sebboh wrote:
the question is not which ones show good 3D, but do you think the one with greater dof has more or less 3D than the one with less dof.
The issue is "separation" between areas in focus and the rest of the image. Please, do not confuse DOF with the way the lens renders the transition between planes. IMO, three-dimensional rendering comes from transition between focus plane and other planes, i.e. if the transition is smooth or "abrupt" (sharp, steep, etc.). From a physical point of view, I guess the transition could be partly determined by how much quickly, or slowly, the enlargement of the circle of confusion in the adjacent planes in front of or behind the focus plane occurs.
A smooth transition is more similar to what our eyes perceive, thus contributing to 3D illusion.
Images with shallow DOF have a visual impact for sure. But this is not necessarily related to 3D illusion.
Ripolini wrote:
The issue is "separation" between areas in focus and the rest of the image. Please, do not confuse DOF with the way the lens renders the transition between planes. IMO, three-dimensional rendering comes from transition between focus plane and other planes, i.e. if the transition is smooth or "abrupt" (sharp, steep, etc.). From a physical point of view, I guess the transition could be partly determined by how much quickly, or slowly, the enlargement of the circle of confusion in the adjacent planes in front of or behind the focus plane occurs.
A smooth transition is more similar to what or eyes perceive, thus contributing to 3D illusion.
Images with shallow DOF have a visual impact for sure. But this is not necessarily related to 3D illusion....Show more →
yes it would be better to have a rapid transitioning lens and slow transitioning lens shot at the same aperture, but i don't have any image samples already available that show a rapid transition vs a slow transition at the same aperture and the transition steepness is mostly determined by how quickly, or slowly, the enlargement of the circle of confusion in the adjacent planes in front of or behind the focus plane occurs. in any case, for many of the images where people think they see 3D pop the transition zone isn't really visible and i just want to see what people are actually looking for since many people seem to be choosing completely orthogonal criteria for deciding if something has 3Dness.
sebboh wrote:
... for many of the images where people think they see 3D pop the transition zone isn't really visible ...
Yes. People believe that the shallower the DOF the higher the 3Dness. We are lucky Giotto and Piero della Francesca died long time ago and never read fora
I don't see any "3D" at all in the first two sets. The two last ones though! And I think the shallower DOF (or rather the larger background blur) looks more 3D in both those. It's not a large difference and mainly because of the very cluttered background.
Edit: However, the 3D-ness would probably decline if DOF was even shallower. Enough is enough.
I would agree that I've no "3D feeling" in the first two sets, in the last two definitely more and to my eyes I'd say more in shots A than B.
At least for me, a tad more blur helps in that regard, but for the effect to be convincing enough the background should remain pretty discernible (like in shots A), so that the eyes and brain are still somehow "tricked" in thinking that they are looking at something "real" and "3D" rather than an "artistic" mass of blur and a super sharp subject sticking out.
The first two pair, I'm just not getting much of the effect, tbh. With the third set, decent effect in both but neither significantly better than the other (I do find it curious that the subject contrast seems to be higher on the larger aperture shot). Last set, the smaller aperture shot (B) has the better effect, though it's pretty strong in both.
Agree on the surface but is there even an objective measure of this? Could we agree on any lenses that have a quick or slow transition?
sebboh wrote:
yes it would be better to have a rapid transitioning lens and slow transitioning lens shot at the same aperture, but i don't have any image samples already available that show a rapid transition vs a slow transition at the same aperture and the transition steepness is mostly determined by how quickly, or slowly, the enlargement of the circle of confusion in the adjacent planes in front of or behind the focus plane occurs. in any case, for many of the images where people think they see 3D pop the transition zone isn't really visible and i just want to see what people are actually looking for since many people seem to be choosing completely orthogonal criteria for deciding if something has 3Dness....Show more →
Mathieu18 wrote:
Agree on the surface but is there even an objective measure of this? Could we agree on any lenses that have a quick or slow transition?
could just do a ruler test, focus on a ruler at a fixed distance and compare different lenses of the same focal length to see differences in the transition. though maybe a textured surface would be better?
realVivek wrote:
I am not alone if you do not ignore the many others who posted here.
I have no clue what you're trying to say.
All I was pointing out a strict logical consequence of what you said.
Nothing more.
Also you just made a "eat shit, millions of flies cant be wrong" kind of argument there. Which yields nothing but a shrug from my side.
Facts dont care about belief.
Just because many people believe something thats actually false doesnt make that any less false, or more true. Just because only one guy believes a thing thats actually true, or maybe even nobody at all believes such a thing, doesnt make it any less true, or more false.
Too finely textured and micro contrast will come into play. Too course and it's meaningless. Ruler is only really good in one axis. presumably the CoC could vary in the saggital and tangential (sorry if my spelling is off). I'd think some sort of fine b&w diamond pattern or a series of b&w crosses on a grid would be ideal, but I don't have a pattern like that... Then you have to see if it falls off different close up (easier to test) or at portrait distance (wee bit harder)...
I think it'd be interesting but don't think I'm up for it!
sebboh wrote:
could just do a ruler test, focus on a ruler at a fixed distance and compare different lenses of the same focal length to see differences in the transition. though maybe a textured surface would be better?
Perhaps lenses that are easy to focus manually with focus peaking identify lenses with rapid transitions between in and out of focus... from personal experience, the ZM 50 Sonnar and the OM 50/2 Macro seemed to have very “peaky” focus peaking with a steep gradient before and after point of focus down around f/2.8 or faster. They were very easy to focus manually. I’ve other lenses that are challenging to manually focus in similar focal and aperture range.
Interesting quote from that article, as one could possibly apply the same strategy to photography by looking for these patterns in the natural or man made environment or perhaps enhancing in post: “Some painters, notably Cézanne, employ "warm" pigments (red, yellow and orange) to bring features towards the viewer, and "cool" ones (blue, violet, and blue-green) to indicate the part of a form that curves away from the picture plane.”
darrellc wrote:
Interesting quote from that article, as one could apply the same strategy to photography: “Some painters, notably Cézanne, employ "warm" pigments (red, yellow and orange) to bring features towards the viewer, and "cool" ones (blue, violet, and blue-green) to indicate the part of a form that curves away from the picture plane.”
Interesting. This is a technique that's long been used in landscape painting to simulate depth/atmosphere, but I hadn't heard of someone using it on smaller scale forms.
WhyFi wrote:
Interesting. This is a technique that's long been used in landscape painting to simulate depth/atmosphere, but I hadn't heard of someone using it on smaller scale forms.
Interesting quote from that article, as one could apply the same strategy to photography: “Some painters, notably Cézanne, employ "warm" pigments (red, yellow and orange) to bring features towards the viewer, and "cool" ones (blue, violet, and blue-green) to indicate the part of a form that curves away from the picture plane.”
it's worth noting that you brain will often ignore stereo cues when they conflict with monocular depth cues (particularly at greater subject distances).
Mathieu18 wrote:
Curious, do either of these look more or less 3D?
the last one looks the most 3D to me, though the differences are very small (flipping between tabs). i think it might just be due to the greater vignetting and distortion.
