New Fast Lenses, Perceptions and Relevance

Just out of curiosity (yeah sure), I have been looking at some of the comparison tests between the new f1.2 Pro Olympus lenses and my stable of, I guess now, "regular" lenses (although the 75mm sits in it's own middle ground for now).

The consensus seems to be;

  • The Pro (and Leica) lenses can test sharper, especially wide open and in the corners.
  • The Bokeh they produce, predictably as this was one of the design directives, is "better" or at least more. This is obviously in response to the current fashion of modern or luxury Bokeh.
  • They are built like the proverbial brick sh*t house and are full frame SLR size and price as a consequence.
  • The lesser lenses are all capable of matching the pro lenses in some circumstances, often the logical circumstances for their use and have shown some dignified character even when assailed by their no-holds-barred younger siblings.

My take away from this goes again to interpretations of need, minimum as opposed to maximum quality and perceptions.

Sharpness

When do we need really big files? The two relevant times are big prints (or prints of cropped images) and looking at small parts of an image on a screen, usually just for testing. There are no other relevant uses in most peoples lives for super high resolution gear.

Printing has a few realities to face. Very large prints are rare, expensive and of limited use outside of galleries. I know from years of experience, reading and observing, that camera resolution and lens sharpness suffer badly from the law of diminishing returns as the print process is the limiting factor. Choice of paper, ink type, printer and size are all going to determine the real end product, not just input. Remember that the world's greatest art works were painted with oils on canvas. Resolution was not a factor.

Printing has it's limits. Ming Thein for example spent a lot of time and money creating a series of super resolution prints that resolved the limits of a 36mp sensor. He was limited in shooting, printing and paper options, maximum print size and in his own words "relevance". You cannot do this at home or even through a pro lab. It took direct control of a multi $10,000 professional printing press (and the print maker) using multiple fine droplet passes and taking considerable time and money.

A good image will still be a good image even if technically weaker than one taken with later or better gear. Technical limitations are far less important that the aesthetic value if the image. The Holga crowd even revel in the flaws of their cameras.

If you need to go really big, small differences in sensor resolution and lens sharpness will not make that much real difference. Even if you did a pair of identical prints for comparison (only reason you would?). There are plenty of M43 vs full frame comparisons on line, generally showing that the difference in prints of quite large size (feet not inches) are surprisingly irrelevant. 

I am not saying that sharpness is not a thing, but visual perception of it is more often than not related to micro/macro contrast, focus, depth of field, colour, tonal separation and lighting than pure sharpness. I even remember seeing many images last century that looked sharper because of strong and tight grain.

If a lens shows low resolution, but high contrast, the perceived sharpness of the images it produces are usually better than the reverse. Much of this stands to be changed in post processing anyway. Sometimes too much fine resolution can even take away perceived sharpness from an image.

 An old (digitally ancient) image of my niece Lilly (now in her teens). The image is a crop taken with a 10D Canon (about 4mp after the crop). The perceived sharpness of her hair is strong and does not need to be any stronger, especially for this type of image. This image effortlessly prints to A3 with "hair splitting" sharpness. More resolution would define every pore or the peach fuzz on her skin but would it make the image better or just too busy? The Bokeh also shows how less DOF would ruin the image (focus is a little too far forward, grabbing the hair), but F2.8 on an 85mm lens saved it from the trash bin. I sometime wonder if this camera would still be more than enough for portrait images.

An old (digitally ancient) image of my niece Lilly (now in her teens). The image is a crop taken with a 10D Canon (about 4mp after the crop). The perceived sharpness of her hair is strong and does not need to be any stronger, especially for this type of image. This image effortlessly prints to A3 with "hair splitting" sharpness. More resolution would define every pore or the peach fuzz on her skin but would it make the image better or just too busy? The Bokeh also shows how less DOF would ruin the image (focus is a little too far forward, grabbing the hair), but F2.8 on an 85mm lens saved it from the trash bin. I sometime wonder if this camera would still be more than enough for portrait images.

As for 1,2,400% resolution comparisons on screen. They are only for photographers. No one else cares.

Testing sharpness only gets us into the bad habit of ignoring other lens characteristics such as contrast, tonal separation, colour, "draw" or such mundane things as handling.

Bokeh

This one is dangerous ground to enter. Firstly, there is no right or wrong, no perfect or un-useable Bokeh. It just is. When Mike Johnson and John Kennerdell first exposed the western world to the Japanese concept of Bokeh in the 90's, they opened a can of worms, that I think even they would sometimes like to re-can.

Bokeh of not simply fuzzy round balls of background blur. Some of the early lenses selected out as stellar Bokeh exponents (Leica 32 f2M) often only showed their true power at longer focussing distances, without exaggerated compression or super wide apertures. One of the things I took from the very first articles on the subject was the usefulness and relevance of Bokeh that created a settled and harmonious transition between the in and out of focus areas of the whole image. The modern trend is to ignore the transition and just go from super sharp to exaggerated blur.

 What would be gained from shallower DOF? An intercom as the hero of the story accompanied by lots of soft mush. If I wanted that, I could either get closer or use the 25mm with it's more trendy sharp/soft Bokeh  (and find something more interesting than an intercom to shoot).

What would be gained from shallower DOF? An intercom as the hero of the story accompanied by lots of soft mush. If I wanted that, I could either get closer or use the 25mm with it's more trendy sharp/soft Bokeh (and find something more interesting than an intercom to shoot).

Super shallow depth of field (as opposed the Bokeh as a whole) is very in fashion at the moment, but it can be over used and even get boring. My best or most useful Bokeh lens is my 17mm f1.8, which is considered by many to be "horrible" or at best "impractical" at Bokeh (as if it was a sport or applied trick), because it holds too much semi-cohesive detail, even at wide apertures. There was a time, not so long ago, when this was considered a very desirable characteristic.

 This image was shot wide open with the 75mm f1.8. If I had f1.2 would I have used it? No, because I needed to balance the amount of blurring with clear rendering of relevant detail. The Bokeh is gentle, sweet and harmonious, but not all pervading.  

This image was shot wide open with the 75mm f1.8. If I had f1.2 would I have used it? No, because I needed to balance the amount of blurring with clear rendering of relevant detail. The Bokeh is gentle, sweet and harmonious, but not all pervading.  

Certainly, my 25, 45, 75 and even the 75-300 are obviously more able to produce the accepted perception of Bokeh, but so are any longer, faster lenses. 

I personally want more connection to all parts of my image. Separating out the main subject form their back ground increases the perception of sharpness and "snap", but where is the story?          I would hate to think I only had two depth of field methods, super Bokeh and landscape.

How much blur is enough or too much? If you study the top tier images used by the masters of portraiture, they use many apertures, not just their widest. Go on, fight it. Fight the urge to go to that (literally) two dimensional safe haven of F-razor thin and explore the relationship between your main subject and some, less or even no blur in an image. 

I know if I bought an F1.2 lens, the novelty would wear off pretty quickly, but even worse, I might not like it's Bokeh or rendering for my style of shooting and would then start to resent the weight and price of the new glass.

The 17 and 45mm lenses at F2.8 are pretty sweet for me (that's F5.6 in full frame terms), handling the balance of in-out of focus elements beautifully and relevantly for their intended applications. Also at that aperture they both stack up so closely to the Pro lenses in sharpness that the differences are effectively irrelevant. The 25mm's look to be too close to split.

One of the things that drew me to M43 is it's near perfect practical application of apertures and depth of field. Full frame is worshipped as ground zero of photography, but why? It was invented out of convenience (availability of motion picture film stock) and considered a small and non professional format in it's day. There is not a real difference in quality until you start looking at medium format sensors and even then we come back to perceived want vs practical need.

Character

Totally subjective, but the character of a lens is often a deeply hidden and elusive beast.

The trend in most comparisons was a difference in colour from lens to lens. Sometimes the older lens was warmer, sometimes not. Apart from warmer images tending to look less contrasty (i.e. sharp), the processing of images from a lens becomes habitual and refined. I even have pre-sets for different lenses tending to make them all fit my preferences, without sacrificing their own look. It has taken me years of use to start to get a true feel for my lenses, but what it has taught me is that knowledge of the lens in hand really opens your eyes to what each offers.

When you look this deeply at a lens, it tends to blur the line between better and worse or stronger or weaker. You start to feel every lens, no matter how poor overall has something to offer*. If I had reviewed the 14-42 kit a few years ago, I would have probably been harsh based on it's build and price, but now I see the many character strengths it has. It and the 75-300 have helped remind me that Bokeh is not a more or less thing, it is a creative element with many degrees of application.

When the lens is far from poor, slightly better does not hold as much promise.

I have learned that the 17mm is the perfect lens for street grabs. It has the right mechanical and optical characteristics to do the job as I want it done. The Pro 17mm or the Panasonic 15mm may offer similar characteristics, but I doubt they would do that job better.

I learned a similar lesson with Canon gear. The nearly perfect, character filled 85mm f1.8 vs the slow focussing, way too big 85L with it's "one look" trick is a prime example.

How would I cope, not being able to rely on my knowledge of the delicate sweetness of the 45mm at F2.2-2.8, the utterly useful 17mm at F2-4 or the surprising quality and forgiving nature of the slow 75-300? Why break something that is working so well. The reality is, if I bought any of the new lenses (money is in the bank), I would most likely still use the old lenses for much of my work, effectively making the lenses twice as expensive if it is used half as much. 

This has already happened with the Pen F, kept in "reserve" while the old OMD EM5's do the bulk of my travel work because they are more expendable and more than enough.

I am not expecting this post to change any one's mind about the new glass. I know how hard it is to resist, but maybe my perspective may help sort out some thought processes.

*An old friend of mine and very talented black and white photographer used to buy up all sorts of lens odds and ends, test them and either keep them (often with their apertures super glued to their best setting), or pass them on with detailed health checks. Some of his favourites cost $10 at a flea market.