[rawr]

WOW! DxOMark - Sigma 18-35mm F1.8 DC HSM A Canon

!!!

I've never seen a DxOMark result with this much excellent green in it, especially at wide open apertures, across all focal lengths. Breaks all the rules.

Attached Images

 

[vodanh1982]

Very impressive. If I were not waiting for the NEX-FF, I would have hit the buy button.

[falconeye]

I've never seen a DxOMark result with this much excellent green in it, especially at wide open apertures, across all focal lengths. Breaks all the rules.

It is definitely a great result. Esp. as writes DxO, because it beats quite a few primes in that range.

However, I have trouble to properly read DxO lens tests and I know nobody who hasn't. Their P-MPix measure is an undocumented weighted mix of sharpness values and the score includes the amount of light captured which is why fast lenses and cameras with a high DxO score produce better lens scores even if the measured sharpness is the same. That too cryptic even for my taste

I normally prefer DxO data over DPReview. But in the case of DxO lens tests, I prefer the reformatted versions by DPReview.

In particular, I looked at
Lens review data: Digital Photography Review

This is a comparison between the 18-35/1.8 (7D) and a 24-70/2.8 (5DmkII) where the MP are similiar (18 vs. 21). I compare equivalent focals and apertures.

And indeed, the 18-35/1.8 is an excellent performer, bringing a level of optical quality to APSC which was formerly confined to FF. Between (apsc) 18-24 and wide apertures, both lenses have similiar resolution. Only at (equivalent) F5.6 and/or 50mm has the 24-70 a (very) slight edge. And at 24-26/51-70mm of course.

So, it is breaking two rules:
-A zoom is worse than a prime (not)
-An APSC camera/lens resolves worse than a FF (not)
Of course, both rules still apply in a more general way, just not for this lens

BTW, the test measures the lens' T-Stop. The Sigma has a T-Stop of T/1.8 and T/1.9, depending on the focal length. The Canon 35/1.4 lens has T/1.7 and a T-Stop 0.3 above the F-Stop is pretty normal.

So, this Sigma lens can be regarded a 18-35/1.5 when it comes to its low light capabilities.

[vodanh1982]

It is definitely a great result. Esp. as writes DxO, because it beats quite a few primes in that range.

However, I have trouble to properly read DxO lens tests and I know nobody who hasn't. Their P-MPix measure is an undocumented weighted mix of sharpness values and the score includes the amount of light captured which is why fast lenses and cameras with a high DxO score produce better lens scores even if the measured sharpness is the same. That too cryptic even for my taste

I normally prefer DxO data over DPReview. But in the case of DxO lens tests, I prefer the reformatted versions by DPReview.

In particular, I looked at
Lens review data: Digital Photography Review

This is a comparison between the 18-35/1.8 (7D) and a 24-70/2.8 (5DmkII) where the MP are similiar (18 vs. 21). I compare equivalent focals and apertures.

And indeed, the 18-35/1.8 is an excellent performer, bringing a level of optical quality to APSC which was formerly confined to FF. Between (apsc) 18-24 and wide apertures, both lenses have similiar resolution. Only at (equivalent) F5.6 and/or 50mm has the 24-70 a (very) slight edge. And at 24-26/51-70mm of course.

So, it is breaking two rules:
-A zoom is worse than a prime (not)
-An APSC camera/lens resolves worse than a FF (not)
Of course, both rules still apply in a more general way, just not for this lens

The DxO scores are funny but the acutance graph is usable. For me, any lens with acutance over 70% at max aperture is a good lens.

[rawr]

The lens is very interesting to me in how it strongly highlights the importance of good optics to realising camera potential. Everyone knew this already, of course, but this lens makes it starkly clear. In front of a good APS-C sensor you could have something quite impressive, for resolution especially. I'd like to see the tests from it mounted on a K-5 IIs.

Of course, I also expect people to stick $10 Tiffen UV or 'protection' filter on the lens, then complain about flare, softness etc etc.

[Aristophanes]

Good results but you're paying for it both in weight and a steep price.

There's no reason why a zoom can't be as good as a prime if you throw enough mass and glass and cash at the design.

[troika]

Hey, you don't have to tell me, I'm from Indianapolis. Ever heard of the Pacers? The Colts?

You're not seriously trying to make a case for the Colts as underdogs, are you? From Payton Manning to Andrew Luck?

[deadwolfbones]

Good results but you're paying for it both in weight and a steep price.

There's no reason why a zoom can't be as good as a prime if you throw enough mass and glass and cash at the design.

$800 is not really a steep price when Nikon and Canon are selling their 17-55/2.8's at well over $1,000.

Hell, Sigma's own 17-50/2.8 is only $230 cheaper.

[Edgar_in_Indy]

$800 is not really a steep price when Nikon and Canon are selling their 17-55/2.8's at well over $1,000.

I agree. Since "expensive" is subjective from person to person, you have to look at the lens relative to other lenses. And this lens has professional build quality with features and performance that are not matched by any other zoom lens out there right now. So Sigma could have charged $1500 for it, and I don't think anybody would have been surprised or said it was completely outrageous.

I think that the lens is a steal at its current price, and I have to wonder if it's going to remain at the price for very long, or if they're just teasing for now to get the word out and to get some people out there using it and preaching the Sigma gospel.

[falconeye]

Good results but you're paying for it both in weight and a steep price.
There's no reason why a zoom can't be as good as a prime if you throw enough mass and glass and cash at the design.

Again, a lack of respect for the optical engineering achievement. First and foremost, you have to throw talent at the design. It isn't replaced by CAM, just like it isn't in architecture.
Moreover, the price is very low and aggressive, actually. The weight is the weight of 35/1.8 mm of glass.

[VoiceOfReason]

The more I see of it the more I like it. It's perfectly suited for the type of shooting I do at anime conventions. Forget a new body if one comes out in the fall, I want one of these. The body can wait.

[Aristophanes]

The more I see of it the more I like it. It's perfectly suited for the type of shooting I do at anime conventions. Forget a new body if one comes out in the fall, I want one of these. The body can wait.

And right here is a summary of what ails the stalled and failing camera industry.

This is exactly why lens prices have gone up, and the K-30 gets re-enacted as the K-50 (and the D7000 as the 7100, and the T3i as the T4i).

Again, a lack of respect for the optical engineering achievement. First and foremost, you have to throw talent at the design. It isn't replaced by CAM, just like it isn't in architecture.
Moreover, the price is very low and aggressive, actually. The weight is the weight of 35/1.8 mm of glass.

I'm not impugning the optical engineering. I am simply saying any optical company with enough resources could do this. There is a tremendous homogeneity in the optical industry now. Differences between manufacturers based on engineering are nil. Canon, Fuji, and even Pentax could have made this lens. They did not for economic reasons.

First, it's a lens that requires an installed base across multiple platforms.

Second, it is a high-priced lens. It's got halo effect, early adopter discounting applied on a lens with a MSRP of $1160. That's right on Sigma's website.

Why do this?

Because dedicated camera sales have stalled or declined. Since you cannot count on many more new customers to the dedicated camera market, you need to discount to move product. Sigma can (in fact has to) do that because they are entirely after-market. That's also why Canikon, Sony, Pentax, Fuji all raised their lens prices recently...to make up for declining unit volumes of cameras. This at a time when the Yen is falling.

Kudos to Sigma for innovating a market niche. However, it is a niche. Look at sales data and what is telling is either a FF large system camera or a small DSLR or mirrorless system camera. The middle is being gutted (seen a D400 on the horizon?). So a large expensive lens for a smaller than FF sensor with a limited FL range would never survive as a Pentax or Canon or Nikon-only offering. It needs to cross brands to survive. The price and size on APS-C are now going in opposite directions from the market data. Lenses like this are no longer driving camera sales. They do not make new customers for the DSLR market.

So it is a niche at even US$800/unit. The APS-C market is going to hit very turbulent times as FF goes downprice and mirrorless steals chunks of the middle. The days of DA* lenses selling for over $1,000 are coming to an end in APS-C because the entire price point of the system is falling. FF is chewing the higher end to pieces.

What Sigma is doing here is swimming upstream where there are still fishermen. They can do this because the price of FF is still too high for many and because there is always a demand for fast glass from prosumers regardless of the size penalty. However, in the long run, big glass for APS-C is problematic. The 18-35 is 300 grams more than the DA*16-50 losing 2mm wide and 15mm long, so Sigma is competing on price and aperture. Nevertheless a huge chunk of consumer activity is going away from big glass for APS-C regardless.

The bulk of APS-C revenues are going to come from smaller system cameras. Big glass, big bodies are going to FF. APS-C DSLR's are going to need smaller bodies (Canon's SL1 and the Pentax K-x get it mostly right...the K-30/50 is IMO a step back in form factor). They will need smaller, lighter, less expensive lenses (DA 35/2.4 and 50/1.8 and 18-135). That's not to say one cannot make money swimming upstream; it's just a smaller market. The actual demand for this glass is probably much less than most expect.

[VoiceOfReason]

And right here is a summary of what ails the stalled and failing camera industry.

This is exactly why lens prices have gone up, and the K-30 gets re-enacted as the K-50 (and the D7000 as the 7100, and the T3i as the T4i).

APS-C happens to be what works best for me for now because I like going long and still having great IQ. FF would probably be good for my anime con stuff, but let's face it. I don't have $2,000+ to drop on a body, then have to go get lenses to work with it. Also, how much of an improvement will the next APS-C body have over my K5 and K-01? Which one would make more sense for me from an image quality standpoint? If Pentax would have released this (at a much higher price I know) I would have saved up to get it. They didn't, but Sigma did. How long will it take Pentax to come up with an answer? How about a truly new camera? The K5 broke ground, but all cameras since then have been using basically the same sensor. Sure, we've gotten improved video (meh, I have a full HD video camera which I use for that), autofocus that works in darker places, and the removal of the AA filter, but no real breakthroughs in it. I'm still going to support Pentax, and in the past year I have with the purchase of a new 18-135, K-01, and DA 40xs, but for me right now glass is the way to go.

[Digitalis]

After taking a look at other image samples from proud owners of this lens I have to agree it is a very good zoom lens - resolution is extremely impressive, astigmatism isn't as bad as I had feared. Though the coma looks pretty nasty - there have been complaints of the "onion ring" effect that appears in OOF highlights, this would be due to Sigma not polishing their aspheric lenses well enough.

[scratchpaddy]

That's not to say one cannot make money swimming upstream; it's just a smaller market.

Sigma loves the smaller market. They're a niche manufacturer; they find holes that the general market didn't even realize were there, and they exploit them.

they are entirely after-market

That's not correct. They have their own DSLR, and three fixed-lens compacts. They invented the fixed-lens APS-C camera market, by the way. Now it's hot stuff.