Pseudo-Hologrammetry
A free motion graphics technique tutorial by mfdh of StoryZoo
29 September 2003
Pseudo-hologrammetry is a fancy name for a very simple effect, in which a two-dimensional image is given a hologram-like appearance. It is a quick, pain-free way to add some life to flat artwork. I came up with this recipe when charged with the task of creating a full minute of interesting video using only twelve still images with just one day of production time available.
Strictly speaking, this recipe does not reproduce the look of a classic flat hologram with its distinctive chromatic scintillation as much as give the impression of flat artwork forcibly "dimensionalised" and projected through some yet to be invented technology, like a very simplified version of the effect seen in the movie Minority Report (Universal, 2002) in which we see flat video footage interpolated into semi-dimensionality and projected into 3D space by an array of tiny light-guns.
The geeks at ILM spent a lot of money and sweat figuring out how to achieve that cool effect. In contrast, the trick described below is cheap and simple. Ready? Let's go.
THE BASIC IDEA ...
Click to enlarge above images ...The original image (above) is clipped into layers (right), which are separated into channels (left), then offset along the Z axis (below) with a screen filter. First of all, the artwork must be divided into two or more planes. In the most basic example we would have a foreground element and a background element; in a more complex example we might have several more intermediate planes. In the example file below you'll note that I have used four planes (background, midground, subject and foreground).
Once you have your sliced layers cut apart in Photoshop (or an equivalent image editing application), create "shadows" for any foreground elements that still appear on the other planes. For instance, an outline of the elephant is created from the subject element, filled with black, and then applied to the layers underneath at 70% opacity. This darkening will represent where source images have been "lifted" to higher layers to force depth. (Do not use 100% black for the shadow, or it won't generate any connective streaks when you apply the volumetric lighting at the compositing stage, below.)
Next, duplicate each of your planes until you have three of each. Use the Channel Mixer (Image > Adjustments > Channel Mixer) to cause each version to carry only one channel of colour information. In other words, your top layer should be set to 0% RED / 0% GREEN / 100% BLUE and your bottom layer should be set to 100% RED / 0% GREEN / 0% BLUE with green between. Finally, change the transfer mode for all of the blue and green layers to Screen but leave the red layers on Normal.
The subject plane split into RGB channels with a normal transfer mode on all layers (left) and a screen transfer mode applied to the G and B channels (right). Import your artwork into After Effects (or an equivalent motion-graphics compositing application) as a composition, so that all elements retain their registration. Reconfigure this composition to your output specifications (720 x 486 for NTSC video, for instance). Add a 3D Camera and toggle all layers to operate in 3D space.
Using the red version of your background as a starting point, offset all subsequent layers negatively along the Z axis. The amount of offset depends on the resolution of the artwork and the intended camera move, but as a rule of thumb I separate background layers by 100 pixels (R= 0, G = -100, B = -200) and foreground layers by 250 pixels (R = -500, G = -750, B = -1000). You can also use the virtual focal length of the camera to adjust the amount of perspective distortion, flattening or exaggerating the separation between planes to suit taste.
Now you can drive your camera around. The closer you get to looking at the artwork square-on, the more alignment between screened layers will increase and the more coherent image will shine through. The shorter your camera focal length, the more separation between layers will be visible on the edges of the area of focus. Swoop, turn, pan or twirl the camera to maximum animated satisfaction.
The subject viewed obliquely (left) and straight-on (right). Click thumbnails to view a larger image. The final step is to add a glaze of volumetric-style visible-ray lighting, which will help gel together the layers and reinforce the feeling of perspective (rather the feeling of cut-up images sliding over one another). I have used Trapcode Shine, but any other light-beams plug-in effect (such as those from Final Effects or Boris) will cut the mustard just as well.
Use short rays of unboosted intensity with a normal transfer mode at 30 - 60% opacity. The key is to keep the rays in perfect alignment with the apparent perspective of our "projected" artwork. This means choosing a central point well-back of the image plane, and trying to keep that point constant relative to the artwork throughout any camera moves; for complex moves, use a null object as a tracking point for the projection source.
The final pseudo-hologrammterisation of the elephant photograph. Click to view a clip of a subtle move (left) or a gross move (right). That's it. Have fun.
[ Other Techniques ] [ MFDH's StoryZoo Studios ]
If you've enjoyed what you've watched here
today, please consider leaving a tip.
