The Future of Television and HDTV - Featured Article by Digital Trends

The Future of Television and HDTV - Featured Article by Digital Trends

Posted using ShareThis

A nice little article which, whilst not strictly about graphics, does look WAY into the future at what might be technology wise for displays which will ultimately affect graphics too. Couldn't resist the Star Wars reference , sorry.

Gametree TV Gaming Model

There are a few ways to deliver games to the set top box or iDTV. The traditional model is that a company provides the games and servers themselves. The games are developed in-house and remain casual in nature. All intellectual property is owned by the company delivering the game service. Operators can either host the game portal themsleves or pay for them to be hosted on a server.Digiquest is an example of this type of model. Typically their games are simple, casual games that target many platforms.

Another way to deliver games is by streaming video. Here the games, usually full Console or PC titles are actually run remotely on a server or gaming console and video of the result sent to the user. This allows any games to be run at all, but has considerable issues of lag and bandwidth to overcome. In addition the server back end could be expensive to setup and maintain. Operators would not run this server service themselves I guess. One company providing such a service is Playcast which I have previously blogged about here.

Free, in the box games exist of course, but they are of little interest commercially speaking.



Now TransGaming Inc are attacking the TV gaming world. Known for their portability layer for the Mac platform, Transgaming have established a gaming service called Gametree. Here the company concentrates on servers, porting apis and crucially apis to support various business models for delivering the games including advertising. The idea is that certain PC games will run on next generation set top boxes. Transgaming, rather than develop the games themselves, enables the porting and integration of existing games onto the platforms and their Gametree server and business model.

This seems like a good idea. So much so, Intel have invested. This is logical as Intel perhaps make the porting easiest with their Canmore and in the future Sodaville devices. Its not clear if the relationship is exclusive.

There is a video available from the Gametree site. Its highly marketing oriented, but it does push the ability of games developers to integrate apis that allow advertising and various purchasing models for the games. Something important until we understand well Gaming on Demand. Also, Transgaming plans to integrate gaming controllers into the platform. An absolute necessity and again reaffirms my view that next generation set top boxes will no longer use the traditional RCU.

The biggest problem of this method of delivery is that the game must run on the set top box which of course is not a PC. This means only a limited range of PC games are applicable. Thewebsite has this to say:

While the graphics performance isn't on the same level as high-end triple-A gaming, it is not intended to be. It is truly amazing the number of titles that are possible on GameTree.tv, including native Linux games and Flash-based games.

Nontheless the problem of which games to deliver is shifted from Transgaming to the games companies, where it should be. The server and supporting APIs are provided by Transgaming, where they should be. Business models are flexible, as they should be. It all sounds rather good.

Intel CE4100 Atom Gives Double Graphics Clock Rate

A few days ago Intel officially announced its new CE4100 device. Its hard at this stage to penetrate the buzz words like "connected", "BluRay", "FLASH 10" etc. and no date for shipments has been announced but one number given is the clock rate of the graphics. This has been doubled.

Camel is already the smoothest graphics of any device available today and the CE4100 ("Sodaville") seems set to keep Intel ahead of the pack in 2010 at least. Doubling the clock rate may not double the graphics performance if, for example, the application is limited by memory access, no improvement will be seen, nonetheless I can't wait to see this in action.

Sony XMB

The set top box is an odd device. Its not a mobile device, nor does it have the power of a PC. When designing a new interface for it, its not always clear which path it could follow in terms of copying technology and learning from those two worlds. One additional possible source of inspiration might be consoles. The Sony XMB (cross media bar) is a UI design (seen above) that is meant to work on a range of devices. It works on both consoles and TVs currently.

The XMB is a mainly 2D interface, a cross of icons with associated text. The user moves both horizontally and vertically and can see at all times, the complete menu path taken to reach the current choice. Text is only displayed where it could affect the users decision making process (ie its contextual) keeping the screen free of clutter. The focus moves but slowly and never far - rather icons scroll underneath the users focus. Unavailable options are faded out. Leaf menus may be displayed in almost any fashion and differently on different devices, the XMB is mainly a way to navigate the menu structure. The mainly 2D approach does not demand advanced 3D hardware. Users note that it elegant and fast. Here is an example on a TV set:



There are many great subtle features such as the way the bars cross each other and icons skip a space to appear on the other side of the bar and yet the user hardly notices. The animation itself with rapid acceleration and deceleration feels almost alive when navigating. Many of the subtle features work only because the interface renders at 50 or 60 frames per second. Such animations cant work at 10-20fps. Rendering speed matters.

Yet, this generic, works-anywhere approach suffers from lowest common denominator problems. Even many Sony Playstation fans complained that the Xbox had a more media oriented approach that simply looked better. Add to that the fact the icons were very small and barely visible on anything but a large screen TV and SONY had problems.

They responded and through firmware updates, upgraded the initial interface substantially over time. The new version 3.0 was released recently:



So SONY listen and adapt. User interfaces are updated after delivery to the users. Media richness is an important feature. 3D per se, is not, it would seem. Performance allows for subtle animations that brings the interface to life.

Seeing (Android) is not believing

At IBC 2009 Kaon were displaying an "Android set top box" on their stand. It was playing HD video with Android graphics overlaid. Apart from one sign it was just one of dozens of set top boxes Kaon were displaying on their impressive stand. This was rather surprising as you would think an Android set top would be big news.

Two other things were quite strange. The graphics seemed fuzzy, not clear at all, and the set top box itself was quite large: I had expected an Android set top box to be quite small, Android coming from the mobile space.

It occurred that there was a good explanation. Although encased in production set top box housing the "set top box" was far from production inside, consisting of two hardware boards. The first was a normal set top box, running the video with a central CPU, and the second was running the Android, on a different central CPU. My guess is that this was effectively a mobile phone HDK. The resolution of output of the Android part was 800x480 (odd, but thats what I was told) and this was then overlaid on the true HD output of the set top box.

Patent: Remote touch screen user interface

Thomson hold a user interface patent described thus:

A method for control comprises a set top box receiving coordinates from a touch sensing screen. The coordinates are interpreted for controlling the set top box, and in accordance with the interpreted coordinates an action is performed. A further method for control comprises a set top box receiving a signal representative of displacement. A control function is determined from the displacement representative signal and the control function is activated. In accordance with the control function a signal is formed for communication.

Its a 2D/3D input system for a set top box from something like a touch sensitive RCU, or more likely a media device. The trick to the patent seems to be that the video from the set top box is used as the rendering of the user interface for the input device. That is, the touch sensitive input device does not render a user interface itself but takes a video stream and displays it. The video stream from the set top box already has the user interface rendered into it. The user then uses normal touch or stylus input and these co-ordinates are transmitted back to the set top box where some smart software converts those co-ordinate inputs into commands for the set top box (which may have more traditional form of input).

Here is a diagram:



Nice.

3D User Interface Conference and Competition

The IEEE runs a conference on 3D user interfaces each year since 2006. In 2010 its is being held in Waltham, Massachusetts USA in March. There is also a competition, open on any platform, using any hardware, for novel 3D interfaces. Its early days yet and the web site is incomplete. The conference looks small and intimate but it is linked to the IEEE Conference on VR and so will have a Virtual Reality slant to it no doubt.