Mathieu18 wrote:
I’m not saying optics doe t play a role, but your comment about your wife, and certainly Derek’s weighty response echo my own thinking on it. Our brains take visual cues from the image, she. The fall off of the image closely matches what are brain expects, we see 3D in the image. When it doesn’t we see flat. I think that’s one reason that predominately we see the discussion around 35-50mm with some outliers. Also why brenzier shots or fast wide shots work, as out version is a narrow cone of focus with wide oof areas. The image is mimicking what our brain expects. It’s plausible that depth perception is somewhat of a sliding scale depending on our vision, so some see it and some folks don’t really register the depth.
Therefore while optics can perhaps ruin the 3D effect, I don’t think they can create it. Lenses that exhibit more of it, probably just match the most common vision, or the way the majority of us “see.”
Second this request from Mathieu... help us out, Sebboh.
My ten year old son and friends spent a couple of hours yesterday going through a card deck of optical illusions. There is likely something instructive for this discussion in deconstructing depth-related optical illusions. Understanding the cues of depth perception, could I intentionally use them in my photography to create images with more “pop”?
Similarly, it seems it might be more interesting to separate bokeh from “pop” and examine stopped down images that still exhibit “pop” to isolate other contributing factors.
I do feel certain lenses, like the Loxia 35, have a lot of depth stopped down vs. other lenses. I often was surprised at how much “dimensionality” the Loxia images had at f/3.2-f/4 compared to my other 35’s (including the RX1, FE 35/2.8, FE 35/1.4). Even my C/Y 35-70 sometimes exhibited this characteristic. Here’s a pic that felt somewhat “3D” to me despite being shot at least f/3.4 (I didn’t note the aperture setting).
Mathieu18 wrote:
Quite interested if you’d expand on this answer, specifically the 5 cues, or if you have some good literature to start with.
No, no, no, no. To bring this discussion to a definitive close, there is but one name: YANNICK KHONG.
I’m shocked his name and authoritative work on the subject hasn’t been brought into this discussion. Sneak preview... it is not at all complicated everyone, it is really very simple: there are only a “few elements” that determine if a lens renders flat images or with great depth perception. Read it and weep.
smpetty wrote:
Of all of the photos in this thread, even the interloping GFX images, your Sigma 20 shots have the most 3D-ness and pop, IMHO
Thank you, @smpetty. It honestly does do it the most regularly for me out of all my lenses (and I've also had the GFX with 63/45 along with faster adapted lenses). The lack of regular separation/differentiation of IQ from the then A7R2 and the GFX 50S for my uses was not readily evident or easily perceptible. The difference in usability however, was very noticeable which is why the 50S was ultimately shipped off to a new owner.
Under the right lighting conditions, the effect can be easily had with even APSC sensor bodies.
Also, it is easy for people to see what many consider to be pop or 3D with a fast telephoto lens (or close to the subject with a fast normal FL lens) since everything in the out of focus areas is obliterated. The look that gets me excited though, is when something is shot with a wide angle lens and has the pop. The ability to shoot an environment, be able to see plenty of the out of focus plane detail, and still have the pop as it were.
Found a few more Sigma 20 shots. Slightly less so than some of the others I posted earlier in the thread, but still there (at least to my eye).
Some samples with telephoto lenses. Again though, much less impressive (to me at least) when compared to a wide lens that gives its own brand of 3d pop.
bushwacker wrote:
This subject been discussed before... look from previous thread maybe 5 years back or more
There's one comparo between Zeiss 1.4/50P vs 2/50 Makro. One has blurry background and the other all sharp same scene.
There's this green Saab sample too.
and then Red Ferrari by Denoir
Those people who are in to this thing know what's 3D and what's not.
look for these names: Samuli, Denoir, Makten, Sebboh, Philip Reeve and many others.
The ability to discern depth (or the illusion of depth) is a basic component of human vision.
I'm not sure that it is necessary to have "what's 3D and what's not" explained.
On the other hand an expert explanation of how optics play a role in this phenomenon (assuming they do play a role) would be very enjoyable to read as would any insights into the physiology of the phenomenon.
I'm sure Sebboh can help us understand the latter component so if any of the above can school us on the former that would be wonderful.
No, no, no, no. To bring this discussion to a definitive close, there is but one name: YANNICK KHONG.
I’m shocked his name and authoritative work on the subject hasn’t been brought into this discussion. Sneak preview... it is not at all complicated everyone, it is really very simple: there are only a “few elements” that determine if a lens renders flat images or with great depth perception. Read it and weep.
darrellc wrote:
Second this request from Mathieu... help us out, Sebboh.
i'll see if i can find some nice reviews that aren't behind a paywall. the short answer is that nothing from neuroscience research is going be terribly helpful in choosing lenses which have relatively tiny differences between them, but has a lot to say about how composition and even post processing can give things more of a 3D feel.
i don't know much about optics research, but i expect most of the work focuses on minimizing aberrations and materials. if anybody has done serious studies examining how subtle differences between lenses impact 3D feel or just viewer preferences, i'd bet it was marketing firms or lens producers and that it would not be generally available.
No, no, no, no. To bring this discussion to a definitive close, there is but one name: YANNICK KHONG.
I’m shocked his name and authoritative work on the subject hasn’t been brought into this discussion. Sneak preview... it is not at all complicated everyone, it is really very simple: there are only a “few elements” that determine if a lens renders flat images or with great depth perception. Read it and weep.
his graphs make the scientist in me twitch uncontrollably.
I know that he and "the angry photographer" claim that 3D is a function of element count.
I have no reason to think that's wrong but I have never seen any research that backs it up.
What, in your opinion, makes it authoritative?
In any case he and Lloyd are preaching to the choir as far as I am concerned.
I love Zeiss optics (and similar lenses).
I'm quite sure he was being sarcastic. Yannick is a fool and dead wrong, even in his most basic assertions. He was put properly in his place by Brian Caldwell, designer of the majestic Coastal Optics 60/4.
smpetty wrote:
I agree that it's 3D-ish, but there is plenty of background blur, especially considering the depth of the scene. The rock right behind the bike parts is blurred.
The point is that it would probably look less "3D" if there was more background blur. I think most of the images in this thread shot wide open with fast tele lenses don't at all look "3D" because it's something that my eyes can't relate to. It would never look like that IRL. Edit: A totally blurred background also eliminates the context of the surroundings of the subject. Which, in many ways, isn't beneficial.
Dave Sanders wrote:
I'm quite sure he was being sarcastic. Yannick is a fool and dead wrong, even in his most basic assertions. He was put properly in his place by Brian Caldwell, designer of the majestic Coastal Optics 60/4.
Not surprised. It seemed to be a very simplistic assessment and I never heard anyone else of substance back his claim.
Well,
Since i switched from the Canon 5D and Contax Zeiss (C/Y and N) to Sony E System i was looking for a Compact (AF) Lens who give me this effect and could not find any for an affordable price.
I have the Batis 85mm and the Zony 55mm but they are just sharp and nice blur but no 3D POP, i would not wonder if they are just Zeiss marked.
The only Lens i have which is very nice in all ways is the converted Contax G 45.
And it pops all the time in 3D.
I think the Batis 40 has a bit of it, but it's not really small. Being a Batis, it's kind of fat....and the hood makes it fairly long as well.
Compared to the G 45 it's huge.
-Tim
rsrsrs wrote:
Well,
Since i switched from the Canon 5D and Contax Zeiss (C/Y and N) to Sony E System i was looking for a Compact (AF) Lens who give me this effect and could not find any for an affordable price.
I have the Batis 85mm and the Zony 55mm but they are just sharp and nice blur but no 3D POP, i would not wonder if they are just Zeiss marked.
The only Lens i have which is very nice in all ways is the converted Contax G 45.
And it pops all the time in 3D.
Appreciate it, and to be fair, I think the composition and post are probably more important than as you say the small differences between optics. I was more looking to shoot more 3D shots than pick "the best" lens. I know Chromostereopsis can play a role but imagine there are other cues as you mentioned that can help convey depth.
sebboh wrote:
i'll see if i can find some nice reviews that aren't behind a paywall. the short answer is that nothing from neuroscience research is going be terribly helpful in choosing lenses which have relatively tiny differences between them, but has a lot to say about how composition and even post processing can give things more of a 3D feel.
i don't know much about optics research, but i expect most of the work focuses on minimizing aberrations and materials. if anybody has done serious studies examining how subtle differences between lenses impact 3D feel or just viewer preferences, i'd bet it was marketing firms or lens producers and that it would not be generally available.
his graphs make the scientist in me twitch uncontrollably.
I find the word “pop” to be too vague and also rather vulgar. Are there any clearer, more precise terms that could help describe what exactly we’re discussing here?
I have a different take here. Like those Renaissance paintings that has the 3D-ness to them, there are 2 important elements that contribute to the perception of depth - shadow, and gradation (which strongly related between both).
We perceive a photo without dark regions or shadows as flat, no matter how sharp they are. And on pictures that does have shadows, but without smooth gradients, will often look flat too.
So the biggest factor here, as some already mentioned, is the composition (particularly lighting). Then, it's whether the camera will be able to capture the gradation of the transitions between dark and bright regions, to give it more refined look.
In other words, having the best camera in the world, shooting in a bad and flat lighting, nothing will look 3D. However, when the lighting is right, and the framing and angle is right, even mobile photo can take pretty 3D looking photos.
Below were taken with Huawei Mate 9 SOOC jpeg.
And this is with aperture mode with blur amount set to taste to avoid unnatural look.
Not to impress everyone with the above and I took the same scene with RX1R and they looked even better. Took these for quick sharing with my friend and surprised they turned out very nice too. Just sharing my opinions here and perhaps some of you might not find my above shots interesting or even look flat.
Mathieu18 wrote:
Quite interested if you’d expand on this answer, specifically the 5 cues, or if you have some good literature to start with.
here's a cute little cartoon of the major cues we use to extract 3D information from images (can people see this image or does it require a journal subscription?):
obviously disparity and motion don't really work for photographs. only blur and texture are really subject to lens differences. there are, of course, other less important cues that are not shown.
there is a lot of interesting neuroscience work that might be relevant trying to figure out how responses of individual neurons across different visual brain regions are bound together to create a coherent percept of an object. the short answer to that though is that we are a long way from figuring it out and a large part of it may come from top down processing...
his graphs make the scientist in me twitch uncontrollably.
