1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

pentax 10-17mm or sigma 10-20mm

Discussion in 'Pentax Chat' started by proview1122, Dec 31, 2007.

  1. proview1122

    proview1122 Member

    I know there has been a lot of info on this subject which to purchase does it compliment other owned lenses etc. My question is how similar is the pentax at 17mm to the sigma at 20 the FOV data in the specifications are quite similar. how do the pics compare, I guess that the pentax is now straightening out the fisheye curves and so is more like the sigma at 10mm, DOF will be a bit different but I can live with that.
    I guess I have just about talked myself into the pentax unless anyone can point me otherwise.

    thanks in advance
    Happy new year
  2. johnriley1uk

    johnriley1uk Well-Known Member

    The two lenses are very different in their design. The Sigma is intended to be rectilinear and keep straight lines as straight lines. The Pentax 10-17mm is a Fisheye zoom and will give a 180 degree image with lots of distortion. However, if you do lanscapes and put the horizon centrally then no distortion need be apparent. I'll try and post a picture here that shows this:


    Pentax 10-17mm with some added Photoshop Texture

    Moderator's note: image converted to a link - please remember the image size rules, thanks.
  3. john_g

    john_g Well-Known Member

    As John has said, the Pentax lens you mention is a fisheye and the results it'll give will be very different when compared to the Sigma... the former will give images that are a bit like looking through a security spy-hole in a door whereas the latter will give results that look like normal photos but with a greater field of view. The Pentax lens that is closest to the Sigma is, I think, the 12-24mm.
  4. proview1122

    proview1122 Member

    thanks for the replies
    I thought that, as the pentax moves toward the 17mm it gets less fisheye (straighter lines. Well that's the impression I got viewing the few photos I can find at the 17mm endgiving a similar field of view to the sigma at 10mm. So I thought that the pentax would be of more use to me at the long end of zoom (very wide angle) then the more pronounced fisheye look at the 10mm end as and when needed.
  5. john_g

    john_g Well-Known Member

    That's an interesting thought. I've no idea whether it does work that way - hopefully someone will know the answer.
  6. john_g

    john_g Well-Known Member

  7. johnriley1uk

    johnriley1uk Well-Known Member

    The pentax lens is actually a full frame fish-eye, so fills the frame and doesn't give a circular image.
  8. proview1122

    proview1122 Member

    But if you look and compare at the 3rd row down second photo named IMGP3030.jpg on the pentax 10-17 review and 2nd row down first photo named IMGP6831_acr37.jpg to me they have the same look about them the pentax at 17mm and the sigma at 10mm. at these settings the field of view is approx 100 degrees for both lenses

    Does anyone have the same photo taken at 17mm with the pentax and 10mm with the sigma? That would perhaps allow me to make a decision.
    thanks again
  9. john_g

    john_g Well-Known Member

    I've not looked at the images because I can accept they'll look superficially similar but if you look at the distortion grids, you'll see that the Sigma is attempting to produce a rectilinear view where the grid will show as perfect squares, whereas the Pentax 10-17mm leaves the distortion uncorrected.
  10. El_Sid

    El_Sid Well-Known Member

    At 10mm the diagonal angle of view is 180°, by 17mm this has reduced to 100°; almost 45% less. This means that at 10mm the diameter of the image circle is effectively the same as the diagonal dimension of the sensor. When the lens is zoomed out the effective diameter of the image circle is considerably increased and the diagonal dimension of the sensor now represents a much smaller portion of the image diameter. Now as the diameter of the image circle increases the radius of any arc within that circle - any straight line or edge that does not lie exactly on the diameter if the image circle will be an arc - also increases. As the sensor is now cropping a smaller portion of the image circle it is also cropping a smaller, shallower, portion of that arc meaning that the visible part will appear comparatively straighter...

    A related effect occurs with full frame rectilinear corrected lens when used on APS format cameras. Any barrel/pincushion distortion inherent to the lens appears less with the smaller frame camera because a smaller area of the image circle is being recorded. This is particularly apparent with wide angle lenses where the amount of correction required to produce a flat image with straight lines is quite significant...

    If you want to see how the image circle changes size on zooming get hold of a cheap simple zoom such as the Sigma 18-50 f3.5-5.6 and use it on a film camera. This lens changes focal length by simply moving the whole optical cell away from the sensor/film with the result that if you start at 18mm on a film body and zoom out you can actually see the image circle getting bigger. I had one and at 18mm on my film EOS you could see almost all the circle (cropped a bit top and bottom) and as the lens was zoomed out the image circle got bigger until at around 24mm the dark corners vanished. At 24mm the lens was borderline for use with film but from 28mm onwards the image circle was big enough to use without problem.

    Note that with more sophisticated zooms only certain groups of elements, rather than the whole unit, are moved, relative to each other, and the cropping effectively takes place within the lens body - the image circle at the sensor doesn't change - so it's not possible to see the effect described above...
  11. johnriley1uk

    johnriley1uk Well-Known Member

    The pentax 10-17mm never becomes rectilinear, even at 17mm. A door, for example, would look as though it was quite severely barrel-shaped.

    The landscape I have posted a link to though does not show the distortion because there are no straight lines to give it away and the horizon is straight because it goes straight through the middle of the field of view.
  12. proview1122

    proview1122 Member

    is there another lens I should consider that gives a wider angle and is a rectilinear lens
  13. Done_rundleCams

    Done_rundleCams AP Forum Ambassador to Canada

    Morning P_V,

    I don't believe there is anything wider in a PK mount then the Sigma 10-20, rectilinearly speaking :).


  14. tazio35

    tazio35 Well-Known Member

    From reading those reviews, the Sigma seems to be a better bet anyway - less of the purple fringing issues. The only negative point I can see is that it has a slower max aperture.
  15. El_Sid

    El_Sid Well-Known Member

    I'd have to concur... I recently had occasion to take shots of the same view with both my Sigma and the equivalent Canon 10-22mm. While the Canon, I think, is marginally sharper at the centre the Sigma was much better on chromatic abberation... particularly toward the corners. I was quite surprised at the Canon in this repect...
  16. HelenEdith

    HelenEdith Well-Known Member

    I've got the Sigma 10-20 and it does a good job for me.



    It can be wise to check the histogram when using it, though, as sometimes the metering is a bit strange. (Most often under-exposure if it doesn't get it right.)


Share This Page