Recently I’ve done a lot of reading in order to make the whole topic accessible to myself. Discussions in the net are very fuzzy, sometimes misleading or at least irritating. There aren’t many websites that try to explain in depth how some 2-dimensional photos look “3D” or “more 3D” than others. After quite a lot of reading I’ve come up with some opinions on the topic which I’d like to share with you in this article.
First off I’d like to point out that there are images that can be viewed three dimensionally. In fact they aren’t real 3D, but your brain makes them look real.
A few years back I’ve experimented a little with stereoscopic images. Stereoscopic meaning that each of your eyes is provided with a slightly different image at the same time – just like it is in the real world.
These stereo images are very cool and first (imperfect) results can be achieved very easily. These images are composed of two shots that have been taken from a different perspective (just like your eyes would view them). The most basic technology to view them is with red/cyan glasses.
If you’d like to try:
- Find a subject that has 3D potential. Take two shots about 5-10 cm apart (to the side). The closer the subject, the smaller the perspective-gap should be and vice versa.
- Try to frame the subject of your choice as evenly as possible. My shots aren’t very successful, because I shot them all without even using a tripod or even better suited equipment.
- Put the images together with Stereophoto Maker and view them on your screen using red/cyan glasses or any other 3D technology that you have access to (like a 3D panel and shutter glasses – then you’d have to use a different output method in stereophotomaker).
- In order to achieve better results than I did, you might have to use a slide rail on a tripod in order to best keep the perspective when sliding 10 cm to the side. Even better are two cameras mounted 10cm apart on one plate. For moving subjects the cameras would have take the two synchronized shots.
First (as I said – imperfect) results might look like this. To be able to see the effect you will need red/cyan (red/green, red/blue) glasses:
Real depth in stereo images (view with red/cyan glasses)
3D that is created by your brain
Images that are in fact flat and don’t use two different perspectives that your brain will form into one 3D image can also have a 3D “look”. This is a topic that is being discussed all over the internet, many times very controversely. The reason for the controversy is that there are different concepts of 3D.
I’m still learning about all this, but at this point I will suggest that 3 different types of 3D should be discerned:
- 3D Pop: The main subject (often being very close) is strongly separated from a background being far enough away that it is merely if at all recognizable. The sharp and contrasty subject pops out of the blur and sometimes seems to jump off the screen.
- 3D by transition: From what I’ve gathered most people agree that images that use some kind of transition from the subject to its surroundings suggest a strong feeling of depth. Most of these images use focus transition. I’ll get into that more later.
- 3D by lifelike rendition. Images that are sharp from front to back that show a great amount of detail with high contrast and especially micro contrast. These images are supposed to capture the scene like it is, give the observer a chance to feel like he’s there, give him the impression to be able to touch things in the scene.
For all scenes some basic knowledge about what can support 3D impression (in any way) can’t be wrong. In my opinion the following article sums it up quite well: Creating Depth in Art and Photography. Many aspects named there require plenty of planning for each photo. So what’s most important in photography?
Light and shadow
Light that helps with the 3D perception needs to show the shape of a 3-dimensional object. To do that, I think it needs to be directional (from one, two sides or maybe even more sides) to form bright areas and dark shadows, the more pronounced the better. Very important are clear cut borders.
A quick drawing will illustrate the basic idea:
flat – depth – more depth with added perspective
I think in general photography one important contributor to the 3D-look is contrast. If the light is flat you will have to pull quite a few other tricks up your sleeve to still achieve a 3D look.
Not long ago I was reading a very lengthy thread on 3D rendition in the German DSLR-Forum. An article by Pavel Kosenko was mentioned and linked. The images in this article were perceived by most observers as excellent examples for the 3D effect. I agree: The very contrasty shots were created by someone who knows the craft of lighting: Depth by contrast
In my opinion the perceived depth in the images on display is mostly created by very high contrast and sharp object borders. Bright areas quickly fall off to very dark grey or even black. Borders of well lit objects have very defined edges with an abrupt transition from bright to black. Shadows are very dark and pronounced.
I’m not talking about the much discussed micro contrast, but rather contrast at a bigger scale, the one that you will achieve by using a steeper contrast curve rather than the clarity slider in Lightroom. Sure, micro contrast also improves the depth perception, but less so than that of the contrast curve.
I will try to give a few examples, that I created from pictures of mine by enhancing contrast with a steep curve. For me the depth of each image improves quite a bit (mostly opposed to the visual quality of the image).
- In the first pair the 3-dimensionality of the statues is much better. Also they seem to stand further apart.
- The first image of the second pair is rather flat while the other much better transports the varying distances of the houses. This image actually got better too.
- In the third pair the forearms of the diver really seem to protrude to the front, but much more so in the second image.
- In the forth pair the square in Pompeji shows plenty more depth when contrast is raised.
The importance of contrast can be comprehended if you think about which light you prefer for your own shots: As a landscape photographer you will certainly pursue the light not long before the sun is setting or just after sunrise, not only because of the special quality and warmth of the light, but also because it can best shape a landscape. This kind of light brings along alternating bright light and dark shadows and gives your image that particularly graphic look.
Bright and dark surfaces taking turns will certainly also be a successful recipe for powerful architecture shots.
In case of your subject being fairly close, the same rules apply, but you can provide the lighting yourself using strobes. This enables you to bring out the shape of things just the way you want it to be. If you’d like to go down this road I can recommend a good read starting here: Lighting 101
Naturally this is the most obvious one. This is what you learned in school, when you were supposed to draw a cube using a vanishing point.
Whenever your image shows lines leading towards a distant vanishing point, your brain will tell you from experience that there is depth.
Typical examples are the columns in a portico, a long row of trees in an avenue or the view towards the altar in the nave of an impressive baroque church.
The lines don’t necessarily need to be straight or actually have a vanishing point. In my opinion it suffices for them to support the impression of distance, like a winding road becoming more and more narrow as it leads away or lines getting closer in the distance.
Depth by perspective
Depth by transition
Three dimenionality is also perceptible when a subject in the foreground is separated from the background. This may be happening in various forms. Most used in photography is probably small depth of field and with it the transition from sharp to unsharp areas.
In my connotation, two terms should be differentiated: Pop and 3D. To me pop occurs, when the subject is separated without transition. Often nothing or hardly anything in the background can be identified while the subject in the foreground is sharp. It can also occur if the background is recognizable but quite far away. This results in a very sudden change of sharpness from the subject to the unsharp background.
3D by focus transition
On the other hand 3D occurs, if the transition is more subtle and smooth and the background still lets you recognize the surrounding area. Ideally several objects like pillars become more and more blurred as they are further away. Most often it will be the ground showing this transition. This also means that it might be wise to choose your shooting angle quite low.
I’ve gathered from reading other opinions on the web and from retracing of my own that the 3D impression mostly is stronger if the background is only a little blurred.
A similar 3D effect will also be there if the transition is achieved with receding contrast. If for instance you take a picture of people walking in the fog or you have several layers fading away more and more in the distance. With landscape images this will often be true. Due to humidity in the air, layers that are further away will appear lighter, less contrasty. To my eyes this doesn’t work as well as focus transition, but it does help.
Depth by receding contrast
Orio Menoni from MFforums shows some images with great depth here: 3D discussion on MFforums. One image in particular has shown me that color can be a very strong component to perceiving 3D: Depth by the color red
If you look at Orios image, try covering the red part with your hand. Then switch back and forth, covering and uncovering the red part. Can you feel the 3D force of the red color?
Other factors of 3D-supporting lens rendition may be the following:
- A vingette will support the impression of a tunnel view and thus enhance the 3D-look in many scenarios. This will especially be true, if you have a vantage point in the center of the frame.
- Close objects will literally bulge out when the lens displays a certain amount of barrel distortion. If used wisely, this may help creating a 3D look.
- Field curvature bending the sharpness of outer parts of the image towards you, may support the quick transition from sharp to unsharp areas and by that the 3D impression.
- Medium or large format sensors will naturally be showing more 3D. It’s about the way the sharpness falls off and how the blur is rendered unsharp yet recognizable. Also sharp areas will have more sublte detail to them, more tones.
The following images were created during a discussion I had on FM forums (after publishing this article, so this is an addendum):
Readers there didn’t think my contrast examples on top were good for showing 3D. They showed me what they thought were better examples. I tried to find as many common denominators as I could and tried to manipulate the Pompeji image from above (contrast examples) to look more 3D.
I added some vingetting, some contrast, enhanced some areas with dodge and burn (to make the border transitions even sharper and more abrupt) and for the most important part added artificial focus transition (Gaussian blur), the transition was achieved using a mask and the gradient tool and some manual brushing for a fairly good transition. This is the result. To me it indeed has a better 3D look than before.
To verify I did two more, only using artificial focus transition with Gaussian blur. I’m not stating that this should be done, I’m only trying to prove that focus transition works.
Creating the 3D look
If you’re trying to create a 3D-looking image (by transition), not the lifelike kind, try to consider as many of the following criteria as possible:
- I think most important are clear cut borders: Be sure that the borders of your subject are well defined against the background, be it by sharpness, brightness or whatever other visual differentiation is possible. If you use transition by focus, make sure that your subject is somewhat isolated against the background on all sides.
- The subject needs to be very sharp and contrasty. Shiny objects seem to work well, like wet things or cars with a metallic surface, because with the right light they make it easy to get tones ranging from black to white.
- There needs to be a gradual transition, be it by focus-, contrast-, brightness- or whatever falloff. This is the one criteria that differentiates 3D from just being Pop.
- Use colors that spring to the front, like red or yellow for the subject of your choice.
- Use a composition where lines show the depth of the scene.
- Instead of lines (or additionally), layers work as well: like pillars showing up at different distances at a different level of sharpness/blur or brightness or contrast.
- Don’t let your suject take up too much space in your frame. If you get too close it won’t help your case.
- Don’t place your subject close to the border, because your brain can’t feel the depth on the other side. It works better, when the depth is perceptible on both sides of the subject.
- To make focus transition more obvious, it helps to do a Bokeh Panorama (also known as a Bokehrama or a Focus Panorama or “Brenizer” image – after Ryan Brenizer who made it popular with his images).
So… do you need a Zeiss lens for 3D in your images?
No – eh well …
At least for 3D by transition or pop it is obvious that it works with other lenses too. It can be achieved with the right lighting, well chosen perspective, post processing and other helpful traits of an image.
I think for all those that don’t want to bother with using Raw, contrast curves, clarity, etc., for those who simply use in camera processing, high contrast lenses, like Zeiss glass, might give their images the edge they’ve been looking for. Also older contax glass may have other traits like field curvature that helps the look naturally. But I’m very sure that the whole thing is far from being limited to Zeiss or Leica. For instance by a fellow reader of the DSLR forum I was pointed to images here (especially the first one with the tree): Topcor 3.5/5cm
Also for those who do use the computer to enhance their imagery, some lenses will give them a head start. From many opinions I’ve read and images I’ve reviewed all over the net, I’ve come to the conclusion that Zeiss focussed on designing lenses for high contrast rendition. It’s not so much about superior sharpness, it’s more about atmosphere.
Some gems that I can’t talk about first hand, seem to be the Contax 2.0/28 “Hollywood” its slower brother 2.8/28, the Contax 1.4/35 Distagon or the 50 and 100 Macro Planars. From my own experience I can say that the Sony Zeiss 1.8/55 Sonnar supports that claim as well.
The secret ingredient of Zeiss glass
I think it’s mostly about their particularly contrasty rendering. By now I would describe it as Micro Contrast, which is another term being conceived in many different ways.
I’d like to define it as followed:
It’s the same thing you get if you use the clarity slider, but at the same time you don’t loose detail. Micro Contrast is the ability of a lens to keep tonal differences clearly differentiated, maybe even accentuated. If you will, it’s accentuated resolution. This is not only about luminance, but about color too, making colors often appear more saturated.
But that’s not all, it also emphasizes larger areas, spreading tonal values widely, I’d call that local contrast. Say we define brightness values in % ranging from 0% (black) to 100% (white). With the exact same camera settings one lens may render the tones on a face with brightness ranging from 50% to 80% while the other shows values ranging from 30% to 90%.
In my experience the latter may be the Zeiss lens. While the Zeiss lens may not be particularly flattering for portraits, its images will look more 3D, because of the stronger contrast. On the other hand in some situations, they may more easily have blown out highlights or areas of complete black, while the other, less contrasty lens still lets you record different tonal values. Which is preferable is a matter of taste or situation. In a low contrast situation, Zeiss glass will give you extra bite. In a high contrast situation, another lens may dim the contrast to a recordable level, but with modern high dynamic range sensors, blown out highlights are quite easily avoided, so this gives the preference to high contrast lenses. Unless of course you prefer more subtle tones in the midtone area. While Zeiss glass is known for this kind of rendering by design, there certainly are othere lenses out there that can keep up.
One forum thread that will support the claims just made, can be found here (the images have been processed, so this may have a strong influence on the outcome):
Zeiss vs. Canon – can you tell?
Which lenses give you the best 3D raw material?
I want to go a little further on this and will do some research with my own lenses. I will try to find out which ones are best suited for the 3D look and I will try to find out if Zeiss lenses really are the ones that excel at this. So stay tuned.