Sigma 10-20mm F3.5 EX DC HSM A-mount lens review by Pallanza

reviewer#10926 date: Feb-17-2013
sharpness: 4
color: 4
build: 1
distortion: 4
flare control: 4
overall: 3.4
tested on:
  • film camera:Film camera
  • APS-C: 6MP6 MP; 10MP10 MP; 12MP12 MP; 14MP14 MP; 16MP16 MP; 20MP20 MP; 24MP24 MP
  • full frame: 12MP12 MP24MP24 MP36MP36 MP42MP42 MP
ownership:I own this lens
compared to:SAL 16105
price paid:600 EUR (new)
positive:Wide angle, build / look & feel.
negative:Has to be stopped down to 5,6 or 8 to get sufficient sharpness. Quality risk.
comment:According to some tests, the sharpness of the Sigma should be comparable to that one of the SAL16105. SLR-Gear shows in its blur index, that the Sigma were even sharper in the corners. Photozone says the opposite. User reviews often moan about bad copies, mainly as to sharpness.

Thus I ordered three copies and compared them with a little test. I pinned a newspaper (the stock exchange tables with their small characters) on a wall and shot it (tripod, level). For each lens copy I set the microadjustment new. A77, raw, auto focus controlled with focus peaking. Processed with DxO Optics Pro 8 to 100%-jpg, using DxO's optical module. Some tests I did double and thrice. Jpgs checked in ThumbPlus7 mostly at 100%, comparing two or more images simulaneously.

Copy A was really bad. Center was ok, but in the corners the newspaper wasn't legible, just black and grey blotches. At 100%, I found in some parts of the images a doubling of contours. Overall: Not usable.
Copies B and C were better than A, but still a bit less sharp than the SAL16105 at 16mm. Center was nearly as sharp as the SAL16105. But edges and corners were mostly significant worse. Stopping down to 5,6 or 8, sharpness of the Sigma is acceptable in the corners too. In practical shots I got the same results.

Thus I cannot confirm the SLR-Gear's blur-index as to the corners of the Sigma 3,5 10-20 mm. This my opinion suits to the results of photozone and lenstip and many user reviews.

If there is enough light for stopping down, the Sigma 3,5 10-20 mm seems to be an acceptable or good lens, and regarding its extreme focal width, even a very good lens.

As to color rating and flare control I do not have any special remarks - it's winter with fog and clouds, any colors out there. Distortion (barrel, cushion) is no real problem for me, using DxO's optical modul. So for these items I give the average points.

Vignetting / dropoff in the corners: with DxO's optical modul not visible. Without: really visible, should be compensated.

Build: At Dyxum's, "build" is defined mainly concerning the mechanical use resp. look and feel of the lenses. As to this, the Sigma 3,5 10-20 mm feels good, worthy. Controls are ok. So normally I would give 5 points.

But as one of three lenses is not usable and two further ones also differ in sharpness, I cannot give 5 points for build of this lens in general.
My little personal experience confirms, that there is a real variation of the technical quality of those lenses. Comparing the tests in the net, you'll find different results for this lens. May be, some testers do use lenses "tuned" by the manufacturer (not unusual in testing business!!). One tester (lenstip) therefore says, they took their testing sample from a dealer and not from the manufacturers representant as usual.

Thus for "build" I want to give only two points.
May be, this is not really conform to Dyxum's lens testing guidelines. But the variance of Sigma's quality really seems to be problem, and this should be stated.
As a customer, I want to get what has been advertised.



rating summary

lens image
  • total reviews: 15
  • sharpness: 4.43
  • color: 4.53
  • build: 4.40
  • distortion: 4.20
  • flare control: 4.20
  • overall: 4.35
Dyxum.com - Home of the alpha system photographer
In memory of Cameron Hill - brettania