Restoration
Taking Media Assets Into the 3G Realm: Automated Restoration Maximizes Quality, Efficiency of Upconversion
Paola Hobson, Product Manager for Conversion and Restoration, Snell
Television viewers today are growing more discriminating and more demanding, as they now have ready access to higher-quality content via 1080p HDTVs and a quickly expanding array of Blu-ray releases. The industry’s widespread adoption of 3G production infrastructures and ancillary systems will enable a further increase in the volume of 1080p content available to the consumer. Unfortunately for broadcasters, content originated in SD – and even in some HD formats – will pale in comparison, particularly on today’s large-screen home entertainment systems, to the highest quality progressive formats.
The shift to 3G will be inevitable for those broadcasters planning to produce and deliver 1080p content. Like the shift to HD, it is a transition that typically will require a controlled migration over time, as permitted by budgetary and technical constraints, among other factors. As broadcasters take on the shift to 3G production, they can take a large first leap by upconverting existing assets to 1080p50 or 1080p60 for 3G transmission.
Among the greatest challenges broadcasters face in upconverting content to 1080p for 3G services is that of preserving or improving picture quality sufficiently for display on today’s larger-screen TVs. Assets that have been transferred from film to HD digital video often retain defects that may have been present in the original or introduced during the transfer. In either case, defects ranging from dirt and dust to scratches and flicker all cause even greater visual issues if they are not addressed prior to upconversion. Without a clean digital master, the broadcaster must rely on image- processing technology to eliminate defects in media and ensure that the finished product boasts high quality and, in turn, high value.
Cleaning With Computing Power
Real-time hardware restoration systems give broadcasters the tools they need to remove picture defects and clean up assets without introducing new artifacts. This is critical because artifacts and noise cause downstream compression, whether for transmission or distribution on Blu- ray, to become very inefficient, wasting bandwidth coding on unwanted elements of each picture. With the right technology, however, even damaged assets can be transformed into content that can be monetized through 3G services.
If broadcasters employ a real-time hardware system at the start of their content restoration efforts, they can remove 85 percent of picture defects in just one pass in the equivalent of the program length. Because such systems typically are semi-automated, restoration staff can perform other tasks during this real-time processing.
Deleting Dust and Dirt
The first step in preparing media for upconversion is to remove dust and dirt. While physical cleaning of original film can help to remove some
34
of this debris, it’s a time-consuming process with the potential to further damage delicate film. When the broadcaster has only a scan or telecine version, rather than the original, then processing is really the only option. Current restoration systems are capable of automating most or all of this task, reducing both the time and cost involved.
Motion compensated processing is a highly effective cleaning and restoration technique, as it allows the user to rebuild a damaged or dirty frame using the frames immediately preceding and following it. By carefully calculating the motion vectors recognized in those frames, a real-time hardware processing system can very quickly address dirt and dust. Phase-correlated motion compensation provides the highly accurate motion vectors required to ensure, particularly in automated processing, that the system can differentiate between defects and objects in motion. Without this capacity, the system may mistakenly “repair” undamaged areas of the picture.
Repairing Tears, Splice Errors and Scratches
Once the preliminary step of removing dust and dirt has been completed, the broadcaster or other media owner can address more substantial damage including tears, splice errors and scratches. This functionality, combined with a fast dirt and dust cleanup, yields a much cleaner piece of media that will compress efficiently during conversion and provide viewers with a crisper and sharper picture.
Unlike manual restoration methods and even earlier software-based solutions, both of which required a significant amount of time and user intervention on a frame-by-frame basis, today’s restoration systems incorporate algorithms that can automatically, in real time, detect and differentiate defects, such as scratches, from picture detail. Next- generation real-time hardware restoration tools can even lock onto the phase of the static picture dropouts caused by scratched 2-inch tape, replacing the lost content using a complex median filter.
Configurable sensitivity parameters allow the user to set detection and repair to an appropriate level, typically with the goal of maintaining all original detail. Masking tools allow operators to protect certain susceptible areas, such as those with very fine stripes, from automated processing. Repair also can be applied less aggressively when specific defects, such as the scratches on old news recordings or other dated content, add character or enhance the overall impact of the picture.
When the defects within individual frames are too severe for interpolation and repair through motion-compensated processing and spatial/temporal filters, the broadcaster can turn to dedicated software tools that allow users to paint into and fix specific areas of damage.
AUTOMATION FOCUS
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60 |
Page 61 |
Page 62 |
Page 63 |
Page 64 |
Page 65 |
Page 66 |
Page 67 |
Page 68 |
Page 69 |
Page 70 |
Page 71 |
Page 72 |
Page 73 |
Page 74 |
Page 75 |
Page 76 |
Page 77 |
Page 78 |
Page 79 |
Page 80 |
Page 81 |
Page 82 |
Page 83 |
Page 84 |
Page 85 |
Page 86 |
Page 87 |
Page 88 |
Page 89 |
Page 90 |
Page 91 |
Page 92 |
Page 93 |
Page 94 |
Page 95 |
Page 96 |
Page 97 |
Page 98 |
Page 99