p.10 #1 · Leica 50/1.4 Lux vs Voigtlander 50/1.2 Nokton vs 50/2 APO
4ALL wrote:
Depth of field is governed by physics, so it's extremely difficult to fully simulate the complete Summilux look, especially its unique rendering.
It would be hard but maybe not quite as hard if we had better spatial resolution in our phone images - i.e how well it can differentiate distances between objects on the Z axis, not sheer number of dots in the depth map.
And the real physical limitation here is that you can only get more spatial resolution by moving the phone’s lenses further apart and that is only possible up to a point and comes with its own challenges (e.g frame jumps when switching lenses, which they’d want to avoid for video).
So we’ll end up guesstimating the distances between objects a lot and that’s where the simulation will fall apart compared to the real lens and it’s infinite depth resolution.
p.10 #2 · Leica 50/1.4 Lux vs Voigtlander 50/1.2 Nokton vs 50/2 APO
'Depth of field is governed by physics'
May I disagree with this quote? Of course I can.
In imaging terms (it's why we take photographs, not as expressions of physics governance), DOF is a matter of perception rather than 'physics' or optical science per se. To a surprising extent, DOF and the pace of loss of focus can be accommodated significantly in lens design.
You can sometimes see steps in the transition from in-focus to out-of-focus content, such that a certain level of de-focus (or abstracted content) is a threshold beyond which the blur level suddenly increases only to more or less stay constant from that point backwards into the scene.
It's an undesirable factor in image depth perception and high levels of lens dimensionality. This phenomenon is very common and it breaks the unspoken visual contract the image has with your apprehension of it, your viewing of it.
It's a consequence of inattention in lens design, with an accelerant added by the insistence on the two plane theory of 3D - a super sharp subject (plane one) fronting a sea of highly abstracted content (plane two), better known as 'bokeh'. Few of them cared about it as they pursued ever greater MTF at high spatial frequencies, hence APO. Longitudinal image appearance was never a priority. Many, like Karbe, actually wanted a sudden drop into de-focus.
Many cine lenses (none of these are apochromatic, by design intent) are now appearing with smooth and extended apparent loss of focus. So who are you going to believe, the physics bozos or your lying eyes? There are also ways of increasing apparent DOF (a little) in post-processing, boosting micro-contrast for example and color separation ('micro-color'). cheers.
p.10 #3 · Leica 50/1.4 Lux vs Voigtlander 50/1.2 Nokton vs 50/2 APO
I'd put it this way: years ago, people used to say that a lens draws an image. Today, they tend to say it renders an image.
Decades ago, lenses had character—they drew an image. Today, many lenses are designed to render it as perfectly as possible.
I also remember when the first digital cameras came out. Everyone was amazed by their huge depth of field, which came from the combination of their small sensors and the very short focal lengths needed to achieve a normal field of view. At the time, many people saw that as an advantage, but looking back, it also changed the way photographs looked.
