[TrueFocus]

This lens is fantastic! Really well made, no wobble, squeeks, or etc. It handles very well and has a reassuring feel to it. I like the finish of the lens and also the way how it looks on my K-5.
Optically it is exactly as advertised. Great at f1.8 throughout the range and fantastic once stopped down to f2.5.

Attached are a few pictures taken with it (plus the obligatory cat picture)

[lytrytyr]

the obligatory cat picture

Your cat doesn't seem to be as happy with the lens as you are!

[rawr]

...Optically it is exactly as advertised.

Flare test on Flickr

I see some flare.
Into the trash it goes.
I hope you weren't using a UV filter on your brand new lens?

[Digitalis]

I was planning on doing a head to head comparison with the sigma 18-35mm f1.8 and my SMCP-FA 31mm f 1.8 limited this week, but due to de-centering in my copy of the sigma 18-35mm f/1.8 I have to return the lens. The issue is clearly visible at infinity focus at 18mm f/5.6 - the left side of the image is sharper than the right side. It wouldn't be fair to compare it to my perfectly centered FA31. I have also noticed some other troubling issues with the optical performance of this lens.

I really can't say much in regards to image quality due to the fact that my copy of this lens was optically compromised, however there is some potential. Along with the lacklustre Sigma quality control ( which, according to sigma is more stringent for "C" "A"and "S" series lenses*) I will mention these issues.

My example of lens flare from the 18-35mm f/1.8 at 35mm f/4:



And some disturbing frog egg bokeh from the 18-35mm at close focus distances:





And an example of the de- centering from my lens, 100% crops, un-sharpened from the extreme left of the frame to the extreme right, the more I zoomed out the worse it got.


* Hypothetically speaking I suppose the foveon sensors Sigma use to test their lenses are less sensitive to slight de-centering due to their 3D structure, Bayer sensors being a plane only a few microns deep would be more sensitive to it.

---

Last edited by Digitalis; 07-14-2014 at 08:16 AM.

[TrueFocus]

Cats usually don't like being photographed.
No, I did not use a UV filter.

Digitalis, I did some bokeh tests myself, but didn't see such terrible 'frog eggs'.
I will do some tests and report back.

[Edgar_in_Indy]

Isn't it kind of hard to really assess de-centering without shooting a uniform, flat subject that is parallel to the lens/sensor?

You said the right side is sharper, but in your image you posted the left side (cow side) looks sharper to me.

[Digitalis]

Isn't it kind of hard to really assess de-centering without shooting a uniform, flat subject that is parallel to the lens/sensor?

You said the right side is sharper, but in your image you posted the left side (cow side) looks sharper to me.

sorry, I re-wrote that part to accurately reflect what I'm seeing. Yes the right side is the part which is affected by the de-centering. Ordinarily the best way to test a lens for centering defects is to shoot a uniform textured subject - but that is boring as hell, so I took the lens out to the Australian countryside and used it like it was intended, and I spotted the defect rather quickly under field conditions.


Pentax K5IIs Sigma 18-35mm f/1.8 "A" DC - 100% crops, both un-sharpened. There is a noticeable loss of fine detail and texture in the tree on the right side.



So when I got home this morning, I put the lens on my optics testing bench and projected a star test pattern through it, rotated the lens in the projection mount and by observing how the image shifted you could see that there was something clearly wrong with the lens I had. I returned the lens to the store this afternoon - it will be taken back to sigma tomorrow and either re-built or replaced outright.

---

Last edited by Digitalis; 07-14-2014 at 08:54 AM.

[TrueFocus]

What is it that you changed between those two last pictures?

[Edgar_in_Indy]

Did you flip the camera upside down for one of the two pictures of the tree so you could compare the two sides of the lens?

[Digitalis]

What is it that you changed between those two last pictures?

The only thing I changed was my perspective, however I didn't need to flip the camera upside down to confirm the de-centering because the subject was at infinity focus anyway. All I had to do was maintain the focal length and aperture.

[Edgar_in_Indy]

I didn't need to flip the camera upside down to confirm the de-centering

My bad, I forgot you're in Australia...the camera is already upside down!

[Digitalis]

My bad, I forgot you're in Australia...the camera is already upside down!

haha, relative to you it is, the scene was broad enough for me to capture it using different sides of the lens - so I didn't need to turn the camera upside down. But at any rate, the optics test bench confirmed what I was seeing on the rear LCD of my camera.

[Uluru]

I was observing examples posted around the web and I am not at all overwhelmed with this new Sigma 18-35. In fact, I do not understand what the fuss is all about? Do not people know there is better than that?
Only relatively decent images I saw are those taken in an overcast day / in shadows, but in sunlight, it leaves much to be desired in comparison with Pentax higher-end optics. I still think Sigma cannot hold a candle to Pentax optics, both from the design point of view, and from many finishing touches lacking.

[DennisH]

Funny you say that; I returned mine because I got good shots in total shade, but in sunlight it horribly overexposed the scene. Never had that problem with other lenses.

Also it front focused, but that's another issue .

[Edgar_in_Indy]

I still think Sigma cannot hold a candle to Pentax optics, both from the design point of view, and from many finishing touches lacking.

What has Pentax designed lately that can hold a candle to the new Sigma 50mm f1.4 or 85mm f1.4? You could also probably mention several other great lenses like the 120-300mm f2.8, 500mm f4.5 and 35mm f1.4. Sure, some of Pentax's old designs from the previous century are great, but I've seen no evidence that they still have the ability to design a good lens.

What have the lens gods at Pentax given us this century, other than a painful case of SDM and meaningless lens roadmaps? My opinion is that whatever credibility Pentax had in the past in lens design is now just a memory.

Let's not forget that the current topic of discussion is an 18-35mm f1.8 zoom lens. That's something nobody has done before, and nobody thought it could be done at this price until Sigma stunned everybody by doing it. Despite the grumbling on here, the vast majority of user reports and professional reviews are very positive, so it surely is not a total disaster. Designing a lens with such an extreme specification at that price point and that can still take a decent (but not perfect) picture is a win in my book.

I don't know who was expecting a lens like this to be great at controlling flare, and this is hardly the first lens to have copies that suffer from de-centering.

Anyway, Digitalis's cows looked pretty damn impressed to me.