So how does tx-transform work?
tx-transform is a film technique that transposes the time axis (t) and with one of the space axes (x or y). An individual frame of film normally depicts the entire space, but only a moment in time (1/24 second). With tx-transformed films, it is the opposite: each frame shows the entire duration, but only a tiny portion of the space - if one cuts alongside the horizontal space axis, for example, the left portion of the picture turns into the "the before", the right one into "the after". How is this achieved? Talking in technical terms, every digital frame of a sequence is split into all of its vertical lines – in the following examples a PAL video frame in 720x576. The first image of a transformed sequence is a combination of all the first (e.g. left) vertical lines of all images in a sequence (or group of images). The second image of a transformed sequence is the combination of all the second lines of all input images and so forth.
Fig.1: Scheme of the column rearrangement between input and output images as it is used in the »tx-transform« process.
In order for the transformed images to have the same resolution and aspect ratio as the input source, the number of input images should be the same as the number of their vertical lines (= 720). If the number of input images is lower, the output image becomes narrower – if the number is higher, it gets wider. Technically frame rate at which you shoot makes no difference as long as you have this number of images. The conversion of time to frames is static and fairly easy to calculate: 720 / frame rate = time in seconds. For example: @ 25 f/sec you would need 28.8 sec to produce 720 frames – if the speed is higher you will need less, if it’s slower more time. When shooting with film, my suggestion is to shoot at a higher frame rate – this helps the model to move in a more natural way. These are the basic technical outlines of tx-transformations.
Fig.2: Sequential images organised in a “block of information” – on the left side in a “normal” film, on the right in a tx-transformed sequence.
More philosophically, »tx-transform« is a way of cutting through the so called "information block" along the space axis instead of along the time axis. Initially, it may seem highly improbable that these "space cuts" would produce discernible images, never mind perceptible sequences of motion. But this turns out not to be the case. The consequence of these "space cuts" through the "information block" is a series of astounding visual effects: houses start to move, heads grow out of themselves, moving trains become shorter and shorter with increasing speed, and much more, but the objects remain strangely recognisable despite the “distortions”.
In contrast to conventional films, the motion of the camera and/or object across the frame is of upmost importance in tx-transformations. In order to be able to use material captured on film for the production of tx-transformations, one must comply precisely with a number of different parameters and a variety of criteria relatings to the relative motion of the camera and the object. For example, the media standard process of omitting an unsuitable segment (cutting it out) becomes impossible since a single missing image in the raw footage would have consequences for the effect of the entire sequence. Nevertheless, the result of a tx-transformation can appear to be completely abstract or completely realistic, depending on how the shot is set up.