Prix
The Prix Ars Electronica Archive is a collection enabling search and viewing of all the submissions since 1987. The award-winning projects are documented with catalogue texts and audio-visual media. All other submissions can be searched by title/artist and displayed with year, category in list form. Please cite the credits (artwork name, artist and photographer) and only use the materials if your article is related to Ars Electronica.
Shadowland
Kazuhiro Goshima
This is a 3D (stereoscopic) film shot with one DSLR camera. Only the shadows become 3D, without any digital special effects.
“Shadows” are cast on the streets by the headlights of cars driving through the city. These are another substance in the streets. Every night, the city itself is overwritten like a retina thousands of times, and no one can decipher its memories.
One of the themes of my works is to express a feeling that cannot be there.
The 3D shadow can have peculiar presence simply because it has neither color nor texture.
3D films using one SLR camera
I have made three 3D films since 2010—all of them shot with a traditional 2D camera, for example a Pentax K7, a Sony HDR-CX700V or a Canon EOS Kiss X5(D600).
The essential factor of 3D vision is binocular parallax. I derive parallax from the slight time lag between the movies projected onto the right and left eyes. There are no digital special effects. I show the same movies to each eye, but there is slight time lag (1-5 frames). If the object in the footage moves sideways (or the movement of the object has a horizontal element), the parallax is produced. It is like the “Pulfrich illusion”. In the first 3D film, *Tokyo Three Dimensional Suite*, I moved the camera position horizontally by walking (i.e. stop-motion animation). Because of the delay between the image for left eye from that for right eye, the distance moved becomes the parallax. In the second film, *t2z*, the camera was moved on a handmade motorized dolly. In this film I shot close-up 3D vision, which is difficult with a twin-lens 3D camera.
The mechanism of 3D shadows:
In *Shadowland*, I shot footage using one fixed camera. The moving element as the source of parallax is the car headlights. The moving lights draw the shadows on the wall in the night-time city. The shadows move dynamically as the cars drive down road. The parallax arises from the horizontal movement of the shadows.
The extent of parallax is influenced by many factors. The shadow moves faster according to the speed of the car, if it is close to the wall, if the object is near the car or if the object is far from the wall. The depth of the shadow is inverted by the direction of movement. There are many cars on the road and many objects throw shadows on the wall. This all gives rise to the complex 3D harmony.
Technical details
In general, processing of images for this work is very simple.
I shot the footage with a Canon EOS Kiss X5(D600). The format is 720p/50fps, because of slow motion and to reduce AC flicker. Although it seems bright to the eye, in a large town at night the light is insufficient to shoot a movie. The limits of my equipment were F1.4-1.8 stop, 1/50 sec. and ISO1600. I searched for a place without unnecessary light. Car headlights are designed to illuminate the ground, they seldom shine upwards. So I looked for a wall on a slight uphill slope. It was difficult to find an ideal place, and it took more time than expected. The footage lasts 20 hours or more.
The converted movies run at 24fps (half speed). After noise reduction, I adjusted the soundtrack. The shadow contrast is sometimes very low and sometimes very sharp. If the contrast is too low, it is hard to feel a 3D illusion. There are also elements apart from a shadow in a frame where I carefully emphasize contrast.
The tone-adjusted movies are copied to the right and left eyes as a parallel 3D format. After this, I can process and edit in stereoscopic view, sending the 3D video preview of AfterEffects and FinalCutPro7 to Sony 3DTV.
In ex-3D work, I adjusted the depth of field to a screen by fine tuning the horizontal position of right and left images. An image looks large if it is far away from a screen. But in *Shadowland* it is important that elements apart from shadow are visible in 2D, because the 3D illusion of shadows is very sensitive. It will easily break down if other 3D elements are shown on screen. It is important that only the shadow is a solid.
Finally, the 3D shadow is completed by attaching a brief time lag to images on either side.
Finally, I adjust the time lag between images on either side to optimum 3D viewing and edit the movie timeline at the same time.
Links: http://www.goshiman.com/hp/03ttls/19_shadowland_e.html, http://youtu.be/49C9PFA1tXw
“Shadows” are cast on the streets by the headlights of cars driving through the city. These are another substance in the streets. Every night, the city itself is overwritten like a retina thousands of times, and no one can decipher its memories.
One of the themes of my works is to express a feeling that cannot be there.
The 3D shadow can have peculiar presence simply because it has neither color nor texture.
3D films using one SLR camera
I have made three 3D films since 2010—all of them shot with a traditional 2D camera, for example a Pentax K7, a Sony HDR-CX700V or a Canon EOS Kiss X5(D600).
The essential factor of 3D vision is binocular parallax. I derive parallax from the slight time lag between the movies projected onto the right and left eyes. There are no digital special effects. I show the same movies to each eye, but there is slight time lag (1-5 frames). If the object in the footage moves sideways (or the movement of the object has a horizontal element), the parallax is produced. It is like the “Pulfrich illusion”. In the first 3D film, *Tokyo Three Dimensional Suite*, I moved the camera position horizontally by walking (i.e. stop-motion animation). Because of the delay between the image for left eye from that for right eye, the distance moved becomes the parallax. In the second film, *t2z*, the camera was moved on a handmade motorized dolly. In this film I shot close-up 3D vision, which is difficult with a twin-lens 3D camera.
The mechanism of 3D shadows:
In *Shadowland*, I shot footage using one fixed camera. The moving element as the source of parallax is the car headlights. The moving lights draw the shadows on the wall in the night-time city. The shadows move dynamically as the cars drive down road. The parallax arises from the horizontal movement of the shadows.
The extent of parallax is influenced by many factors. The shadow moves faster according to the speed of the car, if it is close to the wall, if the object is near the car or if the object is far from the wall. The depth of the shadow is inverted by the direction of movement. There are many cars on the road and many objects throw shadows on the wall. This all gives rise to the complex 3D harmony.
Technical details
In general, processing of images for this work is very simple.
I shot the footage with a Canon EOS Kiss X5(D600). The format is 720p/50fps, because of slow motion and to reduce AC flicker. Although it seems bright to the eye, in a large town at night the light is insufficient to shoot a movie. The limits of my equipment were F1.4-1.8 stop, 1/50 sec. and ISO1600. I searched for a place without unnecessary light. Car headlights are designed to illuminate the ground, they seldom shine upwards. So I looked for a wall on a slight uphill slope. It was difficult to find an ideal place, and it took more time than expected. The footage lasts 20 hours or more.
The converted movies run at 24fps (half speed). After noise reduction, I adjusted the soundtrack. The shadow contrast is sometimes very low and sometimes very sharp. If the contrast is too low, it is hard to feel a 3D illusion. There are also elements apart from a shadow in a frame where I carefully emphasize contrast.
The tone-adjusted movies are copied to the right and left eyes as a parallel 3D format. After this, I can process and edit in stereoscopic view, sending the 3D video preview of AfterEffects and FinalCutPro7 to Sony 3DTV.
In ex-3D work, I adjusted the depth of field to a screen by fine tuning the horizontal position of right and left images. An image looks large if it is far away from a screen. But in *Shadowland* it is important that elements apart from shadow are visible in 2D, because the 3D illusion of shadows is very sensitive. It will easily break down if other 3D elements are shown on screen. It is important that only the shadow is a solid.
Finally, the 3D shadow is completed by attaching a brief time lag to images on either side.
Finally, I adjust the time lag between images on either side to optimum 3D viewing and edit the movie timeline at the same time.
Links: http://www.goshiman.com/hp/03ttls/19_shadowland_e.html, http://youtu.be/49C9PFA1tXw
Special thanks to: Sachiko Kawakita for her music selection and piano performance, my wife Yumi for her explicit advices of value, and all friends for their attentive cooperation.
Kazuhiro Goshima (JP), a visual creator, produces videos and multimedia content. His 3DCG work, *Fade into White #2* ('00), won the grand prix at the 2001 Image Forum Festival. *Fade into White #3* ('01) received an award at the Annecy International Animation Film Festival and also won the excellence award at the Japan Media Arts Festival. A number of his visual works have also been highly regarded at film festivals abroad.