20mm for landscape

[Frank Fremerey]

Thank you, Mike!

[Ron Scubadiver]

I shouldn't be reading this because I would just love to have a 20/1.8 to replace my 20/2.8 AF (not D).

[Akira]

And now we have the Sigma 20/1.4 Art lens. Hopefully, for astro/nightscapes, it performs slightly better than the Sigma 24/1.4 in the coma and CA department.

Shane, this review may offer some food for thoughts...

http://www.lenstip.com/457.7-Lens_review-Sigma_A_20_mm_f_1.4_DG_HSM_Coma__astigmatism_and_bokeh.html

[Hugh_3170]

Come on Ron - be a devil.  It is nearly Christmas and I reckon that you deserve a new lens.  Actually the 20mm f/1.8 is not unreasonably priced IMHO.   

I shouldn't be reading this because I would just love to have a 20/1.8 to replace my 20/2.8 AF (not D).

[Frank Fremerey]

I shouldn't be reading this because I would just love to have a 20/1.8 to replace my 20/2.8 AF (not D).

Sounds like an incredible idea at this time of year

[Shane]

Shane, this review may offer some food for thoughts...

http://www.lenstip.com/457.7-Lens_review-Sigma_A_20_mm_f_1.4_DG_HSM_Coma__astigmatism_and_bokeh.html

Thanks Akira I had recently come across this report. One of the problems I have with lenstip is the number of samples tested and lens variability. I recently posed a couple of questions to lenstip concerning this point, as well as some others concerning their setup,  but received no response.

For example their coma test for the Sigma 35/1.4 was interesting (and useful) when I compared it to my recently acquired (and returned) Sigma 35/1.4. My copy exhibited the same amount of coma at f/2.8 as their copy did at f/1.4, and at f/1.4 mine was horrible. Initially it might be hard to determine if their copy was an exceptionally well corrected copy or run-of-the-mill example. Based on other glowing reports, when this lens was used for Milky Way images, I assumed that it was a run-of-the-mill copy and returned mine for exchange.

Lensrentals are in a better position to comment on variability but it would be nice if they posted aberration image examples of a "typical" lens.

[Akira]

Shane, I wouldn't be surprised if you had already known about this lenstip test. 

The problem of this kind of review is not only the sample variation, but that the reviews only show the optical character of the tested lens when focused at one particular distance.  But there's no doubt hat you had known about that.

I don't think any lens reviews except for those of the Lensrentals are performing the tests with multiple samples.  For example, photozone.de counts mostly on the volunteers to lend their samples for testing.

I myself have experienced multiple defect samples of the same brand (not Sigma) bought new, which eventually led to my doubt for their QC...

[Shane]

but that the reviews only show the optical character of the tested lens when focused at one particular distance

One of the questions I specifically asked them. 

[Akira]

One of the questions I specifically asked them. 

No wonder you asked this essential question.  As you should already know, this interview supports how the resolution power changes according to the ditance and also the amount of lpm:

http://nikkor.com/technology/02.html

[Mongo]

Mongo owns and has used the 20mm f3.5 Ai for many years quite happily. Not the sharpest lens but performs very well between f8 and f11 and set to hyperfocal distance. Mongo intends to keep usinging but if an upgrade is to happen in the near future, it was going to be the nikkor 20mm f1.8. Now, however, there are some serious third party contendors to consider before upgrading. Mongo has the sigma 35mm f1.4 Art and it is a superb lens dollar for dollar and feature for feature (but has no VR whic is of no consequence anyway) .

Maybe Mike's suggestion at the 14mm is worth thinking about and cropping when needed. However, Mongo has always found that 20 or 24mm have been his favourite landscapes lens together with the use of a 14mm as well i.e. 2 lenses

[zuglufttier]

Now that I've used the 20/1.8 a bit I don't see any reason not to recommend it. It's very sharp overall, not heavy and produces nice pictures. I paid 590 euros for a used lens that looks and feels like new. That is not exactly cheap but the Nikon delivers!

[Erik Lund]

14mm is very wide, so unless your up really close and use high MP camera, details will be very small, cropping 14mm to 20mm is not recommended IMHO

24mm is the best compromise for FF and 24 MP Then If wider is needed stitch together 2, 3 or 4 images in portrait mode.

[Fons Baerken]

Problem with wide lenses for landscape and architecture other subjects as well the distortion both in the corners and the horizontal compression, which can only be partly removed.
On horizontal compression the angle and plane of view are important, think of t/s lenses.

[Erik Lund]

With high resolution images form good glass and good sensors T/S lenses are 'almost' a part of the present for images that need corrections, it's amazing what you can get away with without visible traces in the final image.

[Sören Hese]

The 20mm f2.8 AIS is nice as a light companion, but its clearly not a sharp lens. On all digital full frame bodies it only gets acceptable in the corners at the edge to diffraction land (@f11 - where it quickly gets soft again). Even on the D3 it was a soft experience. At least all samples I had the pleasure to test behaved that way. I only grab it when I have to go compact and light and want a center object in the very near foreground to stand out (and I should upgrade to the 20mm f1.8 but I just love the build quality of these old AIS lenses and this sometime makes me forget the drawbacks of these old lens designs).
Usually I am with the 14-24 and its not a heavy beast when you are used to these f2.8 AFS tele lenses (and it includes some of the other classical fixed lenses). If you have a limited budget than go for the new 18-35 AFS version!
Regards! Sören