Sigma 12-24mm v Sigma 15-30mm v Canon 17-40mm L
v Contax Zeiss 21mm Distagon f2.8
[published 2004]
ISO 100 / Full frame Canon 1Ds Mark I with mirror lock up, f8 / 1 second / AWB on a Manfrotto 155 with view camera head, legs locked. All were shot RAW and processed in 8 bit via Capture One to TIFFs and converted for web with BoxTop Pro JPEG. Lighting: single 500W head.
Centre Frame – hanging POS (mid ground) resampled from 270px to 350px
| Sigma 15-30mm @ 21mm | Canon 17-40mm @ 21mm | Zeiss 21mm | Sigma 12-24 @ 21mm |
 |
 |
 |
 |
Foreground
| Sigma 15-30mm @ 21mm | Canon 17-40mm @ 21mm | Zeiss 21mm | Sigma 12-24 @ 21mm |
 |
 |
 |
 |
Edge Frame 1
| Sigma 15-30mm @ 21mm | Canon 17-40mm @ 21mm | Zeiss 21mm | Sigma 12-24 @ 21mm |
 |
 |
 |
|
Edge Frame 2
| Sigma 15-30mm @ 21mm | Canon 17-40mm @ 21mm | Zeiss 21mm | Sigma 12-24 @ 21mm |
 |
 |
 |
 |
Corner 1
| Sigma 15-30mm @ 21mm | Canon 17-40mm @ 21mm | Zeiss 21mm | Sigma 12-24 @ 21mm |
 |
 |
 |
 |
Corner 2
| Sigma 15-30mm @ 21mm | Canon 17-40mm @ 21mm | Zeiss 21mm | Sigma 12-24 @ 21mm |
 |
 |
 |
 |
What to make of this?
The outstanding result is confirmation of the overwhelming supremacy of the Zeiss 21mm – especially in the corners.
The Sigma 12-24mm is clearly the poor relation of the bunch, though it is very consistent in its centre/corner performance (unlike the Canon) – just uniformly softer across the frame. None of the lenses tested showed significant CA.
The Sigma 15-30mm matches or maybe even outresolves the Zeiss 21mm in the frame centre, though it can’t compare to the Zeiss’ edge performance.
The Sigma 15-30mm may have been focused slightly further away than the Canon 17–40mm – which explains the Sigma’s poor foreground showing – and reveals just how terrible the Canon is at the frame edges and corners. Bizarrely the Canon matches the Sigma at the right mid-ground edge, but fails badly at the left foreground edge – freaky, if the Canon’s focus is closer to the camera? Lousy close-range correction, or decentring?
In conclusion, I am hugely disappointed by the second 17-40mm L I have played with in this way, and hugely impressed by the Sigma 15-30mm – just see how sharp it is at 15mm here. As longas you don’t compare the 12-24mm to a Zeiss prime, I think many people are going to be very happy with the world’s first full frame 12mm lens – the image it produced wasn’t conspicuously soft even at A3 / 300dpi. It’s just that the Distagon 21mm is in a class of its own, and the Sigma 15-30mm goes wider, costs less, and performs much better than the 17-40L – at least this one did . . . .
Sigma 12-24mm v Sigma 15-30mm at 15mm/f8
All images shot with a Canon 1Ds Mark I at 100ASA. 1/15 sec at f8 in RAW, processed in 8 bit via Capture One and converted for web with BoxTop Pro.
Frame edge (top)
| Sigma 15-30mm @ 15mm | Sigma 12-24mm @ 12mm |
 |
 |
Frame corner (bottom left)
| Sigma 15-30mm @ 15mm | Sigma 12-24mm @ 15mm |
 |
 |
Frame centre (unresampled)
| Sigma 15-30mm @ 15mm | Sigma 12-24mm @ 15mm |
 |
 |
Another result for the 15-30mm. Corner and centre frame detail is more legible in these 15mm captures by the Sigma than in 17mm shots with the Canon 17-40mm, and clearly better than this 12-24mm sample.
