Frame packing is the key technology that is being adopted in the first generation of 3DTV broadcasting services. Squeezing the Left and Right images in a single High Definition frame allows the service provider to reuse part of the existing production infrastructure and the whole of the existing distribution infrastructure. In most cases, theset top boxes already deployed at the user’s premises for HD services can also be reused. A new frame packing arrangement, called Tile Format, has been proposed. This new frame packing format is designed to overcome the disadvantages of the known formats, providing a satisfactory quality for both 3D and 2D decoding and display: in the first case, the Tile Format is displayed as a couple of 720pvideos, providing the Right and Left views, in the second case, it is possible to let a 2D decoder selectively display only the area of the decoded frame containing the Left view. In this document the technical details for implementing a production chain for a Tile Format service are provided, outlining the format specifications and the technical implications of the setup.
Technicaldescription of the Tile Format
The Tile Format is a frame compatible solution that allows the composition of two videos of a stereoscopic pair, images for Left and Right eye, (L and R hereinafter) in a container video. The two videos representing the right and left views have width of x pixels and height of y pixels, the only limitation being that x and y shall be even. The frame composition isillustrated in the following picture (Figure 1), where the image L, representing the left view, is unchanged and the image R, representing the right view, is divided in two vertically. The right section of the image is further divided in two horizontal slices.
Figure 1 - Frame compatible Tile Format
The resulting image will have the following horizontal and vertical sizes (as expressed in terms ofthe component image sizes): X = x*(1 + ½) and Y = y*(1 + ½) The decoder receiving the image in the Tile Format will recover the left image by cropping the region [(0,0), (X*2/3-1, Y*2/3-1)] and the right image by stitching the three regions obtained by cropping the following regions: R1 R2 R3 [(X*2/3, 0), (X-1,Y*2/3-1)] [(0, Y*2/3), (X/3-1, Y-1)] [(X/3, Y*2/3), (X*2/3-1, Y-1)]
The most immediate example use case of Tile Format is fitting two 720p pictures (the left and right view of a stereoscopic video, respectively) within a 1080p picture. As graphically depicted in Figure 2: the left view occupies the top left corner of the 1080 picture, whereas the right view is tiled into three regions that are arranged according to the strategy shown in Figure2.
Figure 2 - Encapsulation of 720p videos in a 1080p container
The application of the Tile Format to this HD example has a few relevant features: when a native 720p content production is used, the left and right views of a stereoscopic video can be placed within the container 1080p without any subsampling. Besides, one of the two views, left unchanged, can be used as a 2D video component ifa suitable signaling to a 2D decoder is provided. In the trials currently active in northern and central Italy over DVB-T and with continental coverage via satellite, two different resolution are being broadcast: on the former 1080p 25fps is used to preserve compatibility with existing decoders, on the latter 1080p 50fps is used to achieve the smoothest video experience possible.
As already stated, the Frame Compatible Formats allow the distribution of 3D video contents over existing HD infrastructure. Since the 2D compatible video is in the 720p format, 2D and 3D contents can be seamlessly mixed over the same program, serving at the same time 2D and 3D users. The bitrate available for video encoding over the terrestrial channels is 8 Mbit/s, while on the...