Review: Sigma 24-70mm f/2.8 DG OS Art by TDP

[Canon Rumors Guy]

The-Digital-Picture has completed their extensive review of the Sigma 24-70mm f/2.8 DG OS Art series lens.</p>
<p><strong>From TDP:</strong></p>
<blockquote><p>The promise of Sigma Art lens quality coming to the 24-70mm focal length range along with an f/2.8 max aperture and optical stabilization had a large number of photographers signing up to add this lens to their kits immediately on announcement day. Does the Sigma 24-70mm f/2.8 OS Art Lens fulfill Sigma’s “It’s Here to Steal the Spotlight” promise? From the perspectives of many, it was hoped that this lens would get that spotlight for its exceedingly sharp image quality performance and from that perspective, I’m sure that there is some let-down among the hopeful. However, this lens is optically competitive and it performs quite well for the price. <a href="https://www.the-digital-picture.com/Reviews/Sigma-24-70mm-f-2.8-DG-OS-HSM-Art-Lens.aspx">Read the full review</a></p>

</blockquote>
<p>This is definitely a lens that should be on your shortlist if you’re looking for an affordable 24-70mm f/2.8 stabilized lens.</p>
<span id="pty_trigger"></span>

 

[Jopa]

I'm glad Bryan's saying the AF is fast and accurate, but @ 24mm the corners look quite bad compared to the Canon 24-70 II, and even the center is not as good... I think they should have called it just "24-70" or maybe "24-70 Contemporary", but not Art.

https://www.the-digital-picture.com/Reviews/ISO-12233-Sample-Crops.aspx?Lens=1119&Camera=979&Sample=0&FLI=0&API

 

[jolyonralph]

Looks poor at f/2.8 - which means that you might as well just get the Canon 24-70 f/4L IS instead.

 

[ahsanford]

Looks poor at f/2.8 - which means that you might as well just get the Canon 24-70 f/4L IS instead.

Especially when you consider that the Sigma's 0.7x macro mode is underwhelming... in that it does not exist.

That f/4L IS: Sealed + IS + first party AF (ring USM to boot) + 0.7x macro = a heck of a value.

- A

 

[ahsanford]

I'm glad Bryan's saying the AF is fast and accurate, but @ 24mm the corners look quite bad compared to the Canon 24-70 II, and even the center is not as good... I think they should have called it just "24-70" or maybe "24-70 Contemporary", but not Art.

https://www.the-digital-picture.com/Reviews/ISO-12233-Sample-Crops.aspx?Lens=1119&Camera=979&Sample=0&FLI=0&API

That's interesting, as LensTip found the 70mm end to be more problematic than the 24mm end:

https://www.lenstip.com/507.4-Lens_review-Sigma_A_24-70_mm_f_2.8_DG_OS_HSM_Image_resolution.html

Also, under their new more stringent AF protocol, this is what they found:

"The accuracy performance we assess positively, both in the studio and outdoor. For the most difficult combination of the 70 mm focal length and f/2.8 aperture, with the lens positioned before a testing chart, you might count on 88% of hits accurate within just 5% of difference from the best result. A bit worse hits but still completely acceptable (from 5 to 10% of difference from the best result) constituted 7% of all shots.There were 5% of shots considered by us to be totally missfocused (different by more than 20% from the sharpest photo). The results at the shorter end of the focal spectrum were even better due to the increase of the depth of field."

You can read that two ways -- the new LT hit rate test is indeed a high bar, or this Sigma lens whiffs 5% of the time. Again, without a *proper* AF hitrate database, these one-off tests on one lens are only so useful. But caveat emptor, test your new gear before you can't return it, etc.

- A

 

[Ryananthony]

Looks poor at f/2.8 - which means that you might as well just get the Canon 24-70 f/4L IS instead.

Especially when you consider that the Sigma's 0.7x macro mode is underwhelming... in that it does not exist.

That f/4L IS: Sealed + IS + first party AF (ring USM to boot) + 0.7x macro = a heck of a value.

- A

Couldn't agree more. You're also forgetting the almost full pound in weight savings over the Sigma. If one can live with out the f2.8 aperture, the 24-70f4IS looks to be a great purchase.

 

[ahsanford]

Looks poor at f/2.8 - which means that you might as well just get the Canon 24-70 f/4L IS instead.

Especially when you consider that the Sigma's 0.7x macro mode is underwhelming... in that it does not exist.

That f/4L IS: Sealed + IS + first party AF (ring USM to boot) + 0.7x macro = a heck of a value.

- A

Couldn't agree more. You're also forgetting the almost full pound in weight savings over the Sigma. If one can live with out the f2.8 aperture, the 24-70f4IS looks to be a great purchase.

We are drifting perilously into f/4L IS fanboyhood.

Let's get back to the Sigma, apologies.

Presuming this doesn't stack up to the f/2.8L II, how does it fare against it's true rival -- Tamron -- to be '#1 in the #2 business'?

- A

 

[slclick]

And the 24-70L ii remains , YES EVEN WITHOUT STABILIZATION, one of the finest lenses I have ever owned.

 

[Diko]

Looks poor at f/2.8 - which means that you might as well just get the Canon 24-70 f/4L IS instead.

Why is even the talk here?
Canon 24-70 f/2.8 no IS will set you back $1k more.
Why would anyone want to get f4 for 400 less just for the sake of IS? They are mutually excluding each other.
This one is hard to compare since it would provide f/2.8 (could use or not the stabilization).

Having IS and 2.8 for THAT price is perfect for evening events. And considering the better always improving algorithms for removing vignette and distortions... This talk here can be considered abomination.

 

[Jopa]

I'm glad Bryan's saying the AF is fast and accurate, but @ 24mm the corners look quite bad compared to the Canon 24-70 II, and even the center is not as good... I think they should have called it just "24-70" or maybe "24-70 Contemporary", but not Art.

https://www.the-digital-picture.com/Reviews/ISO-12233-Sample-Crops.aspx?Lens=1119&Camera=979&Sample=0&FLI=0&API

That's interesting, as LensTip found the 70mm end to be more problematic than the 24mm end:

https://www.lenstip.com/507.4-Lens_review-Sigma_A_24-70_mm_f_2.8_DG_OS_HSM_Image_resolution.html

Also, under their new more stringent AF protocol, this is what they found:

"The accuracy performance we assess positively, both in the studio and outdoor. For the most difficult combination of the 70 mm focal length and f/2.8 aperture, with the lens positioned before a testing chart, you might count on 88% of hits accurate within just 5% of difference from the best result. A bit worse hits but still completely acceptable (from 5 to 10% of difference from the best result) constituted 7% of all shots.There were 5% of shots considered by us to be totally missfocused (different by more than 20% from the sharpest photo). The results at the shorter end of the focal spectrum were even better due to the increase of the depth of field."

You can read that two ways -- the new LT hit rate test is indeed a high bar, or this Sigma lens whiffs 5% of the time. Again, without a *proper* AF hitrate database, these one-off tests on one lens are only so useful. But caveat emptor, test your new gear before you can't return it, etc.

- A

Any idea what's a hit rate for the 24-70 2.8 II ? I just can't understand if this 88% is good or not? Theoretically it should be a 100, but with PDAF there is always a chance to miss.

 

[Jopa]

^ found it.

24-70 II

"The accuracy of the mechanism deserves also a lot of praise. In studio conditions it missed in just 2% of shots which is a very good result."

That's nice they have real world AF accuracy test results. It's not possible to use their website search though, had to use google.

 

[ahsanford]

^ found it.

24-70 II

"The accuracy of the mechanism deserves also a lot of praise. In studio conditions it missed in just 2% of shots which is a very good result."

That's nice they have real world AF accuracy test results. It's not possible to use their website search though, had to use google.

Yes and no. LensTip recently improved their focusing system from coarse hits / misses -- which was generally favorable towards a hit -- to a graded in focus / just out of focus / true miss. I believe they did this deliberately with the Sigma... 85 Art (was it?) to specifically see if Sigma had upped their game. They've been using it fairly consistently since that time.

So comparing results from both pre & post change may not be so telling. That said, I'd have no doubt a flagship first party offering like the f/2.8L II was a faster/better/more consistent overall focuser.

- A

 

[SecureGSM]

Unless you are aware that Sigma Image Stabilisation algorithm of Art and Contemporary series is not that great (understatement). Couple this with subpar AF performance in low light (EV 5.0 and below) to compliment the picture.

Having IS and 2.8 for THAT price is perfect for evening events. And considering the better always improving algorithms for removing vignette and distortions... This talk here can be considered abomination.

 

[Diko]

subpar AF

Sorry... what is subpar AF... really english is not my native :-[ Not kiddin' you.

Thanks in advance.

 

[SecureGSM]

subpar - A lower standard than customary or traditionally accepted norms, but not entirely unacceptable. For example, a computer's speed may be subpar, indicating that it is running slow relative to other computers, but this does not necessarily mean it is broken or unusable.

http://www.businessdictionary.com/definition/subpar.html

in other words, it is not the best in low light conditions

subpar AF

Sorry... what is subpar AF... really english is not my native :-[ Not kiddin' you.

Thanks in advance.

 

[snoke]

Need shootout. Tamron 24-70/2.8 ABCDEFGH vs Sigma 24-70/2.8 IJKLMNOP. Which is better?

 

[SecureGSM]

https://www.youtube.com/watch?v=hh_r8Pe2wb0

https://www.youtube.com/watch?v=rOpDkPWpMFw

Need shootout. Tamron 24-70/2.8 ABCDEFGH vs Sigma 24-70/2.8 IJKLMNOP. Which is better?

 

[FramerMCB]

https://www.youtube.com/watch?v=hh_r8Pe2wb0

https://www.youtube.com/watch?v=rOpDkPWpMFw

Need shootout. Tamron 24-70/2.8 ABCDEFGH vs Sigma 24-70/2.8 IJKLMNOP. Which is better?

This 2-part series by D.A. comparing the 3 lenses is great. Really gives one good insight into their strengths and weaknesses. Bottom line seems to be, for real world use where most of your shooting is hand-held the Tamron looks like the lens to have. If you have a body with an excellent sensor with in-body stabilization, like the A7r II that Dustin references Paul has (the fellow that loaned him is Canon 24-70mm 2.8L II), then the more obvious choice is the Canon.

A few other aspects to test would be how do all three handle flaring, coma, aberrations? TDP indicates the Sigma doesn't fare well. (Or was that LensTip?) Regardless, zoom lenses by nature (physics) struggle more to control flaring.

 

[jd7]

I see Dustin Abbott has posted his full review of the Tamron 24-70 G2 now (and he posted his Sigma Art review a while back)
https://dustinabbott.net/2017/10/tamron-sp-24-70mm-f2-8-di-vc-usd-g2-review/

Dustin seems very happy with his 24-70 G2 - and he gives it the edge over the Sigma Art - but I'm sure I saw someone here (Jopa?) say recently he bought and returned one because he wasn't happy with the AF. Anyone else given the G2 a go?

 

[jd7]

After watching various parts of Dustin Abbott's two videos (the ones linked above), to my eye (based the images in the videos) I'd rank the lenses in the following order for IQ: 1. Canon 24-70L 2.8 II, 2. daylight, 3. Tamron 24-70 G2, 4. Sigma 24-70 Art (although pretty close to the Tamron really).

Whether better IQ when viewed at 100% on a computer screen is going to be noticeable in a print or when the whole image is viewed on screen, and whether the VC/OS of the Tamron and Sigma mean you are more likely to achieve better IQ in real world shooting (will depend on how often you shoot handheld at relevant shutter times), are good questions - and of course there is a cost difference - but my enthusiasm for the new Tamron has cooled a little.