Summilux E55 1.4/50

NIkon 28/2 AI (1981) wide open

Donuss wrote:i was made with a Zeiss 35mm 2.0

Three dimensionality or 'contour definition', the CZ term, are both much more accurate descriptions of the phenomena than the ambiguous and misleading term 'plasticity', which is a term perhaps best used to describe the skin of female models shot with 'pro zooms' used by 'pros', that is: a 'plastic look'. It is to be hoped it does not catch on, therefore. '3D' has a neat ring to it as well.

3D is best judged in images that, to a large extent, lack the post-processing tricks, happy or designed circumstances of light and most of all, thin DOF. These are all manipulations of photographer and nature that conspire to fool the viewer's visual system.

For example, in viewing an image, our eyes move: from light to dark, from sharp to less sharp, from in focus to out of focus, from colourful to less colourful, from a main subject to subjects of peripheral interest, from clear to less clear, from looming foregrounds to de-emphasised backgrounds.

All these manipulations are easily observable in the linked image provided by AhamB...

http://1x.com/photos/member/17110/33969/

...in which the eye is led to ascertain the image within tight constraints - it cannot wander around looking for secondary objects of interest because there are none. [For many photographers and viewers alike, the art is all about keeping the message simple...not a particularly noble goal, arguably.]

Thin DOF, focus fade, and objects that imply three dimensionality (such as Rustybug's bell on the last page) beg the question: what if the photographer had used aperture f16 - would this image then have any '3D'? Generally the answer is no.

Lenses that exhibit high levels of 3D can do so in less than optimal light conditions and in the absence of overtly obvious visual cues. What it really depends upon is the accuracy of the shaping of objects in image space, both longitudinally (away from the camera) and tangentially (perpendicular to the axis, or 'across the frame' in image space). You can observe it even in quasi-planar subject matter - an example is Anden's fine image of 'Fall colours' in this thread:
http://www.fredmiranda.com/forum/topic/926319/3

To get the best in 3D one needs clarity and excellent resolution, not necessarily of the highest order (Leica output is not noted for high 3D performance) but a certain level of contrast is needed, and it should be consistent in both directions - away from the camera and across the frame.

Since we are talking about human perceptions here, quite far removed from the more verifiable technical issues that seem to obsess us to a great degree, the kind of feelings to look for include: 'Do I feel like I can perceive depth in the image, the feeling I simply can walk into the scene?' (for non-close up compositions), 'Do I feel I can reach out and feel the shape of that object, having observed it depicted so accurately in (seemingly) three dimensional space? (for close ups), and 'Would I feel that way if the image lacked the assistance of the factors listed in para 2 above - the sometimes hidden persuaders?'

Of recent images in this thread, the two by adamdewilde on May 28 exemplify the points made above. Lighting is unhelpful, as it is backlighting in a longitudinal direction, there is no central object that forces eye compliance, light is quite even on all table top objects, sharpness is there for the two nearest objects and is close for the other objects - even the window is drawn quite well, strong colour is present in all objects, as is contrast, and all objects are clear and well lit - crucially there are no deliberate or fortuitous manipulations that attempt to force 3D on the eyes.

Perceptually, however, one feels like s/he could just 'reach out and place an open palm around the vase'...

Now to the lenses adamdewilde used. Zeiss have, as a design goal, the depiction of objects in image space with high levels of countour definition. The MTF charts for these two lenses are shown here:

http://www.zeiss.com/C12571FF00438F7A/0/4DBF8F5EBA9BB366C12576930030E0A3/$file/planar_14_50_en.pdf

and here:

http://www.zeiss.com/C12571FF00438F7A/0/4BF3C3E2245DC957C12576930030E1B0/$file/mp_2_50_en.pdf

Zeiss maintain that faithful contour definition is indicated by MTF lines (all six of them) that are close to horizontal across the image space represented by the MTF chart's horizontal axis. You may wish to take a look at the f5.6 charts for each lens. Much, much more here:

http://www.zeiss.com/C12567A8003B8B6F/EmbedTitelIntern/CLN_31_MTF_en/$File/CLN_MTF_Kurven_2_en.pdf

Click the Binocular icon at the top left of the PDF and enter: 'contour definition' for references.

I questioned the terminology being used too - earlier on in the thread. Contour Definition seems to fit much better! It actually has a meaning where as "3D" is already taken and quite diverse!

Actually, I don't think one can obtain 3D feel or contour definition without sufficient DOF. Thin DOF would generally produce a flat subject on the blur background. Indeed, micro-contrast and lighting are important, but thicker DOF is needed to convey the relationship of the subject with the back/foreground. Even though our eyes would initially isolate the subject in focus, it would move away from it to establish the relationship of the subject with the whole scene. Gradual transition between IOF and OOF is needed to produce the contour definition one might perceive.

However, that doesn't mean shooting at f/16 would be sufficient. Due to the lack of isolation and due to diffraction, the subject wouldn't be able to "pop". I've been told that a wide-angle would be easier to produce the "3D" picture, while a telephoto would be more difficult.

So if we look at the "bell" example, the lit half of the bell and part of the frame really pops out. This is because the far front of the bell frame acts as a foreground element. And the back frame does help to complete this sense being a tad OOF and darker. Notice that the post, which the bell is attached to, is flat. The blurriness of the background in this case is quite irrelevant as long as it doesn't distract the viewer from the subject.

I can see the same concept in Paul Yi's pictures, and some others in the 35/1.4 thread. So in all, "enough" DOF is needed. But what's enough have to be done with trial and error for each aperture and subject distance.

Here's a picture taken with my Voigtlander 125

I'll confess I've only read the first half of this thread, but plan to finish it later. I just had a few snapshots this morning which I thought exhibited a lot of dimensionality, or contour definition, or whatever you want to call it. So I'm just asking for a bit of indulgence, and if others would provide opinions on whether these exhibit a "3D" effect. I'd rather not identify the lens, at least not immediately, for fear that it might bias the results.

The Nikkor 105 AI-S?

maybe its just me, but all I see is a shallow depth of field, I don't see any "3D"

Todd, nope, no 3D as far as I can see, but as you know there is no consensus as to what should be included under the "3D" banner

This is what I consider to be 3D (long live the bench! ):

denoir wrote:
Todd, nope, no 3D as far as I can see


Aw, darn.

Yes, I realize it is difficult to find consensus, but with you and BenV I'm 0 and 2 so far

As for the shots you posted, I see it so clearly in the Leica shot, and in the first Zeiss shot, but not so much in the middle two. Maybe I'll find someone to agree on my shots, and then I can politely reject yours and Ben's opinions.

These were shot with the Voigtlander 75/2.5 SL.

Todd, in the first two I don't see any 3D at all, just a layer in focus and another not. In the third one it gets closer, but I find it a little too saturated, and a little too bright. It could also take just a tad more sharpening for this size. All of that might help. I am fairly sure that "believability" is important for 3D, so when some aspect deviates too much from reality, the 3D gets lost.

FWIW, I don't see such strong 3D in Luka's shots either, except the crates. The Leica shot looks fairly flat to me, I'm sorry to say.

However, I really get a sense of the shape of that moth up above. The flowers and the background not, but the moth I feel that I can reach out and touch and feel the shape of.

carstenw wrote:
Todd, in the first two I don't see any 3D at all, just a layer in focus and another not. In the third one it gets closer, but I find it a little too saturated, and a little too bright. It could also take just a tad more sharpening for this size. All of that might help. I am fairly sure that "believability" is important for 3D, so when some aspect deviates too much from reality, the 3D gets lost.

FWIW, I don't see such strong 3D in Luka's shots either, except the crates. The Leica shot looks fairly flat to me, I'm sorry to say.

However, I really get a sense of the shape of that moth up above. The flowers and the background not, but the moth I feel that I can reach out and touch and feel the shape of.

It seems crazy how much variation there can be in opinions here! I was about to concede that perhaps my third shot didn't have it, but I really think the first two do, for me, and yet you say my third shot comes closest for your eyes.

I don't see anything 3D about the moth. And Luka's last shot, the bench recently posted in another thread, just popped out at my eyes as 3D the moment I saw it.

Well, as I said, no consensus overall

There are a couple of people that post in the Zeiss thread that tend to agree on a common subset when it comes to 3D (Philippe, Makten, Samuli, me and a couple of others) - all my shots above would fall into that subset. I can however produce shots where I see 3D and where they would not. Looking at the alt forum in general there tends to be very limited agreement as to what constitutes 3D.

By the way, in the comments that I received when posting the images previously, #2 (pipes) and #4 (leica) are the ones who have been 'accused' of 3D most by others.


I myself think it's a label for various rendering qualities that in one way or another are related to the spatial interpretation of the shot. In many cases it's a question of degree but also simply that there are different kinds of '3D'. I'll give you a few more shots that according to my interpretation fall into that category - although the rendering quality of these is quite different from the one posted above.



1. The medium format look of the 35/1.4:

Well, I am of the volumetric school (i.e. a strong sense of depth is not what makes me think 3D, but a strong feeling of shape and texture is), yet I see lots of 3D around the bumps on the snails "neck", and also around the edges of some of the leaves on the right.

Is it just me or is "3d" (still I dont think we can confuse this with contour whatever or any lens quality specifically) a way of photographing things and not specific to a brand? All of the true "3d" shots I have seen here look like they have adequate depth of field but still have an unrendered blurry section off in the distance, very similar to our own eye. Its not razor thin DOF, or very deep, and more importantly there are things in the IMAGE that match the DOF. This gives it its volume.

That being said, if anything I would say the Zeiss and even Leica "magic" look comes from the fact that the lens's seem to render the border between the infocus points of the DOF and the OOF points very smoothly, something that other lenses do not do well (ie cream machine nikon 85 1.4). This gives a smooth transition from your point of focus to the bokeh

BenV wrote:
maybe its just me, but all I see is a shallow depth of field, I don't see any "3D"


It's not just you. I don't see any either.

Here are two shots that aren't thin DoF that I feel have a spacial character to them...

The subject matter unfortunately intrudes into assessments of '3D'; not all images taken with lenses known to frequently produce the effect demonstrate it - again unfortunately. I think it is a matter of degree therefore - and it is strongly influenced by light conditions. I am now processing hundreds of RAW images shot with Tamron's surprisingly good 17-50/2.8 and the CY 35-70 and 28/2.8; after weeks of looking at them shot in the same conditions the Zeisses really show a greater propensity to 3D, but not all the time

Much of denoir's excellent 35mm Distagon work does show it very well - the 35/1.4 really shapes objects well despite being quite low in fine detail contrast wide open, as shown by the MTF chart.

Paul Yi's latest images in the CY thread with the Contax 100/2 do so as well (I think), so that is a good example of a tele lens re 3D. You can expect Zeiss lenses to be more likely to give you that effect because they aim for it. Brainiac used to say that Leica images looked like beautiful paintings and showed soft dreamy bokeh rather than the more defined CZ bokeh style. In great conditions probably most lenses will do a fine job of it.

And the 21mm is of course a killer lens for 3D and just about everything else you could ask for in a super wide.

As I said before, getting 3D out of a certain lens requires experience. Zeiss lenses are known for their 3D readiness, and one sees accidental 3D in many photos, but it does help to know what are the conditions that will invariable produce this effect, in order to produce it on demand. In my experience, 3D can be obtained with a choice of subject placement, aperture and lighting (as mentioned by Philip in the above post). The secret is mostly in the framing in order to produce a strong foreground/background relationship. But of course, if a lens is known for flat rendition, it would be very difficult to obtain any 3D effects from it.