- Start: 01/01/2010
- End: 31/12/2010
Projectleader: Luk Overmeire
In this project, we investigate and optimize next generation file-based media production environments, hereby focusing on both storage and network infrastructure and production workflows. In addition, 3DTV is studied as a possible next step in television production.
Next generation media infrastructure
In previous research projects dealing with advanced file-based media production, the WARP storage model was designed. This model offers a much higher efficiency than traditional Fibre Channel storage architectures that are being used nowadays for processing Standard Definition (SD) media production. The higher efficiency of the WARP cluster model allows to offer High Definition (HD) media production. Due to the possibility of independent scaling in network and storage capacity and calculating power, the WARP storage model can be deployed in a very flexible way. Consequently, the transition to fit-for-purpose storage systems becomes smaller.
Already since the design of the WARP cluster, the target goal is to build an architecture that is significantly cheaper, without losing redundancy.
A study of the required degree of redundancy is still necessary. The lower cost of this system is not only attractive for purchase; it also results in a less expensive maintenance cost.
A workflow study of the Digital Media Factory (DMF) at VRT showed that too many FTP-transfers need to be started to transfer media content of different storage clusters of the central infrastructure to the right media service. Based on the WARP storage model, a large amount of these transfers via the LAN (IP) network, could be avoided. This would not only result in an increase of performance, but also in an improvement of the reliability and predictability of the transfers.
In a first phase of this project, a selection will be made of the workflows that can be improved by using a WARP architecture. Possible candidates are ingest, transcoding, in-place editing,…
Different transfer methods as well as network architectures will be evaluated and tested.
Automation in a media production environment
Automation in media production environments is essential to realize complex production workflows, where content needs to be transferred between the central infrastructure and different production systems, while undergoing possible transformations. Ideally, automation of production processes is realized based on the concept of a ‘Service Oriented Architecture’ (SOA). In such an approach, media functionalities are defined as media services, that are controlled via automated processes. It is essential to have a good definition and description of the necessary services. On the other hand, it is important to have a good view on the different workflows. By using a Business Process Management System (BPMS), different workflows can be constructed in a visual way. The whole concept relies on interaction with a service stack (SOA) on the one hand and on human participation in the workflow on the other hand.
The goal is to design and implement one (or more) media process(es) for a relevant use case, further building on the knowledge that was acquired during the Geisha project. This implementation will be also evaluated with respect to further automation of processes.
Complex Event Processing (CEP)
By designing processes, workflows are made explicit and optimizations can be done. Underlying dataflows of these workflows, are not inherently explicit due to the complex character of these media specific data operations. By using Complex Event Processing (CEP), we want to gain more insight in these dataflows. Initially, this will be obtained by monitoring the log-files of the relevant systems in the production environment. As such, errors can be detected more quickly, and in a scalable way.
This technology and approach can also be used for reconstruction of dataflows in the production environment. Based on this approach, integral capacity management can be elaborated.
We will also investigate to which extent an event driven approach can be complementary to a process driven approach, and how the issue of service pipelining can be handled in this context.
3DTV
3DTV is expected to be the next revolution in the history of television production. 3D movies in Digital Cinema have become very popular recently (after a short peak in the 50’s). The current generation of 3D is actually “stereoscopic 3D” which is essentially built on two images that are slightly different (one for the left and one for the right eye).
The recent successful investments in 3D in Digital Cinema is driven by the rise of HDTV and the still increasing screen resolution for home displays. Nowadays, the viewer can enjoy high quality images at home and will hence need fewer visits to the cinema, so differentiation in cinemas versus living rooms becomes necessary.
On the other hand, the effect of digitization of production processes such as post production and projection, has also a part in this. Whereas in the 50s, it was difficult to produce good 3D content, nowadays this has become much easier since it is easier to correct or synchronize both images (e.g. color correction in post production, adjustment of the depth by shifting images, synchronization in time in digital projection).
This technological progress can bring about a unprecedented breakthrough of 3D in living rooms: 3DTVs are available on the market, as well as media carrying technology for 3D content such as Blue-Ray (with sufficient storage capacity),…
However, standardization of different technologies is not yet achieved, people are working on it within organizations such as SMPTE, ITU, …. It is of great importance to be well prepared when 3DTV is coming. In this project, further research on the impact of 3DTV on media production workflows will be carried out.