Sigma 18-35mm F1.8 DC HSM Preview: Digital Photography Review

Design

The 18-35mm follows in the same design idiom as Sigma's most recent lenses such as the 35mm F1.4 DG HSM. The section of the barrel between the mount and the zoom ring is metal, and the central section is composed of Sigma's 'Thermally Stable Composite' in an attempt to balance strength and weight. Rubber grips on the focus and zoom rings, combined with a high quality of fit and finish, give a sense of quality to proceedings. As always, on Sigma lenses, the mount itself is plated brass.

In terms of design and control layout the lens is decidedly conventional, with a large manual focus ring at the front, a zoom ring placed closer to the camera body, and a distance scale and focus mode switch placed between the two. As usual for this class focusing is internal; less conventionally for a normal zoom, so is zooming, which means that the lens stays the same length at all times.

On the camera

There's no denying that the 18-35mm is a pretty large lens - it's 10mm longer than the Tamron 24-70mm F2.8. It's also a fairly heavy lens - essentially the same weight as the more rangey Tamron. However it balances pretty well on high-end SLRs such the Canon EOS 7D shown left, helped by the camera's substantial hand grip. We suspect it's likely to be found on this class of camera most of the time.

On smaller, lighter entry-level SLRs such as the Canon EOS 650D, the overall balance becomes more front-heavy, meaning you'll often find yourself supporting the camera by cradling the lens itself. Frankly, these models tend not to have hand-grips that are comfortable to hold for long periods anyway, so this encouragement to support the lens is no bad thing.

Size compared

For a better idea of its size, here's the 18-35mm lined up alongside Sigma's 17-50mm F2.8 EX DC OS HSM and the recent Tamron SP 24-70mm F/2.8 Di VC USD. It's the narrowest in diameter but longest of the three, and weights almost as much as the full-frame Tamron.

Against the slower Sigma, the 18-35mm is considerably longer and narrower, and weighs 40% more. However, its more substantial build makes it more akin to the aged Nikon AF-S DX Nikkor 17-55mm f/2.8G, whose substantial construction means the fast Sigma is only 7% heavier.

Comparing a the 18-35mm on a mid-range APS-C body to the Tamron 24-70mm F2.8 on one of the latest, similarly enthusiast-focused full frame DSLRs, there's essentially no difference in overall bulk. The small differences in weight and and length between the lenses make no appreciable difference to the handling, making it essentially a wash.

Lens body elements

	The lens will initially be made available in Canon, Nikon and Sigma mounts, as tends to be the case from Sigma. We wouldn't rule out Sony and Pentax models appearing in future. Our preview sample was in Canon EF mount. Unlike Canon's own EF-S lenses it will physically mount on full frame SLRs, but will show substantial vignetting.
	The filter thread is 72mm, and doesn't rotate on focusing. It's surrounded by a bayonet mount for the petal-type lens hood.
	The manual focus ring has a moderately heavy feel to it. There's a hint of stiction with this pre-production version and, based on our experience with other Sigma HSM lenses, we would expect that to be present in the production lenses.
	The zoom ring has a pleasant, fluid movement. It glides smoothly without and play and feels correctly damped, giving a real impression of quality. It has markings at the 18, 20, 24, 28 and 35mm positions.
	A large switch on the side of the lens barrel sets the focus mode. Like on the 35mm F1.4 DG HSM, when it's set to autofocus (as here) the inlay behind it is white; when switched across to MF the inlay is black. This allows an easy visual check of its position even in low light. Here you can also see the 'Made in Japan' label - not something you'll find on all lenses any more.
	The bayonet-mount hood is provided as standard, and clicks positively into place on the front of the lens. It's made from thick plastic, and features ribbed moldings on the inside to minimize reflections of stray light into the lens. Sigma has even added a ribbed grip to make it easier to remove.
	A broad ridged grip covers most of the underside of the barrel between the zoom and focus rings, and provides positive handling when changing lenses.

Comments

Total comments: 152

Let us someday compare the Sigma (810g autofocus but no stabilization) on any APS-C versus the Zeiss 35mm F2.0 (530g pure brass & glass - no autofocus or stabilization) on the Nikon D800. With full-frame optical viewfinder (plus Nikon loupe) I have no trouble with manual focus.
For a walkaround kind of relatively cheap lens I have the Nikon 24-85mm F 3.5-4.5 weighing in at 465g in the other ring corner :-) (the cheap corner, the sharp in the middle and soft in the corners kind of lens). Let me guess: the combined cost of these 2 lenses will be less than the Sigma. This is exactly why I chose full frame - I would eventually pay more or less the same & yes the equipment would be lighter but quality and versatility would be better on full frame - until somebody figures out how to make APS-C 24 Mpixel sensors perform noise-free at 6400iso or more.

0 upvotes

""" 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."""
When we use the same shooting frame in both sensor sizes (APS-C and 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."""
As I know, in normal condition, the gather of light at each aperture is different with others. The lower number, the more of light. But when we have a quantity of light is "a" to be right exposured, that means the light come to the sensor is "equal" in any aperture (in relate to shutter speed), or any sensor size, not only at f/1.8 (in APS-C) # f/2.8 (in Full Frame)
I don't know why the author say that...

1 upvote

Hmm. Could this, for some people, replace a 24mm f1.4 and 35mm f1.4? Reasonable bokeh for full or 3/4 body portraits? Reasonable shutter speeds in dark spaces such as some churches? Sounds interesting to me.

But it's a pity that it's for cropped sensors only.

Comment edited 1 minute after posting

0 upvotes

This ia a 27-55mm lens but it has the unstabilized length of a 70-200mm f4 lens.
Everybody knows how difficult it is to hold a long lens still, and it covers APS-C only,so....

0 upvotes

to me it's easier to hold 85-135mm primes. better to have moderately longer wheelbase (between hands).

0 upvotes

The more I look at this lens the less sense it makes. It should have been a <20mm 1.8 prime

I love my NEX C3 but DX just can't compete with FX in the realm of speed. This whole exercise is a testament to the futility of it all. Better to stick with cheap glass and focus on better utilizing sensors. I forget which company it was that came out with those security camera sensors that were like ISO6553600. They could implement tech on APS-C sensors that would be cost prohibitive on FX sensors to cover the gap.

I mean I have been playing with an aperture equivalence calculator. f/2 on FX is like f/1 on DX. There is no comparison.

Comment edited 1 minute after posting

0 upvotes

I am using "Cheap" sensor with good glass and beating my friend's setup of expensive sensor with cheap glass. If you live on a budget and care about image quality, spend as little as possible on the camera and as much as possible on a good lens ( expensive not always means good )

1 upvote

Definitely not. F/2 on FX has the same DoF as ~f/1.4 on DX. And f/1.8 on DX has slightly narrower DOF than f/2.8 on FX when using equivalent focal length.

0 upvotes

@jaad75: a good lens isn't just about DoF. resolution, sharpness, contrast, colours n the amount of light it lets in matters too.

0 upvotes

or yeah, not forgetting distortion control n dark corners :)

0 upvotes

I really cant understand the concept of the Sigma 13-35 f1.8 lens. Yes, it's bright. But for what situation it's good? For interior or architecture shots you want to use small aperture, you never will shoot full open. For portraits its way too short. It's good news if they have new lens formula -- worth to wait bright normal and portrait lenses.

0 upvotes

Indoors, low light, concerts, fireworks, astro-photography, weddings, corporate functions...so quite a few, and I disagree that 52mm is "way too short" for portraits.

9 upvotes

it's very interesting shooting portrait at 24-28mm equivalent (24mm is challeging to me). you may have a favirot dish but you won't have it everyday breakfast, lunch, and diner.

it's almost useless to me because I don't need it on APS-C except very rarely I may need fast frame rate at wide angle. $1000 is way too much for that extra 2-3 fps.

0 upvotes

It looks like a fun lens, but then again I like crazy projects and designs. I wouldn't buy it though. I have to admit I'm part of the "I'd rather buy full frame"-crowd here. I can see uses for this lens, but none that would make it a must-have for me, and I guess the price point might deterr people as well. I like the fact that Sigma created this lens, but I doubt they'll be able to find the sweetspot between performance, usability and price.

2 upvotes

+1
it looks like a really good lens, a great one I think,
which is useless to me (but still I'm willing to be lured).

2 upvotes

I'm with rondom below in thinking that the "FF equivalent DOF" thing is getting out of hand if that's the key goal of lens' design. We do need to give this lens the benefit of the doubt - it's innovative and it may be very good. OTOH I'm a bit dubious about ultra fast lenses. Everyone bangs on about maximum aperture like it's some kind of religion, but they often fail to acknowledge the downsides....... Fast lenses may have more curvature of field, soft corners wide open, vignette more and flare more. And they're sometimes diffraction limited at an earlier aperture than their slower counterparts. And they're larger. And they cost a lot more. Too bad if you're looking for portability, classic even sharp performance and performance at small apertures. For my interests of landscape and travel, some of the sacrifices for lens speed just aren't worth it.

4 upvotes

bad news:
we don't get portability and image quality at the same time.
we don't get depth of field without sacrificing image quality.

good news:
we can get portability at a lot cheaper price
because smaller the aperture, cheaper the lens
thanks to the low quality of image.

Comment edited 2 minutes after posting

0 upvotes

>> And they're sometimes diffraction limited at an earlier aperture than their slower counterparts. <<

Why? When stopping down the lens to a given aperture, the resulting opening has the same size as with slower lenses. And it's the size of the aperture opening generating diffraction effects.

3 upvotes

Hi Yabokkie. I disagree that you can't have portability and IQ - there are plenty of 24, 35, 50, 70mm lenses from across all brands that are relatively small, aren't ultra fast and offer terrific IQ. I also disagree with the general concept that you can't have DOF without sacrificing IQ. It's a balancing act - you usually don't get optimal IQ wide open and you equally don't get it stopped right down. Each lens has its optimum.
It may be that some cheap lenses don't offer high IQ - we all know that - but it's not true that slower lenses are cheap simply because they always offer poorer IQ than a faster lens.

Hi Managarm, Many lenses reach optimal IQ 2-3 stops closed down from wide open. This is the point between unresolved aberrations of design and the onset of diffraction. If a lens is slower wide open, it will usually be diffraction limited at a later aperture. Obvious example is medium and large format lenses - often f5.6 max, but IQ is excellent to f16, f22, even f32.

0 upvotes

@Rod,

I think we should compare zoom to zoom and prime to prime across formats, and compare zoom to prime for the same format.

0 upvotes

Bokeh shmokeh....
Since when is the lens speed reduced to its dof implication only? Annoying to see the entire discussion going around "full frame equivalent depth of field"
you will be able to shoot at higher shutter speeds with this lens throughout the zoom range. Cropped sensor or not.

9 upvotes

at higher shutter speeds,
if and only if at lower image qualities.
you get both or nothing.

Comment edited 2 times, last edit 38 seconds after posting

0 upvotes

Sigma has hardly been able to produce any lens better then Nikkor.......so whats the big news......I don't think this lens will perform exceptional from any aspect.......

1 upvote

*cough* e.g. Sigma 35/1.4 *cough*

9 upvotes

@managarm: better value too :)

0 upvotes

The disappointing thing is that I thought, with the increased use of aspherical elements that some of the elements could be dispensed with, but this thing has 17 in it. Nothing kills contrast faster than air-to-glass surfaces.

2 upvotes

I find it interesting that Nikon <N> lenses only have one surface (7% of surfaces?) nano-crystal coated and the improvement is visible.

the contrast of this Sigma looks not bad. want to see 4-stop over exposed samples for HDR.

Comment edited 8 minutes after posting

0 upvotes

this is a lens for D300 owners ,if you have a D300 and find the d600 handling and ergonomics a bit rubbish, this gets you back some of the performance that has been lost to progress, and lets you still use one of Nikons best ever SLR bodies

or you could by a pair of second hand primes let your feet do the zooming

0 upvotes

While I like my D800, I'm sorry I sold my D300. It was a great camera.

2 upvotes

D800 owners will be able to use the 1.2x mode from about 24mm up as a lens of 29-42mm f/2.2 equiv. the corners may not be very good but the overall resolution should be higher.

Comment edited 42 seconds after posting

0 upvotes

even when i m using my zoom, i let my feet walk,
as i set the focal length first before composing my shot.
i can't use my feet for zooming as different focal length has different perspective.

0 upvotes

Whooopsy - the graphic at the top calls this the 18-35mm f/1.8 DG. Should be DC of course.

0 upvotes

To add an insult to an injury, Sigma's lens also looks better than Canon/Nikon counterparts. :-) Go Sigma, if nothing else C/N will up the ante or lower the price. They've been getting lazy and greedy as of late.

7 upvotes

depends on your priority,
good looking lens or good looking output image.

1 upvote

You gotta look good, babe. That's half the fun... Haven't you seen Blow Up? Or this awesome video: http://petapixel.com/2012/11/29/a-supercut-of-hollywood-movies-about-photographers/

2 upvotes

We will soon need a sigma rumors website. They are beating Canon and Nikon on the lens innovation lately. And shaming the big 2 with their prices.

8 upvotes

I still dont get this. Again we are simply achieving the performance of 2.8 zooms on full frame. And in order to do it we are handicapping the zoom range and making an aps rig big and heavy and expensive. why not shoot full frame. and if you have an aps c body wouldnt you rather have a lens that was pretty good on dof and low light but had a normal zoom range? I mean if you really want lovely bokeh you should look to standard zoom anyway they tend to not have as nice a look as prime lenses do in these ranges. And they are already plenty of primes that offer f/1.8 and even faster.

3 upvotes

may be for those who are more interested in some weird readings like f-number or ISO number in the middle than the output image.

and those who think full-frame is too high to reach. 18-35/1.8 is a great lens for APS-C.

Comment edited 15 minutes after posting

2 upvotes

Why not shoot FF?
How about lowest price for FF at 1600 Euro vs. lowest price APS-C 300 Euro?
How about weight and size of the camera itself?

Comment edited 2 times, last edit 2 minutes after posting

10 upvotes

Well, you can't really talk about price until Sigma announces price for this lens.

0 upvotes

why don't you all just shoot FF? why are you so poor that you can't afford one? why don't you work out so that you too will have the stamina to carry one around all day? why would you be embarrassed to walk around with a huge piece of expensive equipment? i mean, why would someone steal it? are there poor people around there?

why can you get decent-looking photos without spending thousands of dollars on equipment? why am i feeling ripped off lately?

6 upvotes

@ZAnton: this lens will cost almost as much as a FF body. But to get the same DOF on a FF body, you can use a much cheaper F2.8 lens. The price for lens + body will be almost equal.

And as for weight: FF bodies are not so much bigger nowadays. If you add the weight of this huge lens, it may be even heavier than a FF body and an F2.8 lens.

1 upvote

@mick32
this lens will cost same as f/2.8 lens for FF.
Size and weight: Compare Canon 5D3 vs. Canon 100SL or Canon 650D or 60D

Comment edited 2 times, last edit 2 minutes after posting

0 upvotes

Yea... a D600 is the same size and I think weight as a D7000. The more I think about it, the less sense it makes, though I still salute Sigma for committing to such an exercise. If I had $2500 to spend on a DSLR I would definitely go FX though, no question.

0 upvotes

should compare it against Sigma's own 24-70/2.8.

0 upvotes

Why use APS cameras? They're half the price of full frame so you can afford to update them more frequently; if you drop it or drown it there's less self-loathing; for the price of a FF body you can get two APS-C bodies and equip them with two different focal lengths which is the way God intended us to shoot; they're smaller, lighter, they're easier for people with small hands to use, and most importantly, a smaller camera makes you look like less of a techno-dork when you're out in public.

7 upvotes

@agentul

I had a full frame camera once, a Canon 5D and it was a piece of junk, the shutter release button fell apart at 8000 shots. I may never have a full frame again, at least not until the price fall to less than $1,000. I happen to love my Sony A57 and the quality I get from it.

Comment edited 41 seconds after posting

2 upvotes

i like this assumption that people will buy a new body along with this lens. it feels like a meeting of a marketing team. i know there are some technologically challenged people in the world (the kind of people confused by anything with more than 5 buttons), but there are others that are actually able to replace just one component from a more complex system. in this case "swap camera lens".

man, the paint on these walls is starting to show its age. time to buy a new apartment, i guess.

Comment edited 3 minutes after posting

1 upvote

That's it?

No samples?

0 upvotes

We've been told there are no examples finished enough to be shootable. Any images you may have seen are not from lenses representative of the final performance.

6 upvotes

I agree with Ryan, the only thing that counts in this type of reviews are IMAGES, the rest is bllsht. I don't care about handling, focus motors or 200 aspherical elements as long as there are no IMAGES to backup and justify the tech details.

To put things in perspective, this weekend I've borrowed an excellent Nikon AF-S NIKKOR 70–200mm f/4G ED VR and compared it with my 4 decades old Vivitar 200 f/3.5 manual focus which I bought for EUR 10,- it was only slightly better.

so IMAGES first please

0 upvotes

This will be an interesting exercise.

Sigma has a 24-70/2.8 for FF too, which actually is quite affordable. Comparing Sigma's MTF curves for both (looked up at Sigma DE) the new Art lens for APS-C appears to be a somewhat better performer. And has a more complex construction, e.g., with an aspherical front element.

So, it is a safe bet to assume the APS-C 18-35/1.8 will be significantly more expensive than the equivalent FF 24-70/2.8 which weights the same, collects the same amount of light and covers a larger zoom range.

I have difficulty to see how this can help the APS-C DSLR camp to survive ... As rather it is a good example that ambitious APS-C systems can easily cost more that equivalently specced FF systems, body and lenses combined.

Comment edited 3 minutes after posting

3 upvotes

totally agreed APS-C SLRs are poor performers but how significantly it is I don't have a good number. think it should be a fraction of a stop? (I consider > 0.3 stops harmful)

Comment edited 1 minute after posting

0 upvotes

Lol yabokkie you are talking about of your ***. Bottom barrel Nikons and Sonys of today outperform many professional DSLRs *still on sale today* in many metrics. Not to mention way more people have APS-C DSLRs than full frames, which means it makes a ton of sense to make lenses exclusively for APS-C sensors (as companies do/have). You dont know what you're talking about.

7 upvotes

by "0.3 stops" I mean 20%-30% more expensive for the same value. the "same value" itself may vary for different applications and different users so there will be some tolerance.

0 upvotes

nice debates since the announcement of this lens. yet, if i had this lens, I'd probably use it less than other lens out there. it's truly a beast in size and weight, only perfect for indoor, artificial light shooting to boost shutter speeds at lower ISO. i do hope Sigma sells a lot of this for innovation and taking risk which Canikon rarely does.

0 upvotes

whatever gives you the idea that a f1.8 collects same amount of light as a f2.8?
same DoF maybe, but not collects same amount of light.

i am a wedding photographer and i can share with you the many moments when i need aps-c over full frame.

consider this, low lights and a group of people that are of different distance to me. i'll balance my flash to match the surrounding while making sure DoF covers everyone and my shutter speed to be decent enough for hand held.

FF has its advantage, i have 5D and 5Dii, but crop sensor is not without its merits.

0 upvotes

if you are a photographer BCSeah, you should know that when you get deeper DoF, whatever the format is totally irrelavent, you have no way but to have lower image quality at the same shutter speed, the amount is exactly as you stop down the lens on any camera.

like it or not you don't have a choice.
the god/nature chose it for us.

Comment edited 3 times, last edit 15 minutes after posting

0 upvotes

I am interested in how this lens tests out. I am also interested on what it looks like on a full frame camera, interesting to see it will mount on the Canon 6D. If it only vignettes a little, this could be an interesting combination.

But I think this lens really is for the SD1M. I can' t see a strong argument for it blocking full frame sales. Really a Tamron 28-75mm f2.8 is sharp and priced at $500 New. You get extra 50-75mm range and a cheap camera like Nikon D600 can be obtained for under $2000. Or even a used Kodak SLR/n can be purchased for $500-$800. So Not much of a price difference.
This assumes the Sigma will be $1500. Which I think is practical. This lens however will be outstanding on the SD1M!

0 upvotes

Sigma is really evoluting into a new beast. hope they can come out with awesome 35mm format zooms after most of potential customers bought this one, I won't get it for more than 1000 US ... unless it can perform better than Nikkor 24-70/2.8 between 28-54mm (Sigma on D3200 vs Nikkor on D600).

Comment edited 2 times, last edit 10 minutes after posting

0 upvotes

I would like to know how yabokkie knows this lens is "not very good at tele-end" when there are not any photos on Sigmas own blog site or any other place I can find. Perhaps he/she could disclose their sources so we could see for ourselves.

2 upvotes

take it easy. let's wait and see.
I could be wrong and I hope I'm wrong on this, sincerely.

Comment edited 2 minutes after posting

0 upvotes

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

There are images out there. It's just that some people have no idea what they are talking about. Especially the ones decrying things based on personal bias. A real mook, I'd say.

1 upvote

@rxbot and gl2k,

you can find more samples, including resolution and bokeh tests on the same site by just "cropping" the link.

the final result should come after we compare this on 24MP Nikon APS-C (besides D7100) against 24-70/2.8 on D600.

Comment edited 2 times, last edit 9 minutes after posting

0 upvotes

though no MTF test yet, some Korean samples are quite promising it looks like same or better resolution than slower f/2.8 APS-C zooms (resolution falls near 35mm).

no one should expect good bokeh from 4 stacked aspherical but it's not too bad (from the limited samples).

if AF is also good and the lens robust it could worth more than $1000.

btw, this is also a 30-35mm f/1.8 zoom on full-frame, weird enough but it's world's first f/1.8 "equivalent" zoom (I would prefer a new Nikon prime).

Comment edited 2 times, last edit 12 minutes after posting

0 upvotes

i don't thing sigma will get AF issue solved on this one.

0 upvotes

What AF issue?

0 upvotes

Olympus' 14-35mm f2.0 (28-70mm) weighed in at a similar 900g to the Sigma. It boasts legendary sharpness and is weather sealed.

It is currently $2300 on Amazon, despite being released years ago. I'd expect this sigma to be around that price bracket upon release.

Comment edited 31 seconds after posting

0 upvotes

think we should refer to products of larger market share. small share itself results high price. Oly 14-35/2 doesn't worth the price at all but I acknowledge the cost should not be low.

0 upvotes

The Olympus 14-35mm is NOW $2300. But it was $1800 around launch time. It also sold New between $1400-$1600 When the 43rds cameras were popular. This higher price is partially due to it being an outstanding lens with few whom purchase.

1 upvote

yabokkie: rubbish, as usual. The 14-35 is acknowledged to be one of the highest-performance standard zooms for any format. It's only "not worth the price" if you think 4/3 is somehow crippled by "basic physics"... i.e. from nonsense comes nonsense.

2 upvotes

btw, I don't think EF24-70/4LIS worths the money, either.
if that makes you feel better.

Comment edited 3 times, last edit 12 minutes after posting

0 upvotes

Price point will be its Achilles Heel.
This lens is unprecedented and actually has no serious competition !! Now guess how that translates to the expected price point.
My guess : $1700 to $2300. Which raises the question whether it really adds that much to a lens portfolio, especially if you already own the Nikon 17-50 2.8

0 upvotes

Olympus' 14-35mm f2.0 (28-70mm) weighed in at a similar 900g to the Sigma. It boasts legendary sharpness and is weather sealed.

It is currently $2300 on Amazon, despite being released years ago. I'd expect this sigma to be around that price bracket upon release.

Comment edited 49 seconds after posting

0 upvotes

Oly ZD14-35/2.0 was designed based on wrong understanding of simple camera industry rules and should not be used as reference.

Comment edited 12 seconds after posting

0 upvotes

Kudos to Sigma for taking the lead on this. I do think the zoom range is a little weird, but then again people buy super wide zooms that are nowhere near as fast (10-20mm f/4 for example). 18-35mm on an APS-C is great for street and general photography, and is a popular range for primes. So for someone willing to deal with the heft and drop in IQ vs a prime this is a good buy.

Still though, I would prefer something like a 20mm 1.8. Preferably in E mount. Sigma actually tried a few years ago but failed miserably. Maybe now with more sophisticated computer tech they can right the wrongs and try again.

0 upvotes

actually this lens is not very good at tele-end so an even narrower range for those who want high quality.

0 upvotes

What are you basing that on yabokkie. The sample images I've seen are very sharp. The distortion is very low on both the wide end and the tele end. What is not very good to you at the tele-end that makes you say that it's not very good?

Samples images I've looked at were from here:
http://lcap.tistory.com/entry/Sigma-ART-18-35mm-f18-Preview

1 upvote

it depends on the requirement.
it can be high and it can be low for different people.
but it's difficult for me to think people want this lens for low.

0 upvotes

Again, what are you basing this on? There are barely sample photos out yet.

I do kind of agree though that the zoom range is a little weird. I mean it is good for general photography, but not really much better than a 24mm prime. I have been waiting for years for a super fast 16 to 20mm prime (like f/1.4). I think it could be done for a reasonable price (<$500) and for not much weight (<500g) if it was designed exclusively for APS-C and not APS-C/FX. That would hit the IQ required for a lens of this price and heft.

0 upvotes

> ... a super fast 16 to 20mm prime (like f/1.4).

would call f/1.4 on APS-C fast but not super fast.
think Nikon 28/1.8 (18/1.1 APS-C) is very good.
no need to hang self from the APS-C tree.

Comment edited 6 minutes after posting

0 upvotes

yabokkie do you just argue to argue? 20mm 1.4 on an DX is equivalent to 28mm 2.0 on an FX- wide angle FX doesn't get any faster than 2.0, 1.4 is definitely "super fast". I know you hate APS-C but for people who dont have $2000 to spend on a body, or for people who don't want to carry around a 6lb magnesium albatross + theft magnet they are great. Stop projecting your preferences as objective truths

0 upvotes

> FX doesn't get any faster than 2.0

first it's f/2.2 on FX.

it depends on what you mean by fast. to me fast means fast shutter speed at the same output image quality. I guess it may be at the same ISO for you. I disagree with you because the same ISO doesn't give you the same output, so you are not comparing things at level ground.

what you really need is the image quality right?

0 upvotes

When are people going to be satisfied?
If it’s built from good quality materials… why is it so heavy
If it’s the fastest zoom yet… why it is so short range
Why don’t you sent letters to Canon and Nikon and ask them to build a 16-70mm f/1.8 FF compatible, at 750g and bellow $2,000… and also weather sealed and stabilized?
If price for this Sigma goes above $1,400… I’ll be also screaming Why why why :)

6 upvotes

you forgot "made entirely of metal".

1 upvote

would prefer "made entirely of stone."
both man and woman love stones.

0 upvotes

Sigma must make a 15-45 OS F1.8 or a 17-55 OS F1.8 or something in the middle between this kind of glasses and than we are talking ...

35mm is too short and 18mm is not wide enough especially on a Canon 1.6x crop and the missing optical image stabilization is a problem both for low light photography and for video makers on the go

Great achievement for Sigma, but not enough to get 17-55 F2.8 IS owners to change the glass, IS is more useful than expected and once you get 55mm focal length you wont go back to anything shorter and 55mm at F2.8 is a better portrait lens than 35mm at F1.8

Sigma keep trying, you are almost there ;)

1 upvote

Image stabilization won't stop subject motion, but the stop+ light gathering ability might. So it depends on what you're shooting.

1 upvote

we are on the way to 15-45/1.8,
though I'm not sure when we can get there.

0 upvotes

I know IS won't stop movement, but F1.8 is a little bit more than 1 Stop compared to F2.8 and with an ISO jump it's solved, IS gives you 3 Stops or more and you can't go that much high with ISO ...
as I said OS + F1.8 both are needed to make a real jump ...and more mm at the longest focal length will make this glass really useful

Comment edited 23 seconds after posting

0 upvotes

you are getting yourself confused.

1 stop more light allows to use half the exposure time (so, less blurry fast moving subjects).

3 stops of Image Stabilization means one can use a 8-times LONGER exposure time and not have blurry images of stationary objects.

So they are used for rather different reasons. Tele lens for sports need both, because they deal with fast moving subjects and camera shake is a huge issue at long focal lens.

Portrait or street photography rarely can benefit from IS, because the longer exposure time allowed by IS is not needed or desired when shooting moving subjects. So yea, your IS lens can shoot at 1/4 handheld and keep the background sharp, but you will use it at 1/60 just to keep the people from being blurry, so effectively not use IS.

Comment edited 3 minutes after posting

1 upvote

Sigma should launch their DSLR models with this new lens.

Hear's a tip for your future DSLR's Sigma:

"Drop the 'Merrill'... it sound's... cleaner". --Sean Parker

2 upvotes

OK, on a Nikon DX camera the lens has app. the same light-gathering capability as a 27-53mm, f/2.7 lens on FF, and on Canon APS-C it's (FoV, noise and DoF/diffraction) equivalent to app. 29-56mm, f/2.9 on FF, but in terms of lens design (and exposure) it's still a f/1.8 lens, so it'll be very interesting to see how it performs 'wide open' (resolution), compared to the 24-70mm, f/2.8 lenses on FF.

1 upvote

And now new rumors appear regarding those DG 24mm f/1.4 and 135mm f/1.8 OS.
There was a rumor 2 years ago about Sigma developing fast primes 24, 35 and 135… since then 35 already hit the market

0 upvotes

Brace yourself... the 'equivalent aperture' posts are coming.

12 upvotes

It NEVER fails!!!! LOL!

1 upvote

Does it bother you?

0 upvotes

DPreview already covered this:

"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."

No need to post about equivalent aperture when it's in the article itself. And if their equicalent aperture reasoning bothers you, then take it up with DPReview.

Comment edited 1 minute after posting

2 upvotes

also:
"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."

so it is brighter than a f2.8 (twice) but the sensor is less sensitive. (which does not have to be true so it does not belong in the equation.)

0 upvotes

> so it is brighter than a f2.8 (twice)

that "brighter" has no photographic meaning.
an f/1.8 for APS-C is same as bright as f/2.8 for 35mm format.

Comment edited 3 minutes after posting

0 upvotes

yabokkie. So put it on a full frame dslr and shoot in crop mode. Twice brighter than a F2.8 on the same pixels. So you are going to need a stop less iso , so better iq than a 2.8

0 upvotes

> D1N0,

before we can conclude anything, what angle of views are you going to shoot the two photo for comparison? and what you want to compare?

0 upvotes

The angle of view does not matter since de aperture stays the same. The real difference is sensor characteristics. Which can be multitude. If you take only sensor size as a variable you are ignoring pixel density, pixel sensitivity (cmos, ccd, xtrans cmos?) You'd have to equivalize for all of them. That does not seem sensible to me. Keep lens and sensor separated.

1 upvote

@yabokkie the whole point of f-stops is that focal length (angle of view) has no effect on exposure values.

Therefore if you put an FF zoom @ f2.8 on an FF camera, took a photo, then put this lens @ f1.8 on the same camera and used the same ISO and shutter speed you would have a central circle of image 1 stop brighter than the same part of the f2.8 lens surrounded by black where the image circle cuts off.

1 upvote

did I mention pixel?

what crop means is very clear to me. by cropping you crop off light project on that part of the image, and by cropping you crop off image quality from the frame.

with less light after cropping you get lower image qualty and thus lower speed (which is a weired expression of the lens' capability to get quality image).

Comment edited 3 times, last edit 3 minutes after posting

0 upvotes

cropping means less image, not les iq. iq is dependent on the sensor. smaller sensors tend to give more noise, because of higher pixel density meaning less foton's per pixel. Cropping is not some kind of magical nd-filter.

1 upvote

you cut an image in half, each got the same quality, so you get double the quality? we need a billion pixel sensor now.

> cropping is not some kind of magical nd-filter.
correct. cropping gives you lower resolution, nd-filter doesn't.

Comment edited 2 times, last edit 2 minutes after posting

1 upvote

ofcourse not. That's not what i am saying. I'm just saying that f-stop iq equivalence is nonsensical because there are too many influencing factors you are leaving out of the equation. Want better iq. Just get a full frame.

It is a arbitrary question anyway because modern aps-c camera's do very well at high iso, so in most situations you are going to use this lens in you will have great noiseless pictures. The greatest advantage of this lens is narrow DOF for cute bokeh.

3 upvotes

> there are too many influencing factors

there is only one, the size of aperture.

0 upvotes

> there is only one, the size of aperture.

actually you don't have to look at behind the aperture, whatever optics, whatever format. the speed is decided at a certain angle of view by aperture already (the upper limit, the optics and sensor can only waste less).

0 upvotes

@yabokkie
> there is only one, the size of aperture

So the only thing that matters is the size of the aperture? Sharpess doesn't matter? Sensor tech (say comparing one of the early FF sensors to the latest APS or even m4/3 sensors) doesn't matter? Nothing matters except measuring the aperture with a ruler?

0 upvotes

@Andy Crowe,

I understand what you are saying is that you have no objection to what you quote on lens speed. thanks.

0 upvotes

Yabokkie

The size of the aperture is relative to the focal length. focal length divided by max apparent aperture (in mm) is f-stop 100mm lens with 50mm aperture is F2. So this lens is f1.8 (it's a fixed aperture zoom) the apparent size is going to be between 10mm and 19.44mm. De amount of light focussed in the image circle stays the same wether at 18 or at 35mm. Per square mm on the sensor it is the exact same amount of light as f1.8 on a Full Frame. Any difference in iq after that is due to the sensor. Come op with a noise/sharpness coefficient (iq) for the sensor, and you can say sensible stuf about overall equivalence.

1 upvote

> size of the aperture is relative to the focal length

aperture size measured using a ruler man.

the equiv. f-number is no f-number. it's absolute aperture size for 35mm format. this is what we really need.

same for equiv. focal length. it's no focal length but angle expressed in a weird way for weird people to understand (incl. me that I get 50mm more easily than "46 degrees diagonal").

same angle of view, same absolute aperture size, same photographic result.

Comment edited 3 times, last edit 13 minutes after posting

0 upvotes

"50mm f/1.4 equiv." is the same as
"46 deg AOV, 35.5mm aperture."
same thing in different expressions.

the later term will give you the same photographic result across all sizes of sensors but most people including me won't grasp it easily. I don't like (and don't hate) 35mm format but that's de-facto standard for most people.

0 upvotes

By your calculation you would need a 33.3mm F0.94 on aps-c to equal an 50mm F1.4

0 upvotes

we have a tradition to use rounded-up f-numbers so the calc may give small errors but still correct. we still have T-number issue even the f-number calc is accurate. no big problem we have been doing the same for more than a century.

think f/0.9x are okay for APS-C (mirrorless ones like NEX and EF-M) but things will get more and more diffiicult when f-number approaches 0.5 for 4/3" and 1" (if possible).

0 upvotes

Just like shutter speeds are the same speed on FF or on APSc cameras. So are all the shutter speeds on all formats.
Aperture readings are also the same in FF or APSC or all the other formats. This is standard, Every lens or camera manufacture have to ubide by these rules.
So f1.8 (FF lens) on FF sensor is the same f1.8 (APSc lens) on APSc sensor. Yes the f1.8 FF lens aperture pupil has to be bigger to cover the the larger format but the exposure should read the same. The only difference is the field of view. Depth of field might look like f2.8 on FF.
ISO (wich is also is standard) and the light propertys is another matter due to sensor technology, processor technology, SNR, size of format, size of pixels, numbers of pixels, density of pixels and the rest of the process. It varies from camera to camera.

1 upvote

whatever standard you say there is no way you can use same f-number or ISO to get the same photograph from different formats.

Comment edited 2 minutes after posting

0 upvotes

aperture ~ amount of light
exposure = amount of light * exposure time = aperture * shutter speed
ISO ~ pixel sensitivity (size and tech) * signal amplifier (processing)
image brightness = exposure * ISO = aperture * shutter speed * ISO
The magic happens when the image projection (the light) is captured as a digital image. Unlike the image projection, digital image doesn't lose its brightness when enlarged and that may be the cause for confusion. In the world of photography ISO means something. In the world of electronics ISO is just to make it simple for photographer to understand and calculate. If ISO values were reported objectively for each format then it would be like:
APS-C ISO 100-6400 real ISO 250-16000
P&S ISO 100-6400 real ISO 1000-64000
For example, 18mp APS-C camera can produce same image brightness as 18mp FF, but it doesn't mean that aperture has the same value for whatever pixel size there is, it means that smaller pixels are used with higher REAL ISO values.

1 upvote

Do the people at DPreview understand the concept of the f-stop? An f1.8 lens gives the same exposure no matter how large the light circle. When I shoot medium format the exposure is not affected by the final crop I use of the image. An f-stop of f1.8 means f1.8. The f stop does not go down if you use a smaller sensor or crop the image.

5 upvotes

You are trolling or you really don't get it.

5 upvotes

Yes they understand. Do you understand the concept of reading the paragraph in which they very clearly explain the context of this lens with regards to F/2.8 on FF and ISO settings?

15 upvotes

Of course they understand. Where did they mention different exposures?

Comment edited 2 minutes after posting

3 upvotes

They never said anything about crop affecting the exposure. They are talking about total light gathering ability and its relationship to ISO performance.

FF cameras compared to crop cameras have about a stop better ISO performance when it comes to noise.

So if you're shooting a F1.8 lens wide open at your minimum usable shutter speed on a crop, you will get about the same amount of noise as a F2.8 lens wide open on a FF. That's because you'd be at ISO 1600 (for example) on the crop and about ISO 3200 on the FF.

Comment edited 3 minutes after posting

4 upvotes

@Stephen_C

Sorry, but it's quite incredible that some people don't understand the difference between exposure and light-gathering ability.
Of course DPR understand that F/1.8 on APS-C is equivalent to F/1.8 on FF (or any other format), when talking about exposure. They didn't talk about exposure, though, but about the total amount of light illuminating the sensor. If the amount of light per unit area is the same, then the exposure is the same (given the same ISO), but the total amount of light falling on the sensor is NOT the same, since the sensors have different sizes. The signal-to-noise ratio (and hence IQ) depends on the total amount of light; that's why this is relevant.
Read the text carefully one more time, and see if you understand the difference.

Comment edited 2 times, last edit 3 minutes after posting

11 upvotes

@ Stephen_C:

"Do the people at DPreview understand the concept of the f-stop?"

Yes we do.

"An f1.8 lens gives the same exposure no matter how large the light circle. When I shoot medium format the exposure is not affected by the final crop I use of the image. An f-stop of f1.8 means f1.8. The f stop does not go down if you use a smaller sensor or crop the image."

I'm glad we agree. As nothing I wrote was inconsistent with that, can explain what, exactly, you think I got wrong?

Comment edited 3 minutes after posting

2 upvotes

> f stop does not go down if you use a smaller sensor or crop the image.

this is 100% correct for unit area. things will happen exactly the same within that unit area, say 1 sqcm, no matter how sensor size changes.

but 1 sqcm is 11.6% of a 35mm format sensor,
and it's 27.2% for APS-C (D7100). so you are comparing 11.6% from one image to 27.2% of another.

what do you want to say?

Comment edited 2 times, last edit 2 minutes after posting

1 upvote

It's not that you have it wrong, it's that it is articulated in a way that makes it sound like you're saying the opposite of what you mean.

Whatever your intent, the phrase "project an image that's just over twice as bright onto a sensor that's slightly less than half the area" makes it sound somehow that smaller sensors are less sensitive than big sensors.
I think the trouble comes when you talk about the area of the sensor. It's not the area of the sensor that's the problem, it's the amount of magnification.

If you're mathmatically minded you instantly understand that a picture on a smaller senor requires more enlargement than a larger sensor and you magnify the noise along with the image.

For the rest of us, it's an almost impossible mental leap from point A to B. Maybe say "a larger aperture allows a lower ISO so pictures will show less noise when printed large. Because you have to enlarge APS-C photos about 1.X times more than you do full-frame pictures..." and so on.

0 upvotes

@rfsII
Define sensitivity. If you define sensitivity as the SNR at a standard output size, smaller sensors are less 'sensitive'.
The question is always whether you mean sensitivity per square millimetre or sensitivity per whole image.

The other point is that if you ask people why ISO 1600 is noisier than ISO 800 (for essentially any camera), as long as people don't answer, "Because the sensor receives half as many photons at ISO 1600.", they will have problems understanding cross-format comparisons.

2 upvotes

What you have to understand is that 95 percent of people do not have the level of technological understanding that you do.

Given the endless misunderstandings and flame wars on this site about sensor size/f-stop it is time for the DPR folks to come up with a different set of words to explain the relationship they're trying to articulate.

The way DPR writes about it currently has a demonstrated ability to lead people to the incorrect conclusion that smaller sensors are inherently less sensitive. We know the readers are wrong, but it's not their fault. They're the victims of the English language and its propensity for ambiguity and misunderstandings.

What makes it even more confusing is that there are times when a lens does transmit less light than its rated f-stop, as in view cameras when the bellows are extended and with some macro lenses.

The exposure triangle is very hard to understand for a lot of people. Why make it harder?

0 upvotes

if you mind to do a brain game, imagine we have two sensors of the same pixel count. connect each pixel from one sensor (the upper) to another (the rubber) using lossless fiber.

now project image on the upper and change the size of the rubber. no matter how you change you get exactly the same image. f-number, ISO, unit area exposure are all meaningless from the rubber's point of view. none of these have any photographic meaning.

Comment edited 2 times, last edit 13 minutes after posting

0 upvotes

@rfsIII,

I understand the "amount of magnification" which we have been using for ages but I would rather think the image quality does not change whatever the magnification because I don't use unit area as reference.

I use the area portion of the frame, that I always compare, say, 10% of one image to 10% of another. or it's like a paint on a balloon, it's the same (quality of) paint regardless of the size of balloon.

0 upvotes

APS-C (crop format) is absolutely not dead. At least in Japan it is still the majority of DSLR camera's sold, probable in other places around the world as well.

Nice fast and very useful focal length.
Makes you also wonder if Sigma will complete this with a 35-70 or 35-105mm fast lens.

5 upvotes

sure it's not dead yet.

0 upvotes

It also makes sense for Sigma because all their own DSLRs have APS sensors.

Comment edited 8 seconds after posting

2 upvotes

Sigma produces DSLRs not for profit so it makes no sense here (other than weired marketing). or they better make this lens for SA mount only.

Comment edited 29 seconds after posting

0 upvotes

The reason APS-C was developed was because: 1) it was cheaper to manufacture than FF, and 2) it resulted in smaller cameras. The trends in FF sensor pricing suggests that in not so remote future FF sensor - APS-C sensor price difference will become small enough to not matter anymore. And one can see the problem with the second reason (size and weight) using this new lens from Sigma as an example: once you try to actually match FF cameras IQ, DoF etc., by developing such an extreme lens, you'll end up with your camera+lens setup at comparable weight and size as a FF camera setup. My prediction: APS-C will die fairly soon, and we will have sub-400$ point-and shoots with FF sensors (the way we had sub-400$ FF point-and-shoots back in the film days), and semi-pro or pro DSLRs/mirrorless with FF sensors. Smaller sensors will only be found inside mobile devices (smart phones, tablets etc.)

0 upvotes

APS-C SLRs definitely worth to die. 35mm SLRs will follow and we will live on a planet of mirrorless.

0 upvotes

Yeah, the Sony RX-1 is good indicator that we are marching really fast towards sub $400 point and shoots with FF sensors.

A 24 x 36 mm large piece of silicon chip is not going to get tumble in price to a tenth of what it costs now any time soon. Electronics get cheaper because we are able to cram more transistors per square millimetre on chips, not because the square millimetre of chip is getting cheaper.

2 upvotes

cost of the sensor is only a fraction of the total cost of the camera. double or half the sensor cost won't wildly change the final cost. this is totally different from the old days of Kodak sensors. good Japanese beat mighty Kodak to death.

0 upvotes

as noirdesir said, square millimetre of chip is not really getting any cheaper. Biggest mass production silicon waffles now are 300mm. Intel is doing first steps to begin production of 450mm waffles. It's just that you can put more APS-C sensors there than FF. Also, consider rejection rate. If you have let's say 6 defects per waffle, you can loose up to 6 sensors per waffle. With APS-C you have much better ratio of good sensors to total produced (loss of 6 sensors is not that crucial). This is why price difference will never be small.
Now imagine how Medium Format must feel..

1 upvote

we do get much cheaper sensors than before. the sensors are now manufactured similar as DRAMs, far cheaper than CCDs.

actually the CMOS sensors (C for Canon) already made landscape change, it's behind the large sensor boom of 35mm full-frame down to P&S cameras that the theory by noirdesir and Silvarum cannot explain.

and we are going to save more and more, like LPF, color filters, and in the future even the micro lenses (the sensor design may be changed dramastically when pixel sizes go smaller beyong optical resolution).

Comment edited 2 times, last edit 4 minutes after posting

0 upvotes

aps-c will always be that son of an ff sensor pretending to be ff. differently from 4/3 and 1", it will never be one of a kind format. i will always see it as a "desperate" attempt from manufacturers to step into (and get the development in) digital faster. but, no denying this is an amazing addition, though.

0 upvotes

APS-C is the leading system-camera sensor. Most units sold, and fastest developement. APS-C leads the curve when it comes to sensor technology, it usually takes a couple of years for said technology to reach full-frame, or smaller sensor sizes. A good example is the K-5 and D7000. Only last year did that technology hit the OM-D and D800.

6 upvotes

it is the leading because the manufacturers needed to use a smaller portion of a ff size in order to face less problems about image quality, and thus, make cameras available quickly into the market.

aps-c (and that early canon 1.3 x) was thought out not with the users in mind, but the companies interests only. well, it's only my opinion. i am from the ff time. if you're not, that's another good reason you not to share the same opinion.

maybe, if you look back, you can tell aps-c digital was born on its own and derived from the aps film format, but aps helped to match the companies main objectives, that was to use a smaller portion of a ff sensor.

that is so, that canon proved 1.3 x crop would be enough to get rid of most optical and cost problems. but then, they (canon and others) did like, "why not making it like derived from aps film format? then we could be justified of why using an even smaller portion of a ff.

0 upvotes

this is similar to 16:9 (1.78 x) screen format for tv. rectangles derived from the multiplication by square roots of 2 are visually more comfortable for the eyes (for example: 1.2 x, 1.41 x, 1.71, 2 x, 2.39 x,...).

so, why didn't tv makers choose 1.41, 1.71 or 2 x? why didn't they choose at least a format matching a movie format, like 1.67 or 1.85 x? they didn't because of convenience for themselves, and clearly not thinking about the user.

until today i don't see much logic (in the user sense) in this choice.

0 upvotes