South East Asian Video Project

More detail in the 'Deliverables'.Outline 1. Software or virtualdub filter which can take a text 'script' including timing information (probably in the form of a caption or 'karaoke' script, using a standard SSA style format) and encode this into the top lines of a high-definition video picture before it is H.264 encoded. The data must be resilient and redundant enough to be able to withstand the H.264 encoding. However, the quantity of data to be carried is very small.? ? 2. Software running on a cheap PC at the client end, which can read the picture from a cheap HDMI video input card, decode it, and pass it out as close as possible to real-time. Ideally this software must run on an economical operating system such as Linux. 3. We will require programming which can accept the decoded text and store, display it, or use it to control an X10 automated lighting system. ## Deliverables 1) All deliverables will be considered "work made forhire" under U.S. Copyright law. Employer will receive exclusive andcomplete copyrights to all work purchased. (No 3rd party components unless allcopyright ramifications are explained AND AGREED TO by the employer on the siteper the worker's Worker Legal Agreement).<BR> ? We work in Asia ? We have a client who is building up to 1000 low-cost meeting halls as part of a government-sponsored public education program. ? Whilst the hall is multi-purpose, one of its chiefrequirements is as a 'cinema', showing both entertainment and public education films. ? We are only concerned with the 'cinema' aspect of the hall. ? These meeting halls will be distributed around the country -mostly in rural areas. There will sometimes be little infrastructure there. We cannot rely on anything other than a basic electricity supply (which may cut out frequently, requiring a good UPS), and access to a satellite signal. There will be no access to ADSL or GPRS signals. ? High definition H.264 transport streams will be beamed by satellite into these 'cinemas', stored on a small digital set-top box, and played out according to a pre-defined schedule. ? We are making use of mostly quite cheap home products to achieve this. The budget is very low. But we can still maintain a very high standard of sound and audio presentation. ? But we have a challenge. ? The brief requires that the cinemas be entirely unmanned. The whole operation must be automatic, as we will have no way of training or ensuring the skills of the regional operators. They will have little or no technical ability, and its important that the show goes on, regardless of the local operator. ? ? There are four elements to the automation: ? 1. The film must start and stop at the right times. 2. The lights in the hall must dim on and off at the right times. 3. A small computer monitor positioned outside the hall must show: a. Information about how long before the start of the next programme - probably a countdown. b. Text information about upcoming programmes. ? For item 1, this is simple. The satellite service provider has the ability to send control signals to their playout box. So we do not need to worry about this. ? But items 2 and 3 are our responsibility. The cinemas do not have any 'data' link or modem. The signal sent to the cinemas by satellite cannot include a data signal or cueing signals. It contains only a Transport stream which will go to the settop box and be decoded to an HDMI video signal and an optical digital audio signal. ? ? After much thought, we feel that the only way to send the required cues for lighting automation and simple text information will be by embedding it in the audio and video signal in such a way that it will be passed through the whole chain, and can be read at the cinemas either in the video or audio signal. ? ? One option is to use one of the audio tracks of the incoming stream to carry this data information. ? The advantage of this is that the hardware for a small PC to read this audio track and convert it into data would be quite cheap. ? However, our brief is to provide 5.1 surround sound to the halls. So if we used one track for the data signal, we would lose one channel of audio, and this would not be acceptable to fulfil the brief. ? So we have abandoned this concept. ? We then considered the VANC signal portion - often used for closed captioning. ? We are still considering this, but we have conflicting information as to whether an HDMI cable can carry a VANC signal, and we are not sure whether this signal will be passed through the entire satellite chain. ? So ultimately, we feel there is only one absolutely certain way: ? ? The picture will be projected onto a screen which will have a black mask. ? By zooming in very slightly, it would be very easy to ensure that the top two or three lines of the picture are not displayed on the projection screen. ? We could include data in these top lines. Given that they are part of the picture itself, the H.264 encoding, the satellite chain, and the set-top decoder must include these lines. ? But we would need a customised way of doing this: ? 1. Software or virtualdub filter which can take a text 'script' including timing information (probably in the form of a caption or 'karaoke' script, using a standard SSA style format) and encode this into the top lines of the video picture before it is H.264 encoded. Obviously the data must be resilient and redundant enough to be able to withstand the H.264 encoding. However, the quantity of data to be carried is very small. ? 2. Software running on a cheap PC at the client end, which can read the picture from a cheap HDMI video input card, decode it, and pass it out as close as possible to real-time. Ideally this software must run on an economical operating system such as Linux. ? 3. We will require programming which can accept the decoded text and store, display it, or use it to control the X10 automated lighting system. ? ? The ideal programmer must have experience and knowledge of video signals, available SDKs, and programs such as virtualdub. ? Being a public supported project, the budget is very small. ? If responding to this brief, it would help us if you could: a. give some idea of your experience with video signals, coding, compression, or other appropriate skills. b. suggest your choice of SDKs and HDMI input card at the client end. c. give your opinion on the viability of the project. d. proposed cost of the system development. ? ? Deliverables to include source code, and to be finalized before the acceptance of the proposal. ## Platform See above