Sophisticated solutions preserve the quality of your assets.
Media companies today must work with assets in the file-based domain if they want to take advantage of the latest in IT-based technology. High-quality encoding is one of the key factors in preserving the integrity and value of these assets, which may undergo dozens of encoding processes before they reach the target consumer.
From ingest or encoding in a digital camera during production through to final encoding for broadcast or delivery via VOD, mobile phones, portable media players or Web publishing, numerous processing stages threaten to compromise the quality of the media asset.
If the encoding quality is poor, each encode chips away at the quality of the original material. Downstream versions inherit the flaws of the parent asset, so even small issues upstream can become serious problems downstream. This may result in a financial consequence if the producer or broadcaster feels that quality levels have fallen below an acceptable threshold. Because the value of an asset is tied to its quality, reliable encoding of the best pictures for the target bit rate is critical to any media organization's success.
Formats and workflows
Of the video codecs currently in use, MPEG-2 and DV account for the majority of SD production encoding. HD is a more complicated matter because of its demanding data rate, and different requirements for audio and metadata. While uncompressed HD offers exceptional image quality, the sheer volume of data involved makes it difficult to maintain an efficient workflow in post production.
This problem led to the development of various codecs intended to reduce bandwidth, and therefore storage requirements, while still providing images worthy of being called HD. MPEG-2 (including Sony's XDCAM HD), Panasonic's DVCPRO HD and AVC-Intra, Apple's ProRes 422, VC-3 (including Avid's DNxHD implementation) and JPEG2000 are among the many popular video codecs used for HD work.
Many codecs build on knowledge gained with MPEG-2 development and deal with high-frequency detail, high motion, diagonal detail, gentle gradients and small moving objects in large areas with specific tools to prevent blockiness and other failures that can happen in MPEG-2 systems.
Another downside to working with some newer codecs is that they are considerably more sophisticated than MPEG-2, requiring more hardware and causing software systems to perform more slowly than when working with MPEG-2 content. Furthermore, a lot of interoperability work has gone into MPEG-2 over the past 15 years. Newer solutions may not yet provide as many options or as much interoperability as more mature codecs.
In designing a new facility and choosing a compression format, the content owner or producer must determine the way in which that compression format will be used. A news organization, for example, will often be limited by the capabilities of its newsgathering cameras and the formats supported by its main editing system. As a result, news and sports broadcasters have clear constraints about which format they will choose. More challenging is the design of a production environment, with tape, live feeds and archive material needing to be ingested and converted to the appropriate format.
Each codec has advantages depending on the lifecycle of content within the facility. Is quality for long-term archives most important, or is the goal to get media to air as quickly as possible? Many companies are waiting out the HD compression battle, working with MPEG-2 and using an advanced encoder to transcode into alternative HD formats, as necessary.
If the facility chooses to work with a compression-agnostic wrapper such as MXF, it can more freely build assets and later decide about moving to a newer compression format. Reliance on MXF allows a facility to create transcoded content efficiently and, with the help of metadata, manage media files as simply as a factory manages inventory.
What's more, MXF preserves a history of metadata and media essence, thus providing a future migration path for facilities building a store of MXF masters. As compression formats change, the encoder's MXF support enables the facility to evolve and adapt smoothly to these developments. The Advanced Media Workflow Association is actively working on the use of MXF in automated IT facilities. This work welcomes user input, and companies are encouraged to check the association's Web site (www.amwa.tv) and participate.
The ability to monetize assets through new distribution channels depends on the encoder's ability to support virtually every major format needed for markets as diverse as broadcast, video on demand, portable media players, mobile and the Internet. Encoders that take advantage of a broad spectrum of codecs and use wrappers such as MXF can provide the right content for all these distribution channels while keeping operations relatively simple.
Streamlining workflows
The programming viewers see on TV may have been encoded up to a dozen times on its journey to the home. During production encoding, today's facilities often acquire SD and HD content using digital cameras but frequently end up transcoding to a different format, sometimes unknowingly, as they begin the editorial process. While some facilities edit in the same format used for video capture, more often, the content acquired is converted and transcoded (up-, down- and cross-format converted) to meet the facility's production and editing needs.
Once edited, content may sit in another format on the company's nearline archive, after which it may be encoded again for the broadcast server. Then it's encoded once more when compressed for delivery as part of a cable, satellite or terrestrial broadcast offering. Advanced encoding systems preserve picture quality and simplify the production workflow by providing immediate content access. Figure 1 shows a typical production encoding workflow. Figure 2 illustrates the convenience offered when input, output and conversion work is performed within full automation support. Support for native NLE formats enables fast conversion outside of NLE systems, when required, so that creative staff can get right to work on encoded material.
As content owners and distributors are increasingly called upon to provide premium HD content, advanced encoding takes on a vital role in maintaining the quality of that material. This can be easily understood when realizing that although HD typically offers 6X more pixels than SD, it usually only gets twice the bit rate used today in SD by most people.
A big part of the value proposition for HD rests in its ability to deliver a submersive experience through the quality of picture and sound. So, when addressing HD ingest, the quality of throughput and reliability of ingest are far more critical than in earlier-generation, SD-oriented ingest workflows. While it becomes increasingly difficult to control media as it moves through the marketplace, media companies can use intelligent encoding solutions to make sure that any content they ingest, store, repurpose or distribute meets a high-quality standard.
Mastering and repurposing
In the mastering environment, the integrity of branded assets is paramount. High-end mastering relies on sophisticated picture processing algorithms, and encoders today are chock-full of proven technologies developed for professional broadcast applications. These technologies provide superior-looking pictures in a wide range of professional formats, whether HD masters or mobile media encoded at low bit rates. Content owners must seek out efficient and reliable encoding systems that will allow them to achieve the balance of quality and bandwidth right for each target application. In turn, this will create new opportunities for revenue generation while lowering storage, distribution and transmission costs.
Integrated motion estimation technology is the foundation of key processing tasks, including deinterlacing, frame rate conversion, SD-to-HD upconversion and HD-to-HD conversion. These processing tasks require advanced image conversion capabilities to enable the high-quality, artifact-free output necessary to leverage HD and SD masters successfully.
Phase correlation is the most powerful and accurate motion estimation technique today. It provides the highest level of vector reliability and precision by analyzing incoming images and shifting the pixels in each video field so that they line up perfectly, even for complex scenes where there is fast motion, fine graphics, or mixed film and video material. The accurate shift of pixels enables clean deinterlacing of video to progressive with minimal artifacts. It also provides the high-quality motion portrayal necessary for transparent standards conversion and SD/HD conversions.
Phase correlation technology eliminates issues such as ringing, softness and adaptive mode switching artifacts. With still images, the technology achieves maximum resolution and preserves the fine detail in graphics and sharpness of text. In fast-moving action, phase-correlated motion estimation and motion compensation ensure full resolution and well-defined motion. In complex pictures, a variety of components can be handled without compromise, yielding a crisp overall output.
Sophisticated encoding techniques give content owners the power to create high-quality deliverables from their masters. In the creation of content for low-bit rate applications, advanced encoder capabilities such as automated dynamic reframing and background blurring allow for a significant improvement of sports and entertainment content delivered on small screens. Systems with advanced mobile encoding capability can dynamically distinguish the foreground and the background within the video image, treating each separately in order to concentrate the most energy (or bits) on the parts of the image that are most important to the viewer.
By bringing the most important object, character or athlete into crisp detail, while lessening the detail and prominence of objects or scenery in the background, the encoder can provide a compelling viewing experience regardless of the size of the display. Dynamic reframing of video images and the production of tighter, sharper shots helps to retain the emotional impact of the original content when transcoded for low-bandwidth delivery to portable or bit-starved devices.
The bit rate savings resulting from advanced video processing techniques allow content owners to leverage limited bandwidth to provide more downloads to more customers or to stream additional material on their distribution channels. Companies producing higher quality content can improve the look of their product for the same amount of money or bandwidth, thus maintaining costs while targeting premium customer segments. By establishing a high level of automation in encoding material, a facility owner or manager can reduce operational costs, free up staff for creative work, and improve the throughput and time to market or air.
Automating QC in encoding
Among the greatest advantages of advanced encoders is their ability to perform real-time quality control (QC) during the encoding process, whether at ingest, receipt of external file, or following any subsequent conversion or repurposing operation. Issues with video, audio and metadata often appear at the end of the production chain, long after they first emerge. Until recently, the limited options for controlling the quality of file-based content typically involved having an operator review content on a computer monitor or copy the content to tape or server, and then send it to a dedicated QC suite for checking. A more proactive approach — made possible through the latest encoding systems — allows such problems to be tagged much earlier. This enables the user to identify and resolve issues sooner and consequently minimize the time, resources and money wasted on compromised assets.
While a large visual artifact may be easy to flag, identifying smaller problems that are invisible to the human eye, yet critical to downstream encoders, requires more sophisticated evaluation of content. Inspection of content prior to encoding allows not only for identification of problems with the signal, but also for a before-and-after comparison of QC reports to determine if any new artifacts have been introduced.
Automated QC processes in advanced encoders can measure parameters that eyes and ears don't, such as multichannel audio levels, compression performance, bit rates, metadata and content legality. Data stream can be monitored for closed captions and Dolby metadata. Rather than spending time visually scanning content in a QC suite, staff can use a frame-by-frame QC report to inspect only the exception found by the machine QC stage. This allows the operator to quickly deal with potential problems in content. Easy integration with automation and asset management systems permits seamless checking and encoding of content with little need for human intervention.
Once content is compressed and compared with the original signal, the advanced encoder provides operators with tools for finding the best balance between quality and the desired bit rate. In cases in which content is being captured from a satellite or contribution link, there may be just one chance to make the right decision. The encoder can help operators monitor ingest in real time so that their masters are of the quality they expect and require. With the right system in place, content producers and owners can be assured that they will get it right the first time. This confidence is also particularly valuable in mastering from delicate or deteriorating tape-based archives that might last through just one round of digitizing.
Centralized storage
Most encoding products ingest either to proprietary storage or to standard computer disk, but the greatest workflow advantages are gained when the encoder performs well with centralized storage. Proprietary systems often require ingest to a local disk and then transfer of the ingested file to a SAN or NAS. This process doubles the storage requirement and also creates additional traffic on the network.
However, encoding directly to a SAN is not a trivial matter, and it's difficult for many systems to accommodate multiple users performing read and write operations. By implementing an advanced PC-based encoding system optimized for centralized storage, the content owner can realize time and cost savings while ensuring that media is available almost instantly to all operators.
Managing time code
Time code is a critical component of most media operations, serving as the foundation for frame accuracy. Although time code has been standardized in the baseband video and tape-based domain, the same can't be said for the file-based domain. As a result, time code for a clip on a disk might be indicated in a variety of areas — in the actual video (VBI), in the codec information, in the file wrapper (like MXF and QuickTime), in a private stream of metadata within the wrapper or even as a separate XML file.
Due to the industry's lack of standards, different receiving systems will look for time code information in various places. Consequently, it is especially important to work with a professional encoder that records time code accurately, and in a coherent fashion, in all these different areas. This capability greatly improves interoperability among systems, with the assurance that edit lists and cataloging information will still have accurate time code as a reference.
Advanced encoders also provide a solution for discontinuous time code. In the tape-based world, people often collate different scenes or programs on the same videotape, creating discontinuous tape time code. Because most file-based systems on the market can only handle one range of time code per clip, significant issues arise when a whole videotape with discontinuous time codes is ingested as a single clip. This is particularly true of sports and news material, which have the commonality of being shot on cameras — usually locked to a running clock — with on-and-off recording throughout the day.
Content producers can use an advanced encoder to ingest a whole videotape as a single media clip so that they can have a one-to-one relationship of the asset to their existing database of content. The encoder's ability to track both the discontinuous time code and continuous time code supports workflows requiring ingest of multiple programs or segments as a single clip, while maintaining frame accuracy. This operation extracts content without wreaking havoc on a facility's established library system.
Conclusion
The advanced functionality of a professional encoder brings valuable efficiency to the handling of content in the modern media facility. While content producers and owners can choose from several media formats and numerous workflow models, their selection of a robust, highly automated encoding platform provides the reliability and quality needed to meet the requirements of media mastering and delivery in a demanding and quickly evolving marketplace.
Stephane Blondin is director of product management for AmberFin, a Snell & Wilcox company.
