Can't believe it's not butter... ARRRR
@Paddington basin, LDN
First of all, I'm on a bit of a saturated phase at the moment, bear with me.. Secondly, the APO Sonnar = Nikon 58G on some truly Bokehlicious crack!
\begin{rant}
I've always wondered about the hocus-pocus-y terms "pop" or "3-D" in a lens. I suppose these two are not quite independent as each other since there are a few mechanisms going on:
1. Micro-contrast
2. Ability to slice scene cleanly into "planes"
3. Field curvature and aberrations
I define micro-contrast as the contrast difference between two adjacent points of a certain small distance. So it is possible to have a high micro-contrast (i.e. image very sharp) and a low global contrast (i.e. uniform tonality). Similarly, point two replaces contrast with focus.
Having defined these two things, then I guess a lens designer would optimise these two properties with different profiles. E.g. a peak function profile (i.e. one that looks like _/\_) would be desirable I think, since you'd end up with an image with sharp transition between focus/oof and a sudden "contrast-vignetting" effect which would increase pop, perception of sharpness and a depth effect (3-d ness).
A Gaussian function profile (hill with round top) on the other hand, would produce a relatively smeared image (in terms of focus/oof transition and micro-contrast) but if the maximum height of the hill is larger through the addition of field curvature, then an even more pleasant image of supreme "pop" and "3d" can be found, sacrificing sharpness and introducing unwanted aberrations (these lenses are usually chromatic-aberrate-a-lots).
Whenceforth, I've a feeling that the Nikon 58G has a Gaussian profile with a large height aided by the strategic addition of severe field curvature whereas the APO Sonnar is a lens with a peak function profile for supreme sharpness and edgy focus/oof transition but with a smaller maximum height since there's practically no field curvature. As a result, it may be the case that a flat-field lens is easier to correct aberrations for (easier to line things/rays up maybe).
On the extremely expensive side, a Noctilux 50/0.95 (I've tried one in the shop but remain underwhelmed) is actually closer in rendering to the APO Sonnar but with some blur/aberration characteristics of the 58G, though not all.. It's why they charge £8k for it I suppose. But to be honest, I'd be surprised if there's a world of difference if you compare it with a good copy of the 58G on a 16MP sensor like the Df. (Btw, the 58/1.4 is comparable to a 50/1.2 dof-wise I think.)
So, in the end, you've got to balance how much do you value the 58G kind of aberration-ridden-unsharp-but-incredible-depth pop vs the APO Sonnar kind of perfect-and-sharply-delineated kind of pop.
tl;dr - 58G if you want to pop like Andy Warhol on stilts on ice, APO Sonnar if you want to pop like a precision-fuelled optical diglett with a license to erm.. overkill?! ..nvm
\end{rant}
Anyway, procrastination over, back to bubbles...
Can't believe it's not butter... ARRRR
@Paddington basin, LDN
First of all, I'm on a bit of a saturated phase at the moment, bear with me.. Secondly, the APO Sonnar = Nikon 58G on some truly Bokehlicious crack!
\begin{rant}
I've always wondered about the hocus-pocus-y terms "pop" or "3-D" in a lens. I suppose these two are not quite independent as each other since there are a few mechanisms going on:
1. Micro-contrast
2. Ability to slice scene cleanly into "planes"
3. Field curvature and aberrations
I define micro-contrast as the contrast difference between two adjacent points of a certain small distance. So it is possible to have a high micro-contrast (i.e. image very sharp) and a low global contrast (i.e. uniform tonality). Similarly, point two replaces contrast with focus.
Having defined these two things, then I guess a lens designer would optimise these two properties with different profiles. E.g. a peak function profile (i.e. one that looks like _/\_) would be desirable I think, since you'd end up with an image with sharp transition between focus/oof and a sudden "contrast-vignetting" effect which would increase pop, perception of sharpness and a depth effect (3-d ness).
A Gaussian function profile (hill with round top) on the other hand, would produce a relatively smeared image (in terms of focus/oof transition and micro-contrast) but if the maximum height of the hill is larger through the addition of field curvature, then an even more pleasant image of supreme "pop" and "3d" can be found, sacrificing sharpness and introducing unwanted aberrations (these lenses are usually chromatic-aberrate-a-lots).
Whenceforth, I've a feeling that the Nikon 58G has a Gaussian profile with a large height aided by the strategic addition of severe field curvature whereas the APO Sonnar is a lens with a peak function profile for supreme sharpness and edgy focus/oof transition but with a smaller maximum height since there's practically no field curvature. As a result, it may be the case that a flat-field lens is easier to correct aberrations for (easier to line things/rays up maybe).
On the extremely expensive side, a Noctilux 50/0.95 (I've tried one in the shop but remain underwhelmed) is actually closer in rendering to the APO Sonnar but with some blur/aberration characteristics of the 58G, though not all.. It's why they charge £8k for it I suppose. But to be honest, I'd be surprised if there's a world of difference if you compare it with a good copy of the 58G on a 16MP sensor like the Df. (Btw, the 58/1.4 is comparable to a 50/1.2 dof-wise I think.)
So, in the end, you've got to balance how much do you value the 58G kind of aberration-ridden-unsharp-but-incredible-depth pop vs the APO Sonnar kind of perfect-and-sharply-delineated kind of pop.
tl;dr - 58G if you want to pop like Andy Warhol on stilts on ice, APO Sonnar if you want to pop like a precision-fuelled optical diglett with a license to erm.. overkill?! ..nvm
\end{rant}
Anyway, procrastination over, back to bubbles...