Geometric Distortion & Other
4th May 2019
To the Village
Sigma 100-400 @400mm, f6.3, ISO 1100, Nikon D750
Colour separation, colour clarity, peculiar sorts of contrasts, fanciful terms of other nuances - this is about the lens’ look, its rendition. Hoya, Ohara, Schott, Hikari. Bokeh and sharpness are worlds for themselves, for a while, as nothing is separate for long. The end belongs to the end and that is that - everything is overridden by a good photograph, by a long shot. A sigh of relief. And you can read on as I am not going to delve into all these terms.
So what the differences they are describing are stemming from? The devil is in the details, they say. Seriously, we have nice things to worry about with the choice of gear getting as wide and as good as today.
The details are a diverse landscape where generalisations are the exception rather than the rule. There is a basis for the differences in the look of lenses that we can notice, feel and sometimes imagine by sheer concentration on them. I heard and read people saying ‘Nikon colour’ when describing the look of Nikon lenses for instance. The use of ‘colour’ sums up well the essence, or at the least, you get the drift. It is open to debate to what extent concrete attributes of that look are results of intentional endeavour, or a chance creation occurred during the design and manufacturing. The trend is getting increasingly towards intentional aiming, and we know that the concentrated efforts tend to grow with the price tag. Nikon is traditionally focused in this direction and is building on this tradition so you can and do expect good results in this respect. Paying attention to rendering characteristics of the lenses is hardly uncommon though, and the existing differences in manufacturer’s approach and the results are not fixed in any way. This narrows the space for the possibility of potentially inherent brand advantage. It’s also a truth that the look is never entirely under the manufacturer’s control because the physics cooperate with the intent only to a certain point adding limits to what can be done thus always complicating the ideal design. This makes the results more exciting. They are also more unique for users. But there's a set of factors established in a more fixed way and they relate first to each other and then as a complex to the particular look.
The first attribute or more precisely a potential for it to notice is that each source optical glass can feel somewhat different. They are the secret names in the second sentence of this writing.
The second factor is that the lenses made of glass that came from the same source can also and very often will, feel different. This difference can be slight and it can be prominent. This makes sense when we realise that they, apart from unique optical construction, comprise a greater variety of glass sorts and coatings combinations within that source. These two sources of variations are the basis for distinctions in question (if you ever paid attention to them). They can be noticed in similar devices like binoculars.
The third factor is that the glass and the lens are made to interact with light. They are meant to render the seen. And the light seems at least as complex a factor in practical photography as the particular glass and lens. This makes this three-part interaction produce specific results. It could also explain some of the hard-to-put-finger-on perceptions in the lens look that we can feel or observe. This factor forces lenses to behave (interact in a complex way) rather than ‘be’ and represent a static, easy to describe rendition.
The fourth thing coming into it is that the output of a lens isn’t left on its own in the vast majority of cases. After all, the photographic lens is the delivery part of the recording medium. Leaving the processing aside (white balance adjustment), the lens colour/temperature output becomes part of data that is worked by the underlying correction called white balancing based on individually consistent though varied, because particular, white balance algorithm. Use of a sampling eyedropper or a grey card shot neutralising is the simplest form of this. It is about real colour this time, and it’s adding another variable to the interaction between a particular light, its intensity, type and the properties of the glass/lens of which result we can perceive in a taken picture. This one is different in that it lessens the overall difference despite introducing another variable. Mentioning the white balance, and purely as a matter of interest, we keep a database of WB settings obtained by our method shot individually with each lens for various lighting. This is a part of our WB handling and specification that we help ourselves with when we need to or after we ran into difficulties in the field. It gives us a wide control that way and results that are lens-specific (as opposed to mixed up ones) – and yes, you can tell the difference.
The resulting differences in rendition range from noticeable in some way, subtle, very subtle to barely detectable with the average being somewhere between subtle and very subtle. We can notice them on the processing routines as the file/data respond slightly differently. This applies to the colour as well though the differences show up best on set white balance comparisons (various lenses).
Not surprisingly, the individual perceptions our observations are based on aren’t exempt from variations. Like with more things, perceiving a difference gets amplified when one is particularly adapted or used to a specific medium, a glass and lens in this case, and so it is easier to develop a particularly inclined potentially distorted view. This is long known and allowed for in many fields. There are more objective and measurable differences in the sensory-related responses among us at play as well.
We have used many more lenses in both formats than we currently use. Some of them were indistinct whereas rendering of some caught our eye. There are many relevant aspects. Positives cropping out of the other manufacturers’ production will make us use their advantage. The offer is generous these days for the F-mount. Generally, we have a weak spot for Nikon lenses look. There is more to this obviously. The design and specifications are the soul of the photography’s rather special discipline and the Nikon’s style was always appealing. There were, and especially today are, other excellent systems of all formats, lenses or cameras.
From the camera standpoint, this is supported by Nikon’s way of file handling, including the proprietary RAW conversion. This enters matters of the profiles and thus first steps of standardised processing. High ISO and ISO’s increase in general subdue the differences that might be born from the mentioned above, for example.
And of course, everything mentioned above is overridden by that photograph.
The magical zoom
6th April 2019
Is standard zoom range lens today. 24 and 70 mm wasn’t always the standard. There are a number of variants alongside them that reach to 75/85/105/120 mm or start at 28 mm, and even 35 as the range extends to 150 mm in the Tamron’s portrait zoom. The speed also got more varied, and the modern standard f2.8 is more frequently accompanied by f4, faster and slower variable aperture and even f2 versions.
The zoom is called normal, which may well be valid for the range, but there is one thing that makes it different from the other lenses in this range. It is the zooming function. It changed what was used to be normal before dramatically. That is often downplayed by the people who tend not to use it while those who need it all the time will likely overstate it.
The truth though still is that the 24-70, and any of its variants optically at that level, remains a special lens. It’s because it enables changing focal length, a great many of them, without the need for other lenses and without the need to change them. Granted, if one has a rucksack full of fixed lenses, and the time or the option to change them, they’ve got a workaround, if they accept some compromises like not being able to do this way of zooming quick enough or while shooting. The real trouble with that setup though is that that rucksack and the possibility of changing the lenses are very often ruled out.
Changing the distance to get the required framing isn’t often an option either. But there is a different snag; you are not zooming this way. Instead, you are coming closer or further away from the subject, and that changes the perspective in your photo. In other words, it changes everything, including the elements that you can mind to be changing. This turns us back to the rucksack and those non-options because changing the FL in any other way can only be done by zooming. That way you crop the view to the desired framing and maintain the perspective. This is very different, and it can lead to very different results. You can also do it fast, almost not moving and while shooting, but mainly, you can do it. That is the magic of zoom.
But that is not all the magic there is, right? Some of it must be left we feel. Yes, although using zoom vs prime gives added possibilities there is a wider question to it in photography. It’s because this can ultimately be a big plus for one to get the shot in theory and practice, but it can also be a minus in practice.
Having more possibilities takes more concentration. As it is the result that counts, we need to ask whether the distribution of our attention is optimal for getting the optimal result and even decide what the optimal result for the given moment and scenario is. There can be less than a second for this. So we combine the judgement with our intuition. This is very dynamic, and it means that we can feel rather than think in many instances. Naturally, the photographer as an individual is central to this. We have our own responses and reactions to what we see or what is unfolding in front of us, a different eye, intentions and different opinions about the results themselves.
It’s true that in practice, a restricted choice can help us. And that our concentration will prove more useful elsewhere, often to our surprise. It can also create our distinct style. On the other hand, it can and will quickly destroy the shot or make it impossible. There are obvious answers to this and less obvious as well. It’s the latter instance that rules out universal answers. It simply will always be up to you - one of the main ingredients of photography.
6th April 2019
Uneven illumination is another kind of distortion though it concerns light.
And its correction can also be questionable in many cases. About 95% of my subject shots (an instance where an individual subject plays a major role in the picture) shot wide open are left with full shading in effect (that includes f8 of our 200-500 VR/1.4 TC combination).
The vignetting characteristic of each lens is important and a factor I consider when looking at a lens. So far I don’t remember us adding the shading in processing; it’s always the particular lens with the full effect of the aperture or corrected to a degree or applying the correcting profile (mostly in non-subject shots that include a broad scope of the scene in sharp rendition).
2 June 2017
In September 2014 Samyang released stereographic projection UWA for 135 format that is part of fisheye lens family. The stereographic diagonal fisheye maintains principle of fisheye projection where the linear geometry is bent but unlike the classic fisheyes, it leaves objects shapes in original form. The regular projection fisheye lenses deform shape of objects at the periphery beyond recognition. This is especially apparent in the corners where the fisheye effect naturally draws attention of a viewer (due to the compression). This doesn’t have to be a negative effect but it distinctly separates rendition of a stereographic UWA from typical equisolid angle and equidistant fisheye lenses.
As there are no elements of extreme rendition with stereographic projection, it creates its own category among the UWA lenses.
Another projection useful and intuitive for object recognition (in the entire image) is perspective projection of corrected UWA. The shapes of objects are deformed but easily recognisable and intuitively correct (though exaggerated) while the linear geometry in the image does not change.
It is debatable which projection produces more natural representation of the scene. The stereographic projection preserves shapes of objects locally in remarkably truthful representation and virtually unchanged but it curves the whole space. It exaggerates “its sense” in often intuitively correct way that is in line with our natural perception.
This makes both of these projection types closely related and fit for realistic representation and distinct from non-stereographic projection fisheye.
The fact, however remarkable, has obviously no a priori bearings on artistic, photographic or aesthetic choice and preferences as apparent from the subjective nature of most applications.
26 November 2014
Most rectilinear lenses add some kind of geometric distortion to the picture as they project capture of a subject. One of the most prominent distortions is changed perspective and it is particularly noticeable with the wide and ultra wide angles. It is a kind of distortion that influences remarkably well transition from our own perception of the scene based on our brain and two lenses cooperation to a one lens capture. Prominent changes in perspective and its exaggeration are also extending the tools for a richer expression.
Another kind of geometric distortion of a lens is barrel distortion. Unlike the regular and predictable changes in perspective, this kind of radial distortions are often seen as an optical flaw of a lens and that because it is much less necessary for the design and thus more changeable from lens to lens.
Because the barrel distortion is also a perspective distortion of a sort we can see these two interacting. We will notice that the barrel distortion works its own way towards offsetting the perspective distortion effects by compressing the near corner geometry while making the central area visually more apparent by enlarging it. A mild degree of the barrel distortion in particular can be then often viewed as a beneficial and desirable effect for the overall projection of the scene. It is fine that it is retaining its reputation of a flaw as it is a matter of degree, and it is also less useful universally as a tool. I myself tune my corrections for barrel distortion for each image separately and with some care. Correction that I apply is often partial and many times I am leaving it without any correction whereas sometimes I will choose to correct it fully.
Geometric distortion working in opposite direction of the barrel is pincushion distortion. This curves of the space in the image I correct often fully and in some instances almost always, for example, when typical wildlife is the subject. For other subjects, this I again tend to weight individually.
Moustache types of distortions that are less symmetrical work towards the elongated corner geometry and similar (compression) less effectively but even they can be found beneficial for rendering characteristics of a lens by a photographer. Their irregular curves though often command a default correction.
Observing meaningful details when processing a photograph can make a difference. And just as, e.g. careful framing where the each millimeter plays its role, the symmetrical distortions can help in creating a photograph that conveys its content precisely the way we want. Realising this can also change our view of excessive distortions that inhere in the design of affected lenses because the partial corrections using suitable software have only minimal negative impact on the rendering and resolution quality. And it can, on the other hand, add to the means for a thoughtful processing of digital images using mild distortions optimally for the intent of the final execution.