3D Pop - experiment

Started 2 months ago | Discussions
OP backsidewalkaround Regular Member • Posts: 255
Re: One word: Chromostereopsis
1

DMillier wrote:

I always find that when in the middle of a debate, the appropriate image is never to hand to illustrate the point

Fortunately, I have found some:

http://www.stevehuffphoto.com/wp-content/uploads/2014/05/L1000262.jpg

Perfect example of what I'm looking for. This one gives me that perfect depth impression.

http://www.stevehuffphoto.com/wp-content/uploads/2014/05/L1000267-680x454.jpg

I think this one might work too, but its small size kills it mostly.

In my opinion, the important factor that makes the subject separate from the background with a sense of depth is that while the background is out of focus, it is still recognisable.

If the DOF is so super shallow that the background is a meaningless blur, you get separation but inot that pseudo 3D of the Steve Huff image.

eg

https://www.colesclassroom.com/wp-content/uploads/2018/11/pexels-photo-713312.jpeg

This example isn't even what I call pop, when the subject jumps at you from out of the strong blur. Its a relatively flat portrait of a beautiful girl without any distracting background. For it to be popping it needs more contrast. I would guess the original was better in that regard, but the processing has made the image lifeless and flat.

-- hide signature --
 backsidewalkaround's gear list:backsidewalkaround's gear list
Sony a7R III Sony Vario-Tessar T* FE 16-35mm F4 ZA OSS Sigma 35mm F1.4 DG HSM Art Sony FE 50mm F1.4 ZA Sony FE 200-600 F5.6-6.3 +1 more
paulhoppe_photography
paulhoppe_photography Regular Member • Posts: 107
Re: 3D Pop - experiment

They are all pretty much flat 2D images. As flat as my display goes anyway

No seriously I can not discern very much difference in this regard at least not when looking at the smaller images together in your post. I think this so called 3D pop has much, much more to do with composition and lighting than any lens characteristic.

A three dimensional look in photos is an optical illusion because...well...the image is flat right? And a lens projects the incoming light onto a flat plane be it film or a sensor. Of course things like bokeh, vignetting, sharpness or contrast can aid or detract in this matter. But I doubt that more expensive lenses equal more "3D pop" or that some lens have this pop and others don't.

In fact I have an image that to my eyes looks pretty 3d poppy or layered or however you want to call it and it was shot with a decidedly bad and cheap 35mm f3.5 "Weltblick" lens (which was a brand from a German mail order catalogue). The lens' flaws like softness, crappy bokeh and distortion actually aid in making this more 3D looking. But then this is really just very subjective and you might not see any 3D here.

Nevertheless you did quite a lot of work testing these lenses.

-- hide signature --
 paulhoppe_photography's gear list:paulhoppe_photography's gear list
Fujifilm X70 Fujifilm X-T3 +1 more
OP backsidewalkaround Regular Member • Posts: 255
Re: 3D Pop - experiment

paulhoppe_photography wrote:

They are all pretty much flat 2D images. As flat as my display goes anyway

No seriously I can not discern very much difference in this regard at least not when looking at the smaller images together in your post.

I did upload them in full resolution, so you could look at them at a large size if you have a fitting monitor.

I think this so called 3D pop has much, much more to do with composition and lighting than any lens characteristic.

It certainly is the foundation of it all and you're probably right, without the right composition and lighting it's much less likely to produce an image that gives me the 3D impression. 
However, the lenses contribute too. I'm certain. I've experimented for a while now with several lenses and I do find that some have a much higher hit rate than others. 
One problem with this test probably is that it's done with many already chosen ones, so the differences are indeed small. Yet for me there are some. Non of this matters, though, if you don't see any of it.

A three dimensional look in photos is an optical illusion because...well...the image is flat right? And a lens projects the incoming light onto a flat plane be it film or a sensor. Of course things like bokeh, vignetting, sharpness or contrast can aid or detract in this matter. But I doubt that more expensive lenses equal more "3D pop" or that some lens have this pop and others don't.

Yes. I've given answers to that in previous posts in this thread.

In fact I have an image that to my eyes looks pretty 3d poppy or layered or however you want to call it and it was shot with a decidedly bad and cheap 35mm f3.5 "Weltblick" lens (which was a brand from a German mail order catalogue).

I agree, it looks "layered" - this is a fitting description. The effect is less strong if you can't follow the focus getting less, but it's already there, if your subject is clearly sharper than its surroundings in the back (quite a bit further, thus the transition is abrupt, which is also a clue. 
I find it best, if you can follow the focus getting less on the ground, but on top of the subject there is a clear "break" in sharpness, because the back is much farther away.
Also a flat subject normally isn't ideal, but for the "Layering" it works.

The lens' flaws like softness, crappy bokeh

I don't think those help, but it works despite of it, because it's very contrasty, which is a big help too.

and distortion

maybe, yes, I suspect barrel distortion to help in some cases, but haven't exlored that.

actually aid in making this more 3D looking. But then this is really just very subjective and you might not see any 3D here.

Nevertheless you did quite a lot of work testing these lenses.

Thanks for noticing ;-).

-- hide signature --
 backsidewalkaround's gear list:backsidewalkaround's gear list
Sony a7R III Sony Vario-Tessar T* FE 16-35mm F4 ZA OSS Sigma 35mm F1.4 DG HSM Art Sony FE 50mm F1.4 ZA Sony FE 200-600 F5.6-6.3 +1 more
Sjak
Sjak Senior Member • Posts: 6,356
Re: 3D Pop - experiment
1

backsidewalkaround wrote:

If you view the images in full screen mode (preferably on a large screen), do you get a different sensation of depth with some of the lenses?

Just took a detailed look, but to me, the lens-rendering doesn't contribute to the sense of depth in these images. It does alter things like contrast, but nothing you can't equalize in post.

NB - Not saying that lens rendering is irrelevant, I also use some lenses very much for their rendering. But for the illusion of depth, I look into other tools.

 Sjak's gear list:Sjak's gear list
Pentax K100D Leica M-Monochrom Pentax K10D Pentax K-1
OP backsidewalkaround Regular Member • Posts: 255
Re: 3D Pop - experiment

Sjak wrote:

backsidewalkaround wrote:

If you view the images in full screen mode (preferably on a large screen), do you get a different sensation of depth with some of the lenses?

Just took a detailed look, but to me, the lens-rendering doesn't contribute to the sense of depth in these images. It does alter things like contrast, but nothing you can't equalize in post.

NB - Not saying that lens rendering is irrelevant, I also use some lenses very much for their rendering. But for the illusion of depth, I look into other tools.

Thanks for having a detailled look. I agree: In post you can alter the images to some degree, some of it on top also boosting the effect, like selective sharpening.
However it doesn't take much to overdo it and the effect is gone, because you somehow recognize it as unnatural. An image that looks very much 3D (you know what I intend to say) to start with will always be better.

-- hide signature --
 backsidewalkaround's gear list:backsidewalkaround's gear list
Sony a7R III Sony Vario-Tessar T* FE 16-35mm F4 ZA OSS Sigma 35mm F1.4 DG HSM Art Sony FE 50mm F1.4 ZA Sony FE 200-600 F5.6-6.3 +1 more
Tons o Glass 0 Class
Tons o Glass 0 Class Contributing Member • Posts: 838
On guessing the lenses...
1

At least in this DPR post, images 7 and 8 are identical. The links lead to unique files, but downloading them and comparing them yields zero differences. This begs the question, what lens is missing?

To me, all of the images look about the same when it comes to 3D-ness - I think they all have a mild amount of "it" that is largely in credit due to the scene and how well you crafted/controlled it. The plane of focus choice to give us the most depth cues is very consistent and well executed.

That is not to say that the lenses don't render differently. There are some obvious differences when it comes to global contrast, SA, shapes bokeh balls take towards the corners, color temperature, etc, but these differences don't make or break the 3D effect for me in this scene. They are also not sufficient enough for me to strongly identify lenses, almost all of which I'm unfamiliar with, but I'll try anyway! I'm prepared to look characteristically ridiculous.

I couldn't really observe any indication that any lenses in the shootout have been stopped down (no aperture shapes in bokeh anywhere). With some f/1.2s in the mix, I found that surprising. Perhaps where the aperture would show up in the bokeh the most is obscured by the skater or the lighting is sufficiently gentle there to hide it, and those f/1.4 and f/1.2 lenses just vignette a ton? I don't know. Well played.

Image 1 looks very Minolta-ish color-wise to me, but with none in the lineup, and with Minolta and Leica collaborating on some things apparently, I'll pin it as the Leica R.

Image 2 has to the lone f/2.8 judging by bokeh ball size. Elmar.

When viewing the images in order, Image 3 impressed me. it reminds me of macro lenses, so maybe it's the Makro. If the goal of the shootout was to find a keeper, whatever lens this is might be it for me.

Image 4 is looking like a modern optic overall (very contrasty) which has me thinking it would be the Loxia. It looks like there may have been a light falloff correction applied but that may also have to do with the changing light of the scene. It's the only image where I noticed what might be a hint of a hexagonal aperture shape, however, so maybe it's the 50/1.7 Planar. Undecided.

Image 5 looks Canon-esque to me, and when viewing them in order, it's the first one that looks like it could have been shot with a slightly longer focal length, so I'll guess this is the Canon 55.

Image 6 could be the Pentax. I've seen corner bokeh that looks like that with it.

Image 7/8, I'll go with Nikon for the heck of it.

Image 9 has some especially wonky corner bokeh (similar to what I think is the Pentax) so maybe it's the other rangefinder lens in the lineup, the Hexanon. But mid-frame the bokeh ball size has me thinking it could have been a faster optic... so would it be Pentax then? Dang it.

 Tons o Glass 0 Class's gear list:Tons o Glass 0 Class's gear list
Sony Alpha NEX-3N A3000 Sony a7 II YI M1 Canon EOS 550D +2 more
FingerPainter Forum Pro • Posts: 10,404
Clearly not
1

D Cox wrote:

That seems to be several copies of the same photo,

That's clearly not the case. Examine the position of the subject's right hand in relation to his right knee. Different in every shot, except that #7 is a duplicate of #8.

Another thing that is different in every shot is the lighting. Nothing was done to control the lighting, which AFAICT was unmodified natural, and changed over a period of nearly 10 minutes. The length of the shadows on the ground seems to imply the photos were taken at a time when light would be changing  somewhat rapidly, and the dappled sun on the pavement show differing amounts of cloud shading between shots.

so how can the 3D effect vary ?

A change in both pose and lighting could make more of a change in perceived 3D pop than a change of lens. For that reason, I don't find this test series particularly helpful. A proper test would have constant (hence artificial) lighting and the very same scene (no change in pose so no live subject).

Also the background is too dark for any 3D POP effect.

Why would background colour matter?

More details of your setup are required, I think.

Don Cox

ProfHankD
ProfHankD Veteran Member • Posts: 8,172
Re: One word: Chromostereopsis
5

backsidewalkaround wrote:

ProfHankD wrote:

Experiment all you want, but "3D Pop" is fundamentally NOT a lens property. It is a simple set of composition techniques long widely understood by professionals. Best I can tell, the only thing Zeiss lenses have to do with it is that they combine slightly stronger than average red rendering with costing enough that they tended to be used more often by professionals... who would know the composition tricks.

Thanks for the lecture Prof ;-).

Sorry. I didn't mean to lecture, but this is a VERY frustrating topic for me....

I admire, that you've tried to explore the topic scientifically. I already knew about the influence of color. Still I beg to differ.

As I understand it, you didn't succeed to prove any other clues than Chromostereopsis. But you didn't prove the contrary either.

I haven't proved anything, but I have very strong empirical evidence from about a half dozen surveys (probably 100+ people) that no lens property is consistently seen as producing 3D Pop. In fact, the closest thing to a lens characteristic that caused 3D Pop was lens brand: people told an image was shot using a Zeiss lens tend to see much more 3D Pop than if they don't know the lens brand used. 

3D Pop isn't a thing unless there is general agreement about examples of it -- which there pretty clearly isn't, except for cases traced to chromostereopsis.

For example, some people will say any image with a sharp subject and heavy, smooth, defocus of the background is 3D Pop, which is what a shot like http://www.stevehuffphoto.com/wp-content/uploads/2014/05/L1000262.jpg shows. However, many others (myself included) see that not as a 3D scene, but as rather artificial-looking stacking of two flat layers. If you go back to my 2014 first attempt to define 3D Pop as a set of lens characteristics , my proposed definition was:

My suggestion is that it is the combination of:

  1. Very high sharpness (microcontrast) and texture for the primary subject.
  2. Slight, but very smooth, defocus of more distant parts of the scene around the primary subject.
  3. Some areas of moderately strong colors and tonal contrast, but not the whole frame.

and I really wanted it to be something like that because those are attributes I can recognize, measure, and enhance in postprocessing... hence, it would potentially have made a nice bit of publishable image processing work. But nope.

So there may still be other clues that are too difficult to grasp or to describe.

It seems that I need to elaborate on what I want. ...
I believe it is very difficult to talk about it, because different people have different perception.

That means "it" isn't an it, but a bunch of conflicting opinions. Chromostereopsis, on the other hand, has been generally agreed upon for a very long time. For example, stained glass (one of the first media to be able to make strong blues) was using deliberate chromostereopsis centuries before cameras were invented.

I'd like to put it this way: Every person (capable of seeing 3D) has a set of eyes and a brain to process the images given by the eyes. Differences start with the eyes, since some people can't discern red from green as an example, but there may be many other differences that are never measured or compared, because the eyes work. No reason to talk about individual vision in depth.

Not sure what you're getting at here. BTW, cover one eye and look at the Red/Blue image I posted above -- most people will still see blue as behind red despite having no stereo vision cues at all.

In sum, each of us is allowed to like whatever image properties we like, but there is no such thing as a lens producing 3D Pop. The photographer deserves more credit for their composition skills. 

 ProfHankD's gear list:ProfHankD's gear list
Olympus TG-860 Canon PowerShot SX530 Sony a7R II Sony a6500 Canon EOS 5D Mark IV +30 more
Travis Butler
Travis Butler Senior Member • Posts: 1,627
Re: One word: Chromostereopsis

I would pick out #7 and #9 as the ones showing the most 3D effect to me.

backsidewalkaround wrote:

It seems that I need to elaborate on what I want.

From my experience there are three versions of 3D rendering that need to be kept apart:
1. Verisimiltude: An overall very sharp contrasty image depicts reality as well as possible.

2. Pop: The subject pops out from strong blur. The surroundings are mostly unrecognizable or at least give not depth cues.

3. 3D rendering of a scene: For me those images give me the impression that I'm looking at a 3D scene (that is only 2D of course), but my brain makes me believe it.

I'll add in something from prior discussions on this topic here: Volume.

Separation from the background does help with a 3D effect, but is not sufficient; as someone said elsewhere in the thread, a flat cutout separated from a blurred background isn't 3D to me.

The thing that catches my eye is the suggestions that the object we're looking at has depth in itself - a.k.a. volume. Take a look at the right arm in #7, just to the right of the elbow - you can see a slightly darker stripe, suggesting a flexing muscle, something not present in most of the other versions. Likewise, the play of light across the subject's face, emphasizing the jaw line.

So for me, good tonal rendering - the ability to express a very finely gradated set of tonal variations, giving a sense of volume through the change in light as it plays across a subject- is an important part of the 3D effect.

For an example of my own:

Lumix S5, MD 24-50/4

Do you get a feel of depth from the clouds?

-- hide signature --

Flickr at https://www.flickr.com/photos/the_prof67/ Warning: Heavy Learning in progress.

 Travis Butler's gear list:Travis Butler's gear list
Nikon Coolpix A Fujifilm X20 Panasonic Lumix DMC-GM5 Pentax K-50 Olympus PEN-F +19 more
DMillier Forum Pro • Posts: 23,021
Re: One word: Chromostereopsis

Hi Prof

I am someone who (enjoys?) a colour based 3D illusion on occasion, where objects of different strong colours appear to "float" distinctly over the other as if the foreground were an inch or two above the background.

I've asked about this effect on the forums before because I've tested the effect on family members who were unable to see what I saw.  I have asked an optician about this and got a blank look in response.

For me this effect can be very strong, shockingly so, so it is a bit weird that it is not shared. Someone suggested it might be a chromatic aberration effect.

The oddest aspect of this, is that whenever I mention it to interested people, they often produce examples like yours that are supposed to induce the effect and I find them to be completely flat!  Whatever, it is that I see, I don't think it is quite the standard thing.

Another thing I have noticed is that effect goes away if I shut one eye and sometimes it is stopped by using reading glasses.

The most annoying thing of all is that, although I have seen many examples on the web that create the effect, I can't locate one at this moment to demo the content that creates it for me!

-- hide signature --
ProfHankD
ProfHankD Veteran Member • Posts: 8,172
Re: One word: Chromostereopsis

DMillier wrote:

Hi Prof

I am someone who (enjoys?) a colour based 3D illusion on occasion, where objects of different strong colours appear to "float" distinctly over the other as if the foreground were an inch or two above the background.

I've asked about this effect on the forums before because I've tested the effect on family members who were unable to see what I saw. I have asked an optician about this and got a blank look in response.

For me this effect can be very strong, shockingly so, so it is a bit weird that it is not shared. Someone suggested it might be a chromatic aberration effect.

The oddest aspect of this, is that whenever I mention it to interested people, they often produce examples like yours that are supposed to induce the effect and I find them to be completely flat! Whatever, it is that I see, I don't think it is quite the standard thing.

Another thing I have noticed is that effect goes away if I shut one eye and sometimes it is stopped by using reading glasses.

The most annoying thing of all is that, although I have seen many examples on the web that create the effect, I can't locate one at this moment to demo the content that creates it for me!

Some people see more chromostereopsis than others, and there is even a very small fraction that see it reversed. In short, it isn't 100% either. However, it has way higher agreement than other 3D effects....

 ProfHankD's gear list:ProfHankD's gear list
Olympus TG-860 Canon PowerShot SX530 Sony a7R II Sony a6500 Canon EOS 5D Mark IV +30 more
Sjak
Sjak Senior Member • Posts: 6,356
Re: One word: Chromostereopsis

Travis Butler wrote:

Separation from the background does help with a 3D effect, but is not sufficient; as someone said elsewhere in the thread, a flat cutout separated from a blurred background isn't 3D to me.

Sepration from backrgound helps emphasize a subject, but usually does not contribute much to the illusion of depth.

The thing that catches my eye is the suggestions that the object we're looking at has depth in itself - a.k.a. volume. Take a look at the right arm in #7, just to the right of the elbow - you can see a slightly darker stripe, suggesting a flexing muscle, something not present in most of the other versions. Likewise, the play of light across the subject's face, emphasizing the jaw line.

So for me, good tonal rendering - the ability to express a very finely gradated set of tonal variations, giving a sense of volume through the change in light as it plays across a subject- is an important part of the 3D effect.

For an example of my own:


Lumix S5, MD 24-50/4

Do you get a feel of depth from the clouds?

I think the volume you refer to is a result of the lighting, which we already identified as something that can contribute to perceived depth in an image. An analogy: front-lit portraits are often referred to as "flat" (= lacking depth)

For using lighting, we have several tools, such as our position in relation to the subject, the subject's position in relation to the light source, etc. Sometimes, we do not have control on all of these elements.

Back to your photo: The clouds are well-lit, and you recognized this as an interesting moment. In addition, your image has several layers (car, rocks, mode distant landscape, sky) which also contributes to the illution of depth.

Regardless of 3D-considerations, your composition is nice; too bad the car didn't completely fit in the image.

 Sjak's gear list:Sjak's gear list
Pentax K100D Leica M-Monochrom Pentax K10D Pentax K-1
DMillier Forum Pro • Posts: 23,021
Re: One word: Chromostereopsis
1

What I find most interesting about the occasions an image induces the effect it is 3D in same way that binoculars make things look 3D or that opening and closing one eye makes 3d appear and disappear:  the effect is hard to describe, it's as if everything suddenly gains "spaciousness" around it.

This has to be a brain constructed thing, rather than an optical phenomenon.  Something in the input is tricking the brain into constructing an illusion of height/depth that isn't there.  Kind like looking at pictures of craters on the Moon or Mars. Sometimes they look like bowls and sometimes they look like domes or bumps. And with a bit of effort you can switch the interpretation such that when it is a dome, you can't imagine how you could have possibly thought it was a bowl and vice versa.

I don't think you will find the answer in Zeiss...

-- hide signature --
Sjak
Sjak Senior Member • Posts: 6,356
Re: One word: Chromostereopsis
2

DMillier wrote:

I don't think you will find the answer in Zeiss...

Unless you say the image was made with a Zeiss (regardless whether it's true or not)  

 Sjak's gear list:Sjak's gear list
Pentax K100D Leica M-Monochrom Pentax K10D Pentax K-1
OP backsidewalkaround Regular Member • Posts: 255
CORRECTION
2

Sadly it took a while until someone found my mistakes that needed correcting. I've made two actually and I'm sorry about that.

The #7 photo is missing, instead I accidentally entered #8 twice.

The other mistake I made: The Canon 1.2/55 IS NOT part of this comparison,

it is the OM Zuiko 2/50 Macro instead (I used the Canon in another comparison, hence the mistake). Since I did not yet convey the solution this shouldn't matter too much, but I'm apologizing to Tons o Glass 0 Class who made a really strong effort to find out which lenses were used.

So the lens list really is:

Leica Summicron-R 2/50

Zeiss Makro-Planar 2/50 ZF

Olympus OM Zuiko 2/50 Macro

Leica Elmar-M 2.8/50 (unfair to it, because it's only f/2.8)

Pentax-M 1.4/50

Konica Hexanon-M 2/50

Nikon AI Nikkor 1.2/50

Contax Zeiss 1.7/50 with an 84... serial

Sony Loxia 2/50

Here go the samples again. The missing #7 is quite important in my opinion:

#1

#2

#3

#4

#5

#6

#7

#8

#9

-- hide signature --
 backsidewalkaround's gear list:backsidewalkaround's gear list
Sony a7R III Sony Vario-Tessar T* FE 16-35mm F4 ZA OSS Sigma 35mm F1.4 DG HSM Art Sony FE 50mm F1.4 ZA Sony FE 200-600 F5.6-6.3 +1 more
OP backsidewalkaround Regular Member • Posts: 255
Re: On guessing the lenses...

Tons o Glass 0 Class wrote:

At least in this DPR post, images 7 and 8 are identical. The links lead to unique files, but downloading them and comparing them yields zero differences. This begs the question, what lens is missing?

Thanks for pointing me to my mistake. I've corrected it in my last post now. You actually pointed me to another mistake too, which I correct there too. Sorry.

To me, all of the images look about the same when it comes to 3D-ness - I think they all have a mild amount of "it" that is largely in credit due to the scene and how well you crafted/controlled it. The plane of focus choice to give us the most depth cues is very consistent and well executed.

That is not to say that the lenses don't render differently. There are some obvious differences when it comes to global contrast, SA, shapes bokeh balls take towards the corners, color temperature, etc, but these differences don't make or break the 3D effect for me in this scene. They are also not sufficient enough for me to strongly identify lenses, almost all of which I'm unfamiliar with, but I'll try anyway! I'm prepared to look characteristically ridiculous.

I couldn't really observe any indication that any lenses in the shootout have been stopped down (no aperture shapes in bokeh anywhere). With some f/1.2s in the mix, I found that surprising. Perhaps where the aperture would show up in the bokeh the most is obscured by the skater or the lighting is sufficiently gentle there to hide it, and those f/1.4 and f/1.2 lenses just vignette a ton? I don't know. Well played.

Image 1 looks very Minolta-ish color-wise to me, but with none in the lineup, and with Minolta and Leica collaborating on some things apparently, I'll pin it as the Leica R.

Image 2 has to the lone f/2.8 judging by bokeh ball size. Elmar.

When viewing the images in order, Image 3 impressed me. it reminds me of macro lenses, so maybe it's the Makro. If the goal of the shootout was to find a keeper, whatever lens this is might be it for me.

Image 4 is looking like a modern optic overall (very contrasty) which has me thinking it would be the Loxia. It looks like there may have been a light falloff correction applied but that may also have to do with the changing light of the scene. It's the only image where I noticed what might be a hint of a hexagonal aperture shape, however, so maybe it's the 50/1.7 Planar. Undecided.

Image 5 looks Canon-esque to me, and when viewing them in order, it's the first one that looks like it could have been shot with a slightly longer focal length, so I'll guess this is the Canon 55.

Image 6 could be the Pentax. I've seen corner bokeh that looks like that with it.

Image 7/8, I'll go with Nikon for the heck of it.

Image 9 has some especially wonky corner bokeh (similar to what I think is the Pentax) so maybe it's the other rangefinder lens in the lineup, the Hexanon. But mid-frame the bokeh ball size has me thinking it could have been a faster optic... so would it be Pentax then? Dang it.

All in all, your reasoning is very good, albeit far from perfect. After you see my correction (of BOTH errors) you'll see much clearer, I suppose ;-).

-- hide signature --
 backsidewalkaround's gear list:backsidewalkaround's gear list
Sony a7R III Sony Vario-Tessar T* FE 16-35mm F4 ZA OSS Sigma 35mm F1.4 DG HSM Art Sony FE 50mm F1.4 ZA Sony FE 200-600 F5.6-6.3 +1 more
OP backsidewalkaround Regular Member • Posts: 255
Re: One word: Chromostereopsis

ProfHankD wrote:

backsidewalkaround wrote:

ProfHankD wrote:

Experiment all you want, but "3D Pop" is fundamentally NOT a lens property. It is a simple set of composition techniques long widely understood by professionals. Best I can tell, the only thing Zeiss lenses have to do with it is that they combine slightly stronger than average red rendering with costing enough that they tended to be used more often by professionals... who would know the composition tricks.

Thanks for the lecture Prof ;-).

Sorry. I didn't mean to lecture, but this is a VERY frustrating topic for me....

I can see why, and I can relate. Sorry for bringing it up ;-).

I admire, that you've tried to explore the topic scientifically. I already knew about the influence of color. Still I beg to differ.

As I understand it, you didn't succeed to prove any other clues than Chromostereopsis. But you didn't prove the contrary either.

I haven't proved anything, but I have very strong empirical evidence from about a half dozen surveys (probably 100+ people) that no lens property is consistently seen as producing 3D Pop. In fact, the closest thing to a lens characteristic that caused 3D Pop was lens brand: people told an image was shot using a Zeiss lens tend to see much more 3D Pop than if they don't know the lens brand used.

I guess individual vision must be more different than we think. You also state that not every one can see chromostereopsis, some even see it reversed. It's nice that this seems to be a less widely spread phenomenon, thus you could prove the correlation. 
All the discussion on the web leads me to believe that there may be other triggers for 3D impressions that vary more widely, so that you can only find smaller groups, seeing the same thing, which would then explain, way it isn't possible to prove a general rule, but it may apply to some.

3D Pop isn't a thing unless there is general agreement about examples of it -- which there pretty clearly isn't, except for cases traced to chromostereopsis.

For example, some people will say any image with a sharp subject and heavy, smooth, defocus of the background is 3D Pop, which is what a shot like http://www.stevehuffphoto.com/wp-content/uploads/2014/05/L1000262.jpg shows.

I already said, this is a prime example for my vision. I have a very strong impression of depth with this photo and it shows the typical properties for photos that give ME that impression: 
The subject is well lit from the side with clear directional light, given strong shadows on the other side.

There is a front that is out of focus, you then can follow transition of focus on the ground, getting into focus and back out.

The subject stands in the middle of the frame allowing you to feel that transition around it.

The top part of the boy is positioned much more to the front in comparison to the buildings in the back. This makes the transition from sharpness to the blurry background very sudden, making the boy stand out.

Other helping elements might have further improved the impression, like vignetting, the boy standing on a street with converging lines, etc.

However, many others (myself included) see that not as a 3D scene, but as rather artificial-looking stacking of two flat layers. If you go back to my 2014 first attempt to define 3D Pop as a set of lens characteristics , my proposed definition was:

My suggestion is that it is the combination of:

  1. Very high sharpness (microcontrast) and texture for the primary subject.
  2. Slight, but very smooth, defocus of more distant parts of the scene around the primary subject.
  3. Some areas of moderately strong colors and tonal contrast, but not the whole frame.

I would support that suggestion, but add some more clues, like described above.

and I really wanted it to be something like that because those are attributes I can recognize, measure, and enhance in postprocessing... hence, it would potentially have made a nice bit of publishable image processing work. But nope.

So there may still be other clues that are too difficult to grasp or to describe.

It seems that I need to elaborate on what I want. ...
I believe it is very difficult to talk about it, because different people have different perception.

That means "it" isn't an it, but a bunch of conflicting opinions.

Maybe it's not conflicting opinions rather than different vision (plus the inability to really describe it well).

Chromostereopsis, on the other hand, has been generally agreed upon for a very long time. For example, stained glass (one of the first media to be able to make strong blues) was using deliberate chromostereopsis centuries before cameras were invented.

Agreed.

I'd like to put it this way: Every person (capable of seeing 3D) has a set of eyes and a brain to process the images given by the eyes. Differences start with the eyes, since some people can't discern red from green as an example, but there may be many other differences that are never measured or compared, because the eyes work. No reason to talk about individual vision in depth.

Not sure what you're getting at here. BTW, cover one eye and look at the Red/Blue image I posted above -- most people will still see blue as behind red despite having no stereo vision cues at all.

So you admit that with flat vision (one eye only) the brain creates some form of depth that isn't there. Why shouldn't the brain do it in other ways too.

In sum, each of us is allowed to like whatever image properties we like, but there is no such thing as a lens producing 3D Pop. The photographer deserves more credit for their composition skills.

I never said, that one lens produces 3D Pop every time. I agree that without proper lighting and composition it will never happen, BUT, if lighting and composition are well chosen as a foundation, some lenses will help more than others to trigger the illusion. 
By the way, at least in my samples, I don't favour the Zeiss lenses.

-- hide signature --
 backsidewalkaround's gear list:backsidewalkaround's gear list
Sony a7R III Sony Vario-Tessar T* FE 16-35mm F4 ZA OSS Sigma 35mm F1.4 DG HSM Art Sony FE 50mm F1.4 ZA Sony FE 200-600 F5.6-6.3 +1 more
OP backsidewalkaround Regular Member • Posts: 255
Re: One word: Chromostereopsis

Travis Butler wrote:

I would pick out #7 and #9 as the ones showing the most 3D effect to me.

backsidewalkaround wrote:

It seems that I need to elaborate on what I want.

From my experience there are three versions of 3D rendering that need to be kept apart:
1. Verisimiltude: An overall very sharp contrasty image depicts reality as well as possible.

2. Pop: The subject pops out from strong blur. The surroundings are mostly unrecognizable or at least give not depth cues.

3. 3D rendering of a scene: For me those images give me the impression that I'm looking at a 3D scene (that is only 2D of course), but my brain makes me believe it.

I'll add in something from prior discussions on this topic here: Volume.

Separation from the background does help with a 3D effect, but is not sufficient; as someone said elsewhere in the thread, a flat cutout separated from a blurred background isn't 3D to me.

The thing that catches my eye is the suggestions that the object we're looking at has depth in itself - a.k.a. volume. Take a look at the right arm in #7, just to the right of the elbow - you can see a slightly darker stripe, suggesting a flexing muscle, something not present in most of the other versions. Likewise, the play of light across the subject's face, emphasizing the jaw line.

So for me, good tonal rendering - the ability to express a very finely gradated set of tonal variations, giving a sense of volume through the change in light as it plays across a subject- is an important part of the 3D effect.

For an example of my own:

Lumix S5, MD 24-50/4

Do you get a feel of depth from the clouds?

I can see what you mean. It's true your image also has a lot of depth. Light and shadows are perfectly shaping the landscape just like a painter would do it. So you're absolutely right that way.

BUT, it doesn't trigger the same illusion for me as do other images, mostly with a transition from focussed to unfocussed areas.

On the other hand maybe it would, if you had included the ground (to follow the focus towards the car from unsharp to the sharp back, starting from a low POV. Sometimes my brain just needs a little clue to "flip the switch".

-- hide signature --
 backsidewalkaround's gear list:backsidewalkaround's gear list
Sony a7R III Sony Vario-Tessar T* FE 16-35mm F4 ZA OSS Sigma 35mm F1.4 DG HSM Art Sony FE 50mm F1.4 ZA Sony FE 200-600 F5.6-6.3 +1 more
Gesture Veteran Member • Posts: 9,037
Re: CORRECTION
1

Snapshot.  Can view @ 100 percent.

OP backsidewalkaround Regular Member • Posts: 255
Re: CORRECTION

@ Gesture: That's very nice and it definitely has quite a bit of depth. While you succeed to prove to me that chromostereopsis works, I see the depth more in the green leaves especially the one extending to the front.

On the other hand the chromostereopsis works in an irritating way with the middle branches of the leaves behind the trunk. They seem oddly misplaced.

Personally I'm more fond of depth in images showing a scene on a larger scale, ideally from up close to infinity maybe a person standing in the street between tall buildings. Those fascinate me most.

-- hide signature --
 backsidewalkaround's gear list:backsidewalkaround's gear list
Sony a7R III Sony Vario-Tessar T* FE 16-35mm F4 ZA OSS Sigma 35mm F1.4 DG HSM Art Sony FE 50mm F1.4 ZA Sony FE 200-600 F5.6-6.3 +1 more
Keyboard shortcuts:
FForum MMy threads