[Klaus]

Here we go ...

 

http://www.opticallimits.com/canon_eos_f...anon1124f4

[borisbg]

Great review Klaus! That was quite fast. How did you feel manuvaring these two babies on field?

Under the flare section: "e.g. the Nikkor AF-S 14-28mm f/2.8 ED", shouldn't that be 14-24?

[Rover]

Ah... the most eagerly anticipated review in recent memory is here. The 1,5 months of downtime surely helped to stoke the anxiety of the user base.

Looks like a great 14-24mm lens to me, with 11mm for extra craziness when the situation merits it. Pity that the price is so exorbitant though... Tell me Klaus, are you by chance going to retest the Sigma 12-24 on the SR? Then a (faint) point of comparison might emerge...

[soLong]

Deleted.

[Brightcolours]

Thank you for an accurate review yet again Klaus….for me this is a lovely lens on a lovely camera.

 

One thing i suppose that i've never understood about measurements is comparing my field experiences to the charts results. Maybe mine were just poor copies of the lens (i tried 2 copies) -  but, i had a 1635 that measures here like this at 16mm - 

 



but on my camera at that time the corners were totally smeary and unusable. Whereas this lens at 11mm measures like this - 



but i'm not seeing this on your pictures or mine and i'd be happy to use my shots without cropping....or....why i also check field results for myself where possible.

 

And yes this gear is expensive but still lower priced than the high end gear was not so many years ago, while the quality is still higher now….the future of large digital? 

 

….and now i have 11-400mm covered without sharpening :-)

At times the difference can be explained with the MTF testing distance. 51x the focal length makes it 11 x 51 = 56cm subject distance. I bet your shots were focussed on something a bit more distant?

[soLong]

Deleted.

[Scythels]

A typical ISO 12233 chart is more in the range of about 4-5' wide by 3-4' tall.  I do not know how big Klaus' chart is, however.  Testing anything under 20mm on FF you are basically kissing the chart.  11mm looks quite ridiculous. 

 

The corner vs center results may also be influenced by the chart lighting.  This is an issue even the people who wrote imatest do not know how to quantify the effects of.

[Klaus]

Have a look at the sample images - at 11mm the corners aren't hot even at the medium aperture settings.

Thus the charts are in sync here.

 

But yeah, the chart to lens distance was ... interesting. The object magnification is identical across all tests though. Of course, the performance will differ at other magnifications (which is why it always makes sense to take multiple tests into account). However, it is also not a good idea to use a ultra-wide lens at infinity focus (-> hyperfocal distance).

Of course, there's a trick to increase the "size" of the chart - just use one of the corners as the center (increases the resolution by a factor of 4 for one corner). I did that at one stage in the past but the results weren't that different.  We are not talking about object magnifications of sub 1:10 where the situation is very different, of course.

 

Yes, an even illumination of the test charts cost me SOME nerves with this lens ( in fact this aspect is always messy but even more so here). Imatest doesn't like high CAs - this is more relevant than an even illumination of the chart. At 3px at 11mm, the effect is still manageable but I reckon it had a certain impact at the borders - somewhat more in the corners (even higher CAs).

[Klaus]

Just to illustrate - this is a 11mm corner at f/9. 

Thus about as good as it gets.

[soLong]

Deleted.