gdanmitchell
Offline
• • • • • • •
Upload & Sell: Off
|
 p.63 #18 · which lens has the most 3D POP? | |
Comparing those two images to draw conclusions about lens performance is a fool’s errand.
They are not lit the same, and they are not cropped the same.
Whatever other differences there are shrink to nothing compared to what we typically do with even minor adjustments in post.
For the umpteenth time, all of the micro-this and micro-that that people keep going on about is essentiallymake meaningless when comparing excellent lenses from different manufacturers…
… and this thing that people refer to as “pop” does not come from lenses. It comes from all o fthe other decisions that photographers make regarding composition, lighting, focal length, aperture choice, color relationships, depth of field, and more.
Learn about those things and how to use them and you can make photographs from any camera and lens “pop.”
philip_pj wrote:
3D (dimensionality) specifically in portraits
I’m looking here at this example comparison (by Dustin Abbott) of the Viltrox 35/1.2 and Sony’s GM 35/1.4. The methodology is sound (same lighting, aperture, expression, head tilt, etc) which is very important if we are to make valid conclusions. Some of the factors will certainly overlap with those that determine 3D at longer focal distances and with different arrangements of image objects (motifs). The focal distance is also similar to what users would employ, around 1.5-2m. It’s at 21:28 here on YT:
GM VS LAB: 35mm Showdown | Do We Have a New Champion?
The first point to make is that focal planes are very thin at regular focal distances at f1.4, even in 35mm lenses. Many results flow from this often-unstated aspect of portraiture. And in fact, it also confirms the need to closely consider the entire out of focus ‘bokeh field’ for any fast lens, because they are designed partially or primarily for this usage.
Put simply, in low light or to avoid clutter in backgrounds, you don’t have much choice other than to use wider apertures, and you quickly come to appreciate reliable and enjoyable bokeh, which can keep the focus on the subject while providing some context, if it is managed well. Bokeh is a spectrum characteristic, ranging from the fashionable full blur to more gentle contextual backgrounds (and foregrounds).
But this wafer of the focal plane is the only section of any image shot at wide apertures for which we can speak accurately of ‘sharpness’ and micro-contrast. Reviewers who gleefully point to aberrations like fringing, LoCA, SA etc. at wide open aperture, and wax lyrical about sharpness wide open tend to pay no attention to other critical elements that, taken together, influence how our portrait images look. Most of these affect the images’ out-of-focus regions which (very important) include focus transitions.
The ones I want to mention here all apply at the fine level; images are constructed from the ground up, something reviewers intuitively know so they habitually use crops and charts to illustrate their comments.
Micro-contrast
Most know what they need to know about micro-contrast (apart from the fact that all MTF is micro-contrast, as it appears in images). This is the attribute that instils sharpness but it is only valid for planar subjects such as test charts - ‘the focal plane is where the image is formed, and *its thickness is minimal, essentially a point or a thin surface’; ‘it's a small distance, measured in millimetres.’* If you are looking at a face such as in these images, the focal plane might be as little as 5% of what you see.
Micro-focus
I know you have never heard the term before, nor have I. By this word, I mean the measurement between two points separated by longitudinal space from the lens. It refers to the difference in focus between these points, as manufactured by the fade character of the lens: how fast the fade is, and whether the focus fade is linear or if it changes its rate (fading faster or slower with longitudinal (axial) distance). These points may both be inside out-of-focus regions or one may be on the focal plane. The macro attributes all apply to bokeh points as well: contrast, definition, color, etc. How lenses respond to small changes of focus will impact the image significantly.
Micro-brightness (tonal gradation, micro-gradation (h/t Dirty photography club)
All pixels have their level of brightness, they are the raw material for our familiar image histograms that graph pixels counts in 256 increments (values: 0-255). This term describes the ability of the lens to accurately depict differences in brightness at the pixel (or local) level. The value of this measure is that it may show discrimination between adjacent parts of a subject in our images – in this case small sections of a face. Note that tone depiction applies to pixels not only located on the focal plane, but also inside the transition zone or full bokeh (where full abstraction leads to amorphous clouds of blur).
Analysis
It may be clear to you that one image here has better image depth than the other. It’s immediately apparent from the global level, the left image’s face is just easier to look at and it conveys that sense that you ‘understand’ it. Macro contrast is stronger, but micro-contrast is close to equal. And a warmer color palette helps as well, but these are insufficient to explain all of what we see – the extra life, the better shaping of features like the chin, lips and cheeks. The Sony image has large regions of rather flat and plain bright tones that expand out from the in-focus parts - these interfere with identifying the small facial regions by smearing a flat and near-uniform color across most of the face, in focus or not.
In the left side image, we can see better definition of the skin blemishes around the sides of the cheeks, far from the focal plane. Shadow detail in the right side image is lighter and highlight detail is decreased by lower definition of features. Contrast in the hair is weaker and ear definition is less apparent. The chin in the Viltrox image is very well drawn, but the lower face is flatter and less clear in the Sony image.
Conclusion
3D in faces is accentuated by small motifs that signal depth to our visual system. I’d expect a smoother histogram with fewer spikes. So micro-color and micro-brightness are playing a big role here, helped along by the better retention of facial structures in out-of-focus areas in its focus transition. They are all stronger in the Viltrox such that the left side image shows a clear step up in terms of clarity and presence.
It’s more faithful to how this face would look in real life. It’s almost as though the Viltrox lens was prepared to make better use of the light falling on the subject. Does it matter? If you want to photograph faces that resemble what we see in real life, yes it does. That Viltrox was able to do so well is quite a surprise. I hope that this exercise demonstrates the value of 3D for portraits as well as other photographic applications.
...Show more →
|
