Landscape lens odyssey Part III - Optical aberrations and phenomena affecting landscape image quality

The 3rd post in a series looking at how I arrived at a stable of lenses for my landscape photography as a Canon 1Ds3 owner and dedicated tripod user. 

The most succinct observation that I've read is that optical lens design is - 'the art of shifting aberrations somewhere where they are least seen' 

For an introduction on lens aberrations with illustrations and insight this link is worth a read http://www.lensrentals.com/blog/2010/10/the-seven-deadly-aberrations

I hadn't intended posting this as I'm not an optical designer and make no claim to have any profound insight into optical design and some of what follows may be well intentioned, but highly erroneous thinking. There may be optical aberrations and phenomena that I've discounted, that actually do impact on landscape imagery. Nonetheless, what follows is a highly subjective probably erroneous and abridged overview of lens aberrations, optical phenomena and their potential impact on images of landscape scenery that I prefer to photograph. 

Lens distortion -  I can't recollect an image being affected by low to moderate levels of pin cushion or barrel distortion, if you stitch panoramas you may feel differently.

Coma (Comatic aberration) - A concern for astrophotography, being honest I've read the terminology and am still no wiser on any effect for normal landscape scenarios at stopped down apertures.

Colour fringing (Axial and Lateral chromatic aberrations) -  What is perplexing is the nature of colour fringing isn't easily characterised by aperture, focal length, subject, it's a heterogenous phenomena in my experience and can show up anywhere in an image, but generally shows up towards the edges/corners of an image around high contrast edges. Snow patch edges, horizons, highlights and skylined foliage are prime candidate subjects for colour fringing, with the signature fringing of magenta, purple, green, red, blue and yellow.  Some colour fringing can be easily removed from some image subjects in post and at other times it is a tedious, time consuming if not impossible task to remove in other subjects. There is some speculation on the cause(s) of colour fringing on the internet, even sensors are suspected, but there are lenses that exhibit minimal colour fringing they are designated as being very expensive apochromatic or nearly apochromatic slightly less expensive.  I'll confess to an acute sensitivity to purple tree branches.

Flare, veiling flare, ghosting etc - not desirable and very challenging to correct, best avoided in my experience, unless the lens has the coatings up to the task, use of the lens hood and/or shading the font lens elements with a hat/hand is advisable.

Field curvature and astigmatism - manifests itself as unfocused and partly focused areas and elements in an image.  

Illustration of Field Curvature and why focusing in the corner, (the blue intersection point of lines) is a good idea with the lens stopped down (increased depth of field). Also a reason why MTF charts can show a drop in performance across the image.
Arribution : By BenFrantzDale (Own work) [CC BY-SA 3.0 (http://creativecommons.org/licenses/by-sa/3.0) or GFDL (http://www.gnu.org/copyleft/fdl.html)], via Wikimedia Commons

If you want to verify the effect of field curvature, it is best to give some thought to an actual landscape view that has subject matter which will highlight field curvature/astigmatism, visit it and with careful live view manual focusing across the image from the corners to centre, make a series of images, recording the focus distance  and repeat through an aperture range. You will have characterised your lens copy and established how field curvature afflicts that lens, what apertures provide enough depth of field to mask the effects of field curvature and the actual minimum focusing distance to achieve near/far sharp landscape images. Wide angle zoom lenses present an interesting challenge as field curvature and astigmatism will vary across the range of focal lengths, as there is a power law relationship with the angle of view ~ double the angle of view and field curvature/astigmatism increase by a factor of 4.

There's also evidence to suggest that some lenses have different field curvatures for different spatial frequencies, so fine detail should be carefully examined/pixel peeped.

Colour rendition - It could be an elephant in the room, emperors new clothes or a white elephant scenario. There are 'cool', 'warm' and 'neutral' colour lenses and this may potentially enhance or adversely impact a landscape image. I don't know if a lens colour rendition is relevant with digital sensors, colour filter arrays, camera digital white balance, colour saturation settings and RAW software, lens profiles ... but I could speculate that a lens colour contribution must have some impact on an image. There are ISO standards for the colour contribution index of a lens, limited manufacturers literature, some speculation that getting better colour involves investing in lenses that include the words APO, exotic glass, aspheric lens elements and marketing material which proclaims 'vivid' 'crisp colour' 'natural rendition'. There is a price to pay as normally the lenses are expensive.

Sharpness/resolution/acutance - I'll define it as the ability to resolve fine detail across the whole of a landscape image, at stopped down apertures and manually focused in live view to allow for field curvature. I can easily trade some centre sharpness for improved edge/corner performance.

Diffraction - with a 21mp sensor I once took a set of images between F5.6 and F22, pixel peeping unsharpened images and a loss of sharpness can be detected, in print with appropriate sharpening the difference was far less apparent. I've since read articles online by other photographers who have reached a similar conclusion on far bigger print sizes etc etc I believe; lens aberrations; insufficient depth of field; technique shortcomings and atmospherics, all have a bigger impact on image quality, imho.

Sunstars - The number of aperture blades for some landscape photographers is a critical aspect, I rarely incorporate the sun into compositions, so it isn't an issue. Aside from sunstars, I've no idea if  there's any benefits to 5,6,7,9,14... aperture blades.

Vignetting - I can live with moderate amounts at stopped down apertures, if you are stitching images this may be more important.

Focus shift - where the point of focus moves as the aperture changes. It may be astigmatism or field curvature, a combination of either or something else, I've no idea. Another reason to check focus in live view, having recomposed a scene and to test out in the field.

Rendering of out of focus areas - is something that I've never considered, as I'm normally trying to maximise depth of field to keep everything in focus. 

Prime or Zoom lenses - this is a subject deserving a blog post all of its own, as it defines to some extent a landscape photographers philosophy on image capture in the field. 

My own philosophy on lens quality is a pragmatic one, which concerns those lens optical aberrations and phenomena with the aperture stopped down, that adversely impact in my subjective eyes close scrutiny "print sniffing" of an A3 print of landscape scenery from a 21MP sensor. Some may consider this a low bar, I'd counter that there's a surprising number of lenses that fail to meet this remit.