Considerations for Capturing and Logging

The recent FCC notice (MB Docket No. 04-232) has challenged broadcasters to potentially come up with a plan to retain recordings of their programming for some yet-to-be-determined period of time, "such as 60 to 90 days."

The recent FCC notice (MB Docket No. 04-232) has challenged broadcasters to potentially come up with a plan to retain recordings of their programming for some yet-to-be-determined period of time, "such as 60 to 90 days."

Although the July 8, 2004, Notice of Proposed Rulemaking (NPRM) has and will certainly be challenged in various venues, ultimately, if adopted, broadcasters may need to archive the programming outside the 10 p.m.-to-6 a.m. period.


TV stations have, for decades, recorded their primary off-air program stream to such media as VHS (in extended-play mode), to logging VCRs (like those employed in surveillance systems), to slow-scan frame recorders, and more recently, to logging disk-drive recorders, including home TiVo sets.

In most instances, these off-air recordings were dated and time-stamped with burned in timecode and logged for the purpose of creating an affidavit to verify a commercial actually ran in the time period for which it was contracted.

News departments have often "TiVoed" other local stations for producer or news director review during a later department meeting. Third-party services have also recorded local channels, especially during ratings periods when the cost of advertising was high and missing a spot presented an even greater complication for competing sponsors in major markets.

In most cases, the content was not kept for an extended period (usually less than a month), and the quality of the recording was sufficient only for validation, often without audio when a slow-scan frame-only recorder was employed.

The recordings were seldom, if ever, intended as potential evidence or protection against a possible indecency proceeding.

So for the purposes of identifying what alternatives in recording/logging technologies might be available, should the FCC's NPRM hold water, let's take a look at some potential schemes that employ non-tape, e.g., disk-based systems for the practical recording of content for long-term archive.


One of the first decisions that should be made is to determine the quality of the image recording. Postage stamp-sized QSIF files in MPEG-1 might be sufficient for strictly validation purposes, in which case a PC-workstation with a low-quality tuner and capture card, plus a gigantic hard drive, might be sufficient for several months of recordings.

Stations that employ a Web-enabled encoding system for streaming programs to the Internet might easily multipurpose them to a hard disk, making it possible to use simple Windows Media or other PC-based capture software for the same purpose.

A more sophisticated approach would be to employ one of the MPEG-based capture and logging systems. These systems, principally designed for media asset management, provide the means for capture-and-logging at higher resolution, plus they produce a low-resolution proxy of program content, often in an unattended hands-off operation.

The applications included with these loggers feature metadata structures that link the traffic and/or automation systems for synchronization. Search engines permit rapid location of the stored program content via a Web-based retrieval of proxies. Recordings can be dual-encoded with the low-resolution proxy, typically MPEG-4, and pushed to a DVD-RAM or other linear data tape archive, and stored for an indefinite period of time.


The advantages of a logging system versus the VHS or slow-scan logger are its ability to capture both the audio and video, integrating metadata from outside sources and enabling a network architecture so that others (beyond the FCC and a station's legal council) can make use of the content on a desktop workstation in a nonlinear retrieval function.

The logger can further act as an aircheck service, permitting segments from the higher-resolution recording to be recovered for duplication or other purposes.

Loggers can be purchased with different levels of components. Some packages might include the proxy recording station as a dedicated MPEG-4 encoding station that performs a real-time recording of the off-air signal.

Other options include an MPEG-2-to-MPEG-4 converter that transcodes MPEG recordings from another video server, say at the server's 12 Mbps, MPEG-2 4:2:0 recording format--to the MPEG-4 domain. The transcoding process essentially reduces the 46.5 MB file for MPEG-2 4:2:0 at 12 Mbps to as low as a 0.7 MB file (MPEG-4 at 150 kbps for 30 seconds).

Coupling an archive manager to the proxy system, on the other hand, can run the gamut in terms of costs and features to deploy.

If the broadcaster already employs an archive manager and library system for server backup, adding handles into the system for managing proxies might not be a terribly complicated activity, given the product and the fact that some peripherals may already be in place. If developing a proxy archive from scratch, finding an integrated solution from a single vendor might prove more cost-effective in the long run.


Dealing with the instances of a single program stream, such as the NTSC analog channel, is rather straightforward in today's technologies.

The costs of the encoders, hardware, disk storage and even physical media (optical or linear tape) are relatively mild, and if you employ an IT specialist already, a system might even be built out of standard components available at most computer discount warehouse companies.

However, the catch in saving your "programming for matters of decency" comes when you switch to digital channels, whether a single standard-definition stream or a high-def/multichannel environment. Now the solutions get a lot more complicated.


A straightforward approach to capturing each of the program channels would be to dedicate a single receiver (set-top box) to each designated stream (e.g., DTV-1, DTV-2, etc.) Direct the analog outputs of each STB to a dedicated MPEG-4 streaming encoder and capture the signals to a set of conventional PC-workstation-like/server-based hard drives.

If the STB has date and time-of-day burn-in capability on the analog output, turn it on and leave it on. The devil is obviously in the details, but the alternatives bring greater challenges.

The more challenging approach would be to capture the aggregate 8-VSB signal off-air and record the full transport stream (most likely as ASI) into a video server capable of recording the full 19.38 Mbps ASI signal.

Although this will consume a whopping 20 Mbps of disk space, it will honestly and truly capture all the PSIP data, all program streams and any other ancillary data (including broadcast datastreams) for further use or analysis. This method keeps the station honest and serves as a tool for technical diagnostics in the future.


To further address the absurd--that is to keep the programming for between 60 and 90 days (as is suggested in the FCC's NPRM)--stations would need to stream 9 GB/hr to a hard drive continuously, which in round numbers means between 8.6 and 13 TB of disk space to meet the objective time periods.

There is very little market for the logging-and-search platform that would allow the content to be returned in anything other than a transport stream, and certainly no immediate mechanism for browsing or generating a proxy of all the content for easy access or verification.

The bottom line is that the capture of the digital broadcast signal for the practical retrieval of program content and verification will most likely remain in the analog domain for at least the duration of analog television broadcasting.

But there is hope maybe the decency requirements suggested by the FCC will become self-monitoring and the need for mass storage of ASI transport streams will fall on deaf ears.