbenn

That it doesnt AF properly.

1 upvote
dash2k8

Use the camera's microadjustment to fix it. We do that with other lenses, so I don't see why we can't do it with this one.

0 upvotes
NetMage

Did you read the review?

0 upvotes
dark goob

Why oly cant has?

2 upvotes
Bluetrain048

Indeed. Why can't has :(

1 upvote
pancromat

my highest respect to the sigma engineers for this exceptional lens, but from my point of view it makes absolutely no sense. i stay with the APS-C sized sensor, because i want a good compromise between IQ, handling and compact size of the "package", my canon 60D with a tamron 17-50 (no VC, trusted, lasted over three bodies) comes close to this. if i accept to carry a lens this big, and to carry it with comfort you need the larger camera bodies, i would go FF. in that zoomrange there is no creative benefit from 1.8, i'm pretty sure. so a nice compact 2.8/17-55 or so with that excellent IQ would have exhilarated me much more.

1 upvote
PCorvo

In your point of view the 28, 30, 35 primes makes absolutely no sense, is that right?

3 upvotes
pancromat

imho, absolutely. my photography now is about immediate reaction on the changing subject. changing lenses decelerate (and collects dust). don't judge me wrong. i experienced the creative challenge and the verve of a many nice primes (except there is no equiv. to a 1.4/35 FF for APS-C DSLRs, right). for several decades (beginning in the film era) i believed in primes, but i have overcome this religion.

Comment edited 38 seconds after posting
1 upvote
Frank_BR

In the glory days of the film era, the zooms were loved by the amateurs and despised by the professionals. Today, it is just the opposite. Go figure…

2 upvotes
Anadrol

Well, that you don't need the lens, doesn't make the lens less useful for others.

6 upvotes
CeleryBeats

Awesome lens. Awaiting Full Frame equivalent.

10 upvotes
geoson

I don't think that a 35mm equivalent is a sure thing. If anything, this lens may have been made with a goal of keeping APS-C users from moving up to 35mm. This is a large lens, and the 35mm version would be even larger. For some users on the fence between APS-c and 35mm, this may be enough to keep them in the "cruiser-weight" division.

2 upvotes
forpetessake

You're awaiting what? The 24-70/2.8 FF lenses were available, well ... forever. They also have better range and better resolution on FF.

7 upvotes
Anadrol

You should start going to the gym, the FF version will weight 5 kilos !

Comment edited 57 seconds after posting
1 upvote
zodiacfml

It is valid and possible. A FF equivalent of sensible size and cost would probably weigh and sized the same. The compromise would be 2x zoom, worse vignetting and CA, soft corners, less detail wide open, and not so constant aperture, like f2 or f2.2 at the longer end.

I'm actually baffled why it is an EF mount despite APSC coverage. I guess Sigma wants to force Canon to produce FF cameras with APSC crop modes similar to Nikon.

Comment edited 4 minutes after posting
1 upvote
CeleryBeats

Calorie burner and awesome optics in one package! Even better :P
I believe in evolution. Some people thought this lens to be impossible as well. One day we'll be holding our 24-500 1.0 compact lens and look back at the good old days :P

Comment edited 4 minutes after posting
3 upvotes
NetMage

@zodiacfml All third party lenses for APS-C use EF mount as far as I know - I bet the manufacturers see no reason to support another mount format since their lenses have to work with Nikon and other cameras that don't have APS-C specific mounts, they can't engineer an APS-C lens that takes advantage of the EF-S extra inside camera length.

0 upvotes
marleni

Sigma has shown courage and great skill to produce this outstanding lens.
I really hope we will see more high-quality lenses like this one soon!
Please Sigma please make a more versatile mid-zoom-range with F1,8 next.
Thank you :-)

1 upvote
Sad Joe

And so say all of us ! Except people who don't like 3rd party lenses or those who recall that Sigma have often had problems with AF in the past. No, I'm not blasting Sigma - this looks to me to be a sure fire winner !

0 upvotes
JDThomas

One thing I have to disagree with is two of the "cons"

"restricted zoom range". How can the zoom range be considered restrictive? Maybe if you compare it to a STANDARD zoom. But this is NOT a standard zoom. It's in a class of it's own. It's a super-fast zoom. You don't buy this lens for range. You buy it for low-light capability. If you look at it from that point of view it's not restricted. If you are going to lump it in the standard zoom category then ALL f/2.8 zooms must now have in the Cons column "restricted aperture setting"

"physically large for a standard zoom". Again, this lens is NOT a standard zoom. If you want speed you need big glass. To be perfectly honest. This lens a actually SMALL considering what it does.

39 upvotes
1vwGTIdriver

i totally agree with this comment ......

1 upvote
Andy Crowe

But this lens has been marketed by Sigma as being an alternative to a full frame camera with an f2.8 zoom, so it's perfectly fair to make the comparison and yes compared to an 24-70mm f2.8 zoom it does have a more restricted zoom range.

4 upvotes
nawknai

Agreed 100%.

"Limited zoom range"being a con is a laugh. That's a "pro" of this lens. That's why people would buy this lens!!

It's not an 18-250 mm lens, but that was never the point. Do reviews of prime lenses all say "limited zoom range" as a con?

Comment edited 1 minute after posting
3 upvotes
JDThomas

@Andy Crowe: "But this lens has been marketed by Sigma as being an alternative to a full frame camera with an f2.8 zoom"

Please show me where SIGMA has specifically said that this lens is an alternative to a full-frame camera.

I'll save you some time. You can't show me that because SIGMA never made that statement. It's being marketed as the "first wide-angle to standard zoom lens to achieve a large aperture of 1.8. Designed specifically for APS-C sized sensors".

Sigma makes no claim that this lens is an alternative to FF. They say it's a FASTER zoom for APS-C.

0 upvotes
photogeek

The lens weighs 1.8 pounds, and it is quite gigantic. With such size and weight the range does seem quite limited, I agree with the reviewer.

1 upvote
JDThomas

@photogeek:

The Nikon 17-35mm f/2.8 weighs 1.7 pounds, is the same width and is only a 1/2 inch shorter. I've never heard anyone complain that the Nikon 17-35 was a gigantic lens with limited range.

So in the same size package you get 1 1/3 of a stop more light. Seems pretty fair to me.

0 upvotes
Dazed and Confused

@JDThomas

The Nikon 17-35mm f/2.8 is a full frame wide zoom, not a DX standard zoom.

I do not understand why everyone is stating that the limited zoom is not a 'con'. It is. It just may not be one that is a big deal to you personally, but it's still a con compared to a 17-55 2.8. It's up to you to decide if you want speed or range.

The issue here seems to be that people do not understand that 'standard' in standard zoom means focal length - nothing to do with speed. Just like primes - wide, standard and telephoto.

1 upvote
JDThomas

It is NOT a "con" it is COMPROMISE that you must make when making the decision to buy a ZOOM lens with an f/1.8 aperture.

By your reasoning the Nikon 24mm f/3.5 PC-E should be lumped into the same category as the 24mm f/1.4G. They are both expensive, large, and heavy. Is the small aperture and lack of AF on the PC-E lens a "con"? No. Because it a specialty lens. As is the Sigma 18-35mm f/1.8

The only people that see it as a CON are people that want ONE lens to do EVERYTHING. This is a fast zoom. There is no other alternative, hence this lens is an a class of it's own and cannot be compared to a standard zoom.

In any case the the lens isn't even a "standard" focal length if you want to break it down to the brass tacks of your focal length descriptions. A 42mm equivalent is at the bare minimum of a normal focal length. So for all practical purposes this is a WIDE-ANGLE zoom not a "standard" zoom.

Comment edited 54 seconds after posting
2 upvotes
cordellwillis

JDThomas makes VERY good points. Since the Sigma is a lens in a class of it's own how can the range be a con? "Con" against what? Against a FF 24-70, 17-55, ....? Ok, so the flip side of those is the "con" of 2.8. All of these have compromises. It all boils down to which compromise works within your needs.

1 upvote
Andy Crowe

@JDThomas Oh yeah you're right, the bit I'm thinking of is actually the DPR editorial tacked on to the beginning of the original press release

0 upvotes
Dimit

It really seems an excellent ''short wide zoom'' lens.Let's also forget about mediocre bokeh as the focal distance is not the one relating to such an ability.Let's also forget about speedy auto focusing in the widest aperture as due to the narrow dof can't be stellar anyway.
The real benefit for such a short zoom lens would actually be to behave on a 90% level equally to the relevant fixed prime at any of the focal lengths in regard.My idea is that it does.A perfect choice although not extremely useful.

0 upvotes
marbo uk

I would have thought the zoom range would be a pro for this type of lens not a con. If i bought this lens it would be because of the zoom range at 1.8..

Comment edited 2 minutes after posting
1 upvote
Jose Rocha

It's a bit of a specialist lens. With under 2x zoom and falling in the focal lengths known as "standard" zoom lenses, it hardly replaces a 18-55, it's too big and too heavy especially for APS-C standards. Better think well first before spending the dollars! It will replace a few primes in the same range but should not be very practical for day-to-day shooting, I believe...

2 upvotes
Mais78

Inconsistent shot-to-shot AF rings a bell: same issue of the Sigma 35mm f1.4 (I owned 3 copies).

4 upvotes
InTheMist

Should use only Nikons for focus testing - much more precise, at least on one side ;)

Seriously, great performance Sigma from an enthusiastic 35/1.4 Art owner.

1 upvote
Sad Joe

This is what would worry me - Sigma (sadly) have a history of such....my one and only Sigma lens packed up on me - something in in over 35 years of using an SLR had never happen too me..Today I stick with Pentax for Pentax, Nikon for Nikon and Canon for my Canon's.....

0 upvotes
falconeyes

I would have been keen to learn about AF consistency using the new 70D's dual pixel live view AF.

1 upvote
Andy Westlake

Unfortunately, we don't currently have a 70D to test that out. Sorry for that.

0 upvotes
Matt

Are you kidding?

People are buying them in the stores and using them. How can you not have one and have started a review?

Other sites will be lightyears ahead of you

1 upvote
Andy Westlake

We've already published an [extensive preview of the EOS 70D](http://www.dpreview.com/previews/canon-eos-70d) based on a pre-production camera, including our usual set of test shots and a first look at the Dual Pixel AF system. We're due to get a full production camera soon. Sadly though we haven't had a 70D and an 18-35/1.8 in the same place at the same time.

Comment edited 30 seconds after posting
0 upvotes
makofoto

My 70D arrives on Wed ... from Amazon

0 upvotes
CarstenKriegerPhotography

I have been using the 70D and Sigma 18-35mm for about a week now and so far the live view AF is very consistent and accurate compared to the viewfinder AF. I haven't tried it out with fast moving subjects yet but so far it looks rather promising.

3 upvotes
Matt

@ Andy, preview? great but other have REVIEWS out.

You cant get a 70D? Are you for real? Maybe fire the person in charge of procuring products and get someone who can either get it straight from the manufacturer or can get his ass in a store and BUY it!

1 upvote
Adrien S

@ Matt, do you have to be obnoxious?
It's not because they don't have the one body you want that you have to try and give them such a scolding...
The preview they made was as good as any of your reviews since it was made with an actual 70D. Yet I believe they made it before the release date, with a body which wasn't for sale or given to them permanently by canon. That's why they don't have it anymore.

0 upvotes
Matt

@Adrien. I dont want a 70D But if the #1 camera review website can not get in the #1 new camera from the #1 camera manufacturer after it IS IN THE STORES, then someone is not doing their job right.

You can buy the 70D in the stores .... If they dont have one then its incompetence

3 upvotes
KentG

You seem to be under the misconception that reviews happen in the present and not in the past. Reviews are just like articles. They get done, then written up, then proofed, and then finally get published. Chances are when this was started they were not in the stores as yet and they only had a pre-production model somewhere in someones possession. And likely not the reviewers.
Commenters need to have a better grasp of reality.

Comment edited 24 seconds after posting
0 upvotes
sandy b

You seem to be under the misconception they buy all the equipment they test.

0 upvotes
Matt

@sandy b

I seem to believe that if I ran a review site I would get the camera wherever I have to so that I can have a review first and collect the revenue through web traffic. If I dont get it delivered by Canon, I will go out and buy it as it pretty much wont cost me anything anyways.

But thats just me

1 upvote
Josh152

Doesn't Amazon own DPR? Couldn't they just send a 70D over for testing and then DPR could send it back and Amazon could just sell it as "used-like new" and still make some money on it? It is completely ridiculous that DPR seems to be the only website that couldn't get a review copy of the 70D. Especially since their parent company has them for sale and in stock. Even DigitalRev's youtube review for the 70D is out already.

0 upvotes
rfsIII

Don't blame the dudes. The failure by the Canon PR department...they should realize that after Popular Photography, DPReview is the most important source of information for American photographers.

0 upvotes
photogeek

I don't get why people use "normal" zooms — this is the range where it's incredibly easy to "zoom" by just getting closer or farther from the subject. I get UWA and tele zooms, but I haven't owned a "standard" fast zoom for well over a decade now. I just have 35 and 50mm primes instead, and 80% of the time I just leave the 35 on the body.

3 upvotes
keithselle

I agree. I got a 35mm on mine body all the time. My main lens.

1 upvote
InTheMist

In my world there are fences, railings, rivers and streets.

29 upvotes
Andy Crowe

Zooming with your feet changes the perspective though, and there's not always enough space to move back far enough to get a wide view with a ~50mm equiv prime.

18 upvotes
Comitant

Zooming with your feet it utter BS. If you take a lot of photos often time you have to shoot within a second to capture the moment and there is not time to plod around.

20 upvotes
Silvarum

@Comitant: And always using zoom can easily make you lazy. Some people prefer photos with well balanced thought-out perspective over hasty snapshots, even if that means missing the moment.
No photo is better than bad photo.

1 upvote
makofoto

I often need WiDER and can't get further back in time if I'm shooting say a cloud formation. :-)

I fact I now often use my iPhone in Panorama mode to get really wide

0 upvotes
imsabbel

The thing is, the cool part of the lens is the 18/1.8 with damn fine quality part. The rest is all bonus - yes, i could crop the center of the image, but going in to 35mm is easier and higher quaility.

And some times, you just cannot go back. Try making images indoors, for example - which, conveniently, is also where you really enjoy having a fast lens.

1 upvote
Suntan

18mm at 7 paces is a completely different perspective than 35mm at 14 paces.

It may be "about the same" to you, but that doesn't mean we arethe all so generic in our composition.

1 upvote
photogeek

@Comitant: that's the thing. I don't take a lot of photos, and I delete about 80% of the few photos I take. I'm not interested in creating mountains of garbage that I will never have the time to dig through later on. If a picture is not excellent, or if it doesn't have documentary/sentimental value, it goes straight to the trash bin. Thankfully, this no longer costs anything.

0 upvotes
photogeek

@Suntan: yes it is, but if I need wider perspective, I'll just put on a 24mm prime. Still lighter than hauling a monster zoom around, better image quality, cheaper, smaller. What's not to like? Or I could even put on a 14-24mm zoom, if I expect to be shooting wide for a while (I have a FF camera). That way I can get much wider and get exaggerated WA effects, which I kind of like.

0 upvotes
photogeek

@imsabel: indoors I don't need a fast lens at all. I just bounce flash off the ceiling, and set my lens to whatever aperture it performs well at (about f4 for primes, f/2.8 if I have a longer lens on and need to blur the background).

0 upvotes
KentG

Well for one thing you won't be getting better image quality with primes, period. And you don't have to miss something by taking time to put a prime on or change one. I used primes for over 30 years. Now I generally use top zooms because there is little difference between them and primes, sorry. A bag full of primes will weigh as much as this lens of course.

1 upvote
rfsIII

Focal length is used to control the size and amount of the background behind your main subject. The wider the lens, the more background you'll see. The longer the lens, the less background you'll see. There are all kinds of reasons for this, but that's what it's for.

I know that from an optical science point of view that this isn't completely true, but for a working photographer it's a great rule of thumb.

0 upvotes
Frank_BR

In my opinion, the conclusion that "… the lens is effectively diffraction limited… (because) there's no measurable increase on stopping down" cannot be drawn from the measurements.

A perfect (diffraction limited) f/1.8 lens produces an Airy disk with a diameter of 2.4 µm, and a ultimate resolution of little more than 6000 LW/PH for an APS-C sensor (22.2 x 14.8mm). The corresponding MTF-50 would be about 3000 LW/PH, which is considerably higher than 2000 LH/PH measured by DXO. It is clear, now, that the measured "low" value of 2000 LP/PH is not directly related to diffraction limits, at least for the 1.8 aperture.

In conclusion, it is impossible to say from the data that the measured value of 2000 LW/PH was given by sensor limitations (3648 pixels height), lens limitations (aberrations and/or diffraction), or a combination of these factors. Probably the (excellent) behavior of the Sigma lens at 1.8 is due to a combined effect of aberration and diffraction.

5 upvotes
jhinkey

Exactly my thought when I read that. I have several lenses that don't get any sharper when stopping down when used on my D800 and they certainly are not diffraction limited.

0 upvotes
yabokkie

I think we can talk about perfect f/4 for 35mm format
(or perfect f/2.5 for APS-C or perfect f/2.0 for 4/3 ...)
which will need 470MPix (stacked pixels, need more for color filters, say 1.5 times) to fully resolve.

the calc is: 24 * 36 * (1477 / 4)^2 * 4
we may have perfect f/2.8 primes but let's go 500MPix first.

Comment edited 2 minutes after posting
0 upvotes
Jun2

It's almost perfect lens for now, I am not worrying about future bodies have higher pixel density. For me 16-20M is enough. 24M is bit too much. I never print 20X30 inches.

0 upvotes
yabokkie

I need more pixels because I never print (except for some seniors). people often think in detailed formats they have than what they need.

what we need is higher resolution that can equal or better our eyes which got well over 100 million pixels.

0 upvotes
KakoW

If my Nikon 35mm f/1.8 is having trouble nailing focus at f/1.8, I can't say I'm surprised the Sigma is the same. It's a very thin DOF. It's still the lens of the year and an insane bargain. I mean, 4 spherical and 5 SLD elements in a 800$ lens!

7 upvotes
io_bg

The 35/1.8's AF may be on the slower side but it pretty much always nails it for me.

0 upvotes
imsabbel

Yup. I noticed some issues with the focus too, at times.

Just out of spite I switched to f/2.8 and everything looked perfect. So its not worse than all the other lenses I own, its just that only the sigma has the shallow DOF to make it matter if the focus is set to the last meter.

0 upvotes
nelsonal

Looks like it could just about work on a 1.3x crop. Might be fun to try.

2 upvotes
pca7070

Yes it works!

0 upvotes
David Naylor

I own the Sigma 18-35mm f/1.8, the Canon 17-55mm f/2.8 IS, the Sigma 50-150mm f/2.8 IS, and two more lenses.

I don't find the Sigma's autofocus to be any more reliable or unreliable than any other lenses'.

This brings be on to a bigger question: Is it even technically possible for the lens to affect consistency in this regard? It is the camera's autofocus sensor that reports "sharp" or "not sharp", so how can this be the lens' fault?

I have always found this to be a problem with phase-detect AF, regardless of which lens I'm using. Is there really a lens out there which always delivers perfect focus using PDAF?

I have a personal theory that states that the wider a lens is, the harder it is for the PDAF sensor to tell the difference between in-focus and out of focus. This seems to be very much true for the two UWA lenses I have owned, as well as the wide end of my standard zooms.

4 upvotes
Andy Westlake

The Sigma 35mm F1.4 DG HSM focuses more accurately on the same camera bodies that we used to test the 18-35mm.

6 upvotes
yabokkie

from AF sensor's point of view, a lens' aberration (esp. peripheral for PDAF) and out of focus bokeh (may surpass the length of an AF sensor) at open can make problems. this means medium to not too large aperture lenses may perform better (say f/2.8 to f/4).

from the control point of view, some lenses may have heavy AF groups, non-accurate motor, old set of control commands, and bad execution to make AF slow and instable.

then Sig 35/1.4 may have well controlled peripheral aberration judging from the AF performance.

Comment edited 3 times, last edit 3 minutes after posting
0 upvotes
InTheMist

@David

If I'm not mistaken, a camera with phase detect AF actually says "focus to 1.667 meters" when it's not in live-view mode.

0 upvotes
makofoto

how does it know to focus on the eyes instead of say nose or side of the face wide open?

1 upvote
NetMage

For PDAF, the AF sensors are in fixed locations in the viewfinder, and chosen by the photographer, so the target for AF is whatever the photographer puts under the AF sensor.

0 upvotes
D1N0

Great lens but boring bokeh. Lacks character (or you guys are no good at making interesting bokeh shots).

0 upvotes
JDThomas

But isn't that what the masses want? If a lens has any character people tend to complain that the out of focus areas are "harsh" or "nervous" or "blobby".

10 upvotes
yabokkie

lens design, like everything else, is a compromise of conflicting requirements and you can get much better quality if can afford taking out one or two specs and put into minor list.

the Nikon 24-70/2.8, the king between 2007-2012 and still top class zoom, got not so good bokeh either and field curvature is not corrected well.

0 upvotes
Jose Rocha

@yabokkie: "Not so good bokeh"? For a standard zoom lens I think it's superb, and at 70mm it's beautiful. Of course, it's not f/1.8, as you know.

0 upvotes
yabokkie

> Of course, it's not f/1.8, as you know.

both of them are f/2.8 equiv.,
whatever the job an aperture can do they can do the same.

Comment edited 13 seconds after posting
0 upvotes
makofoto

@yab, Ahh Nikon, http://www.lensrentals.com/blog/2012/11/canon-24-70-mk-ii-variation

0 upvotes
D1N0

>both of them are f/2.8 equiv.,

not on an aps-c sensor :p

@JDT the masses must be boring as well.

Comment edited 3 minutes after posting
0 upvotes
JDThomas

@D1N0: "the masses must be boring as well."

Haha. No Comment.

0 upvotes
steelhead3

What you are saying in the conclusion is that the OVF of the Canon was not up to the task of confirming focus. If you had tested on a Sony with EVF, maybe the conclusions of mis focus would have been different (focus peaking and magnification views).

3 upvotes
Andy Westlake

Difficult to test this lens on a Sony at the moment, as it's not yet available in Alpha mount. Also bear in mind that the real problem is inconsistent focusing wide open, and it's not obvious a Sony body would do any better.

7 upvotes
yabokkie

it might be an interesting idea to make an accurate shift adapter for Sony E mount to test all full frame SLR lenses on a NEX body and you could sell it as "DPReview shift lens adapter #1."

Comment edited 49 seconds after posting
0 upvotes
Andy Crowe

@yabokkie what the shift adapter be for, in the review? Many FF cameras these days support CDAF so the lenses' CDAF potential can be tested on a FF camera no problem.

Also this lens is APS.

0 upvotes
photo nuts

EVFs disgust me. I used one on a mirrorless camera close to a year; I ended up dumping that EVIL camera and all the accompanying lenses. Heartily ran back into the comforting embrace of my OVF DSLRs.

Just say no to EVFs.

Comment edited 47 seconds after posting
3 upvotes
Flying Snail

Forget EVF - how about using the display?

0 upvotes
Andy Crowe

@photo nuts I suspect you may have been spoilt by decent large pentaprism viewfinders, I found the EVF on even the original Panasonic G1 far better than my previous DSLR's small pentamirror viewfinder in low light and at least as good (helped by the better magnification) in good light.

1 upvote
wansai

@photo nut

what didnt you like about using an evf? and what camera was it?

the evf on the nex7 and omd work really well. i find them to be excellent performers. sure in properly lit scenes i prfer an ovf but once you get into more challenging light situations, the evf outperforms its counterpart.

i have often found ovf to be difficult to use with heavy backlit scenes and under low light situations. with practice and experience you can guestimate the proper focus but it cant beat the live view the evf's offer. evf's are a lot more immediate. there's no guessing involved. you dial in your settings and see how it affects your shot. the ovf can only compete under ideal situations. all else, the evf is more useful.

1 upvote
Langusta

Well deserved GOLD award Sigma.

19 upvotes
JDThomas

Wow. First comment, but I have nothing to complain about...

5 upvotes
Total comments: 333
12