DVB 3D-TV is a new standard that partially came out at the end of 2010 which included techniques and procedures to send a three-dimensional video signal through actual DVB transmission standards (Cable, Terrestrial or Satellite). Currently there is a commercial requirement text for 3D TV broadcasters and Set-top box manufacturers, but no technical information is in there.
Nowadays 3D television technology is already in its first steps regarding its standardization, now the major 3D market is in theaters and Blu-ray Disc players with stereoscopic systems, but in the near future it will be extended to diffusion, and later Free viewpoint television will come into our homes, which means the need of new coding and transmission standards.
The implantation of first generation of DVB 3D-TV will be staggered:
Matrix of signal formats for 3DTV:
|Compatibility level||1st generation 3DTV||2nd generation 3DTV||3rd generation 3DTV|
HD service compatible (CSC)
|2D HD + MVC (L,R formed by matrixing: depth info)||2D HD + MVC (Depth, occlusion and transparency data)|
HD Frame compatible compatible (FCC)
|Frame compatible + MPEG resolution extension (ex. SVC)|
Conventional HD Frame compatible (CFC)
|L and R in same HD frame|
Conventional HD display compatible (CDC)
Regarding how a signal once it's decoded is sent to the display, current stereoscopic systems use a frame-sequential 3D signal. Left and right frames are sent alternately to the display and by diverse systems like shuttered glasses or polarized glasses are then shown to each eye. This involves that the real frame frequency halves the video frame frequency.
In Phase 1 system, only frame-sequential 3D is allowed, using frame compatible (CFC) format. This is made by a spatial multiplex that combines the left and right video sequences in one HD stream which is coded with H.264 as a single image. This allows to handle video as normal HD video using typical channels and interfaces like HDMI, which is possible in this 1.4a version. Frame compatible (CFC) model is also compatible with 2D HD mode in the same channel, adding some signalling for switching from 2D to 3D.
There are basically two ways to do spatial multiplex: Side by side and Top and bottom, but additional spatial multiplex formats have been proposed in order to improve picture quality by providing a better balance between the V and H resolution.
Side by side (SbS) format just put the left and right images one next to the other in an HD image. Because of this, a horizontal decimate is required which causes halving of horizontal definition. DVB 3D-TV supports following SbS formats:
Top and Bottom (TaB) format put left and right images one above the other in a HD image. In this case, vertical decimate is required which causes halving of vertical definition. DVB 3D-TV supports following TaB formats:
The main function of signaling for frame compatible 3DTV is to signal the presence of a 3D or 2D video stream. It must be also possible to include in the broadcast signal information about the pixel arrangement used to decimate the master HDTV full samples/line pictures to create the anamorphic version, if 3D is available. It's interesting to signal also for 3D receivers the 3D events that are available, for which 3D availability should appear on EPG. For future service compatible (CSC) 3DTV, signal that a 3D version of a 2D service or event is being simulcast, and vice versa will be needed.
Multiview video coding is a compression standard appended from H.264/MPEG-4 AVC which allows send stereoscopic 3D video without resolution loss due to spatial multiplex, and reducing overhead of sending 2 HD images up to 50% and in a single video stream. It is used in Blu-ray players, but at the moment it's not applicable to broadcast because of the processing time for encoding, which uses motion compensation algorithms. Nevertheless, there have been some experiments with MVC 3D broadcast by Fraunhofer society over 2nd generation DVB (DVB-T2, DVB-C2 and DVB-S2).
Total Free viewpoint television can be reached by capturing multiple views and extracting 2D+depth information from them to create a 3D model of the scene. Currently this system is being investigated, but the coding complexity and great bandwidth requirements make current broadcasting applications using Multiview Video Coding impractical, so a totally new compression scheme and capture techniques need to be investigated.