There is considerably more to 3D television success than just engineering techniques – both established and fresh production considerations are vital, too. Philip Stevens talks to a number of experts in taking a closer look at the 3D production learning curve from a director’s viewpoint.
Since Sky Sports first launched its 3D output in August 2010, the broadcaster has covered a variety of events using that technology. These have included football, rugby, the Ryder Cup and US Masters golf, the US Open Tennis, darts and the Haye v Harrison boxing match.
Understandably, much of the attention over the past year has been focused on the engineering achievement of generating 3D output. Beyond that, manufacturers of 3D televisions have been keen to display their latest wares. But what about those charged with the responsibility of producing the programmes? What, if any, new production techniques have to be employed when directing 3D programmes? Can the 3D production facilities be used for 2D output – or are there separate outside broadcasts for each standard? As far as costs are concerned, it would make sense to have just one production unit generating pictures for both formats. But is that practical?
“Usually there are two autonomous OBs,” states Andy Finn, a programme director at Sky Sports. Finn has more than 20 years’ experience of directing sports output and is a regular when it comes to 3D coverage. “This enables the best coverage for both sets of viewers. When it comes to many sports, low camera angles for 3D coverage tend to produce the best results. Of course, each venue has different considerations, such as seat sales and sight lines. When it comes to football matches, the number of 3D cameras we employ is considerably less than conventional coverage. In most cases, we use eight 3ality camera rigs for 3D productions.”
At the moment, because of restrictions with the mirror boxes and the physical size of the lenses on side by side rigs, only 22:1 lenses are available for 3D sports coverage. Doubtless, manufacturers are working to overcome this restriction. Similarly, the weight of rigs is also an issue for both handheld cameras and jimmy jibs.
Finn explains that there is an exception to this two separate OBs configuration when it comes to the coverage of darts.”Here we use a single 3D OB unit, with the right eye output of the camera rigs being used for 2D transmission. This is because space is generally limited at a darts venue, meaning there is really only one good position to capture important action. For this reason, it is space-efficient and cost-effective to treat the event as a single outside broadcast.”
Mention was made of the fact that the right eye output of the 3D rigs is used for conventional coverage. As the image from this source doesn’t pass through the mirror on the camera rig, it is generally sharper than the output from the left eye lens, and therefore better suited for transmission.
Scanners used by Sky Sports generally have JVC 3D monitors for the transmission and preview output, while all sources are viewed as 2D images on Vutrix production monitors. Sony stereography monitors are used for engineering.
It has often been said that directing for 3D requires some different disciplines. For instance, because the brain is unable to assimilate the images as quickly, fast cutting is not appropriate. “I can understand why there might be a problem with, say, a rock concert – but the cutting of pictures for sport is determined by the pace of the event. Admittedly, I think my pace of cutting has changed somewhat for 3D, but for some events – golf, for example – quick shot changes have never been correct.”
The key, Finn believes, is always to be sensitive to the needs of the viewers, and if over-cutting reaches a point where it is hard to watch, then the pace is wrong. He says that hands-on experience of directing this type of coverage will quickly reveal what is best practice.
Similar considerations have to be given to the use of graphics. Two factors – convergence and depth budget — are crucial for the successful use of graphics for 3D programmes.
Convergence involves the ability of both eyes to turn inwards together, enabling them to look at the same point in space. Depth budget is used to describe the maximum amount of depth consistent with acceptable stereoscopic viewing. In reality, this depends on location of the viewer and the size of the display.
“The last thing viewers want is to feel the graphics are on the end of their nose,” declares Finn. “Also, you have to think about the image behind it. For example, during a darts match where you have a player walking towards the camera to retrieve the darts from the board, a badly designed caption could make it appear that the individual is in front of the graphic. It is vital that the convergence issue is considered along with the screen plane.”
He continues: “When it comes to depth budget it is important that the graphic does not appear to have a great deal of depth because our eyes simply cannot tolerate that effect.”
Vizrt UK provides the graphics support for the Sky Sports 3D programmes. Its managing director, Roy Jenkins, says that 3D graphic operators must be good Illusionists. “After all, 3DTV is an optical illusion. The graphics are not really popping out of the screen, they just appear to be. And it is all too easy to ruin the illusion if you don’t follow the rules of how the illusion works.”
Although it is normal for a lower-third 2DTV animation to come in from the left or right, this doesn’t work for 3D because the golden rule for all stereoscopic images is that they must never touch the edges of the screen. “Since your eyes know in which plane the edges of the screen are, if a graphic that appears to be sticking out from the screen is allowed to touch one side, the eyes register the conflicting information and the illusion is lost. 3DTV graphics change their perspective by animating from the top or bottom to the centre or by fading in and not by flying in from outside.”
Jenkins explains that unless full-frame graphics are shown, it is best not to inject too much depth-changing into the scene. Graphics overlaid on video and whizzing about at all 3D depths could conflict with the stereoscopic content of the video itself.
“When it comes to live stereoscopic production, a graphics operator can feed changing depth information to the graphics scenes automatically using GPI signals, according to whichever camera is on-air. This enables a director to know that the graphics will always have the appropriate depth perception for any camera view. These important tweaks adjust the separation distance between each eye depending on whether a graphic occupies the front, main or background layers.”
With an increasing demand for Sky Sports 3D production on the horizon, OB provider Telegenic has just announced its fourth truck equipped for that format is about to be built. “We have been using our T18 and T19 units for Sky’s 3D production,” explains Terry James, Telegenic’s director of operations. “The latest, T21, was built to come online at the end of May.”
Telegenic reports that with several Premiership football games to cover each week, the investment has been justified. “As far as directors are concerned, there is very little technical difference between the trucks. However, we are purchasing Sony P1 cameras for the new unit – and these will likely then be spread on rigs across all the trucks,” reveals James. “Our third truck will house space for 24 cameras, or 12 3D positions.”
He goes on to say that Sony P1 cameras will, among other considerations, make it easier for mirror and side-by-side positions in stadia where space is at a premium. This will provide greater flexibility for directors in their coverage of 3D events.
James then echoes a comment made earlier by Finn. “When it comes to cameras, directors are concerned with the lack of long lenses for coverage of certain sports, most notably football. However, I believe that we are entering an era where all major sports events will be covered in 3D. And that means technological developments will increase.”