Here's an example of what I'm talking about from large format.

This is an 8x10 contact print (i.e. from an 8x10 inch negative), printed as a cyanotype (UV-sensitized ferric ferrocyanide printing) onto calligraphy parchment. It's sort of a work in progress, as I'm reprinting it with a goal of getting deeper shadows. I toned it in black tea to get the cream-colored highlights.

But the luminosity of the highlights, and the gradations of highlight detail (especially on the boat on the right and in some of the buildings) really makes the textures stand out.

Subtle differences but the top photo is my vote for best and middle as worst.

Textures do help bring out the 3D effect. So it maks snse that a large format would be benificial. Hre's on on a small format where I think the texture helps the 3d effect:

Paul, if what you are getting at is dynamic range may contribute to the 3d effect then I agree. Nice shot by the way!

I also think that the differences a lens renders contrast also effects our perception of depth. This can be seen as the profile of the different line pairs on the MTF charts. I know that some lenses will show a ring around slightly oof large objects and that makes it stand out.

One thing for sure that effects the 3d effect is sensor/film size since the magnification changes the rapidity at which lenses transition from in focus to out of focus areas. I think this is why medium and large format have more of this 3d effect. Some correct me if I'm wrong.

EricH wrote:
Paul, if what you are getting at is dynamic range may contribute to the 3d effect then I agree. Nice shot by the way!

Thanks, Eric (though I'm not sure if you were commenting on my photo or the nice one from the other Paul!)

I don't mean dynamic range per se as much as I mean contrast. Ansel Adams (for one) was extremely systematic and deliberate about tone placement and contrast. Contrast is related to dynamic range, but you can obviously contract high dynamic ranges for printable contrast, and expand low dynamic ranges for more vibrant contrast. I think the presence of detailed highlights (Ansel's zone 7) and textured highlights (zone 8) are extremely important, especially if there are textured and detailed shadows (zones 2 and 3) to counterbalance them. The midtones hold it all together, but our eyes get drawn to the detailed highlights.

One thing for sure that effects the 3d effect is sensor/film size since the magnification changes the rapidity at which lenses transition from in focus to out of focus areas. I think this is why medium and large format have more of this 3d effect. Some correct me if I'm wrong.

Yeah, as I've said that's absolutely true, but this obviously isn't a factor in photos (like my cyanotype) in which everything is in focus. Ansel disdained selective focus, and despite knowing his work fairly well, I can't think of a single shot of his in which something is out of focus. He shot at f/32, f/64, f/90, and used tilts to get all subjects in the plane of focus. And yet his prints, when you see them in person (and I've seen about 40 original prints of his in person), look so 3D that you could just about walk into them. His extraordinary sharpness, texture, and detail are part of it. His control of tone and contrast is most of it.

Eric, If you were referring to my shot. While the dynamic range probably helps (this was a 30s exposre at ISO6 - Kodak SLR/c), I think the textures on the rocks really bring out the 3D effect.

DrPablo, I'd comment on your shot, which I'm sure is excellent, but your host is blocked on my work computer.

Strange, it's PBase...

At any rate, here's the link if you want to look:

http://www.pbase.com/drpablo74/image/77837184

DrPablo - thank for the link but the problem is that PBase is blocked at the root level. Anything at PBase can't be viewed from my work computer. Some of its It's content is too controversial. Almost all photo hosting sites are included. It's like that for all major corps that I've worked for.

Paul (DrPablo) Yes I was refering to your shot but now that I've seen the other Paul's (pdmphoto) shot I can say both are nice.

I've been going through a bunch of my images looking for ones that really exemplify the 3D look - and I really do see that my Leica lenses are giving me this even on my canon bodies, but it is more pronounced on my DMR than on my 5D. Now I don't know for sure whether this has to do with the fact that the DMR has a crop factor of 1.37 or that the DMR has more dynamic range. All I can conclude so far is that a portion is due to the lens and a portion can be attributed to the sensor. Now as far as the lenses are concerned it appears that the older lenses render as much 3D as the new ones, and I think the new ones are sharper. I am now thinking that while there appears to be number of factors that all contribute to the effect, that the contrast rendering for large objects is an important factor. Of course I am just an arm chair lens junkie and anyone with real optics knowlede is welcome to set me straight.

I Am Enjoying The Debate, Discussion.
Many Ideas Have Been Put Forth, And Some Ideas Dispelled...
One More To Muddle Over. This Shot Has, In My Opinion, The 3D Effect.
It Was Shot With A D30 (Not 30D) & A Canon 50mm f1.8. How Does That Fit Into Some Of Our Ideas....

bathman wrote:
Where does the 3D look come from?

I tried to do a picture which has a bit of a 3D look with 3 different lenses and I can not tell there is much of a difference between the pictures. Maybe some of you will see the difference, can anybody really tell which shot was taken with a Zeiss?

I would like to know where the 3D looks comes from and what the photographer would need to do to get some 3D out of his pics. Thanks for sharing your know-how

3:-D

IMO, the "3D look" has more to do with the sensor than it has with the lens.
Photos taken on Sigma DSLR's with Foveon X3 sensors are often referred to as looking very 3D.
But if you mix Zeiss with Foveon it can certainly help to give better 3Dness IMO...SD10 + Carl Zeiss 85mm f1.4 Planar T*:

Seems to me it's background separation inmany of the images people are posting. Not necesssarily a halo, but a really crisp separation between subject and the rest of the image.

Great thread! I've looked at all the images included here and 90% of them I would consider to be flat, I think some people are confusing 'sharp' with 3D. I'll try to explain what it is that I think my brain is trying to do when it perceives an image with great 3D qualities and I'll use the superb shot of the M4 on page 3 by Rico as an example.

If I scroll the image on screen so that I can only see the bottom half of the image starting at the lens then I see a very sharp image that has great detail but no depth because it's all in focus.

Now, if I do the opposite and scroll the image so that I can only see from above the lens to the top then my eye is instantly attracted to the detail in the in-focus and sharply defined area on top of the camera around the dial and the "Ernst Leitz GmbH Wetzlar Germany 5cm" inscription. At the same time my eye is perceiving sufficient detail in the background blur to try and resolve it into focus but it can't and it doesn't want to let go of the very detailed subject and I think this 'confusion' is causing a perception of depth.

So the attributes you need for me to perceive a 3D image is a lens that can resolve tiny detail in a subject yet render just out-of-focus areas with the right level of detail.

Anybody agree/disagree?

My two cents:

I think that the main ingredient is the combination of the relative sharpness of the focused portion plus the transition from focused to blurred.
This combination is of course more apparent in the pictures with a shallower DOF. This is why - I think - medium format images look more 3D than 135 format images, and, why wide aperture images look more 3D than f/8 or f/11 images.

I have many Carl Zeiss lenses and in my opinion their main quality is how they render the transition from the focused parts, which are always very sharp (as Zeiss are quality optics), to the blurred parts, where many Zeiss lenses have a bokeh that is -in my humble opinion- of superior smoothness (for being modern day lenses, that is).

Carl Zeiss lenses can combine both the sharpness and detail in the focused part, and the smoothness of the blurred parts, in a single shot, and the perception of this difference gives the 3D effect. In my very humble opinion.

Of course there are other factors concurring as well. The direction of light is very important: lateral lighting always increases the sense of depth. But these are things that everybody knows so it is useless that I repeat them here :-)

I think you're both making good points, but again, unless your argument is that prints never have a 3D look if the whole image is in focus, then you simply can't rest on out of focus quality or transitions to explain a 3D look.

For a lens or a format to nicely render out of focus areas is neither necessary nor sufficient to explain this 3D look.

In images that do have selective focus, I think that the transition between in focus and oof areas can exaggerate a 3D look. But I'm firmly convinced that the 3D look comes from having rich and bright textured highlights, dense shadows, and subtle midtone transitions. Those are the features that make a photo feel like you can reach in and touch it.

This is a picture of mine taken with a lens that has fairly crappy bokeh, the Canon 50 f/1.8. I think the picture indeed has a 3D effect. But do you really think it comes from the background? I think it comes from the play of light and texture on the guard's face and uniform.

As I said earlier I think it is the tonal range/play of light and I think zeiss lenses tend to bring out more details (probably because of the coating and not optical formulae) than canon lenses

I'm an owner of three Zeiss lenses (for Hasselblad) myself, so I concur, and it probably has to do with both coating and correction of optics. If it were just coating, it would be pretty cheap for Canon to match that quality of Zeiss lenses. Incidentally, I think Schneider and Rodenstock lenses are every bit as stunning as Zeiss if not more, though the format difference makes it an apples/oranges comparison.

But even if you have cheap lenses, as I think I've illustrated above, you can achieve a 3D look if you use texture and lighting wisely.

zakk9 wrote:
. . . but it's a well established truth within most forms of visual art, that some colours seem closer (warm reds in particular) and some seem more distant (cool blues).

Brent's excellent river bend on page four of this thread is a good example. The red in the background pushes the cliff and the river forward, making them coming "out of the frame".

Edited by zakk9 on Apr 27, 2007 at 08:56 AM GMT

Great thread! There are probably several elements at work here as many have stated but I wanted to emphasize what zakk mentioned in his post because I see its influence in several of the color photos posted, the latest being Dr Pablo's.

Warm colors do in fact focus closer than cool colors. Based on what I have read, it is a "human" form the chromatic aberration that we so dearly love in our lenses .

You can try this out yourself by placing a red rectangle on a blue background in Photoshop. Look at it in a dark room for added effect.

Here is a paper (PDF) on the effect. You can jump to the conclusion if scholarly papers are not your cup of tea (delayed pun intended ). For those of you arguing that there are other elements at play, you will be pleased to read that the paper concludes that as well.

The Real Effect of Warm-Cool Colors - Bailey Grimm Davoli

Very interesting article, Dennis.

Of course there are other factors at play, because there are plenty of apochromatic camera lenses, as well as black and white images that have a sense of depth. But this is one interesting facet of it.

I might have missed it, but the article didn't mention the blue effect of atmospheric haze as you look at distant objects (which is incidentally the reason why blue-blocking filters will make distant objects look much sharper on B&W film). I suppose it makes sense that "warm" wavelengths would be preferentially focused on/near our fovea, i.e. the area of highest acuity, while "cool" wavelengths would be focused more peripherally, i.e. it would preferentially attract our attention to things that are closer and "warmer".

Okay, so far I have found these persuasive as elements:


- good sharpness and microcontrast on in-focus areas
- use of lighting to bring out textures on the subject
- use of DOF
- use of perspective
- use of color
- use of framing

The cross lighting on the human subject in DrPablo's latest example seems to help a lot.