Skip to main content

Captioning systems for IP video

What is CVAA you might ask? It’s the “Twenty-First Century Communications and Video Accessibility Act of 2010” and some aspects of it take effect in about two weeks. CVAA Title I addresses delivery of captioning and descriptive services over IP. Title II intends to make it easier for people with disabilities to consume content. Two FCC Report and Orders have been issued to mandate compliance with CVAA, complete with timetables.

Issued last August, FCC Report and Order 11-126 addresses Descriptive Video Service (DVS). It can be found here. The most relevant information is on page 38, Appendix B. FCC Report and Order 12-9 was issued in January 2012 and addresses captions on IP video content. The most relevant information is on page 81 in Appendix B if you want to look. Otherwise, read on and get what a broadcast engineer needs to know right here.

New FCC rules
Development of Closed Captioning began in the 1970s. The FCC mandated NTSC Closed Captioning in 1993 for TV sets 13in and larger. NTSC captioning uses EIA-608, the familiar system that employs Line 21 for caption data. ATSC digital broadcasting uses EIA-708. It is transmitted in the picture user data packets and uses the DTVCC Transport Stream. It allocates 960bits/s for two backwards compatible CEA-608 style captions and 8640bits/s for CEA-708 captions. The ATSC A/53 standard defines the encoding scheme.

FCC Report and Order 11-126 reinstates the FCC’s earlier DVS rules that were overturned in court in 2002. In part, it describes guidelines for the insertion of audio narrated elements describing visual elements during natural dialog pauses. The content isn’t affected for non-DVS viewers because the original audio track remains intact. Typically, the DVS track or tracks are broadcast on a digital television SAP channel.

The FCC 11-126 Report and Order tells multichannel video programming distributors (MVPDs) with more than 50,000 subscribers that they must caption the Top 5 non-broadcast cable networks and pass through DVS on the Top 4 broadcast networks. These rules take effect in about two weeks, on July 1, 2012. Compliance is required by October 2012. To be in compliance, a station must air a minimum of 50 hours of prime-time or children’s programming with DVS each quarter.

The FCC 12-9 Report and Order regulates captioning and DVS on IP delivered content. It includes video over any IP network and is specifically not limited to the Internet. Content delivered over IP that has not been broadcast isn’t subject to this regulation. It extends the intent of closed captioning to new devices and services, such as second and third screens, and is focused on full-length programs.

The FCC also redefined MVPDs and broadcasters with regard to content delivery over IP. We’re all now Video Program Distributor/Providers (VPD/VPP) because we license content from the Video Program Owners (VPOs). The VPO owns and licenses content to VPDs for distribution. The VPD/VPP then makes that licensed content available to consumers using IP technology. The FCC’s definition of IP delivery technology is broad and includes mobile video, iPads, OTT, TV sets with Internet, and DVRs, online or offline.

Notably, professional and commercial video equipment is not included in any of the mandates, including autologging systems, display-only devices or devices not technically feasible to deliver closed captioning. However, some broadcast autologging systems do have the ability to capture multiple channels of audio and search captions by keywords, all of which is useful for all sorts of things, but not a legal requirement.

The TT solution
EIA-608 and EIA-708 are no longer valid because they are broadcast-specific and part of a delivery standard that depends on the display device to do the decoding. Because IP video has so many standards and formats, the FCC decided the best way to handle captioning and DVS is for VPOs to produce content with timed text formatted captions known as SMPTE-TT, before it is delivered to VPDs, leaving the decoding to software in the player. The SMPTE-TT standard is also known as SMPTE ST 2052-1:2010.

SMPTE-TT is a profile of the W3C Time Text Markup Language. It handles CEA-608, CEA-708, DVB subtitles and WST, and preserves many of the critical elements of CEA-708, such as position, transparency, font, color and timing. In the Preserve format, SMPTE-TT preserves the look and feel of the original captioning format rendering exactly as it was graphically rendered within the original program. In the Enhanced format, captions are normalized to look the same from any content played on a particular device. More detailed information about SMPTE-TT is available at

Compliance requirements
Only full-length content is regulated. Short content clips are not regulated. The Report and Order only regulates content delivered by a VPD, either handed off to the VPD by a VPO or created locally by the VPD. This makes the use of SMPTE-TT virtually mandatory across platforms as programming is produced or repurposed for re-airing and/or the Internet.

Several deadlines loom for VPOs and VPDs. Pre-recorded content over 24 hours old, which has already been broadcast and not edited for the Internet, must be compliant in two weeks, by July 2012. Content that is live or near-live (less than 24 hours old) will be required to provide captioning over the Internet by January 2013. Content specifically edited for the Internet must be captioning-compliant by July 2013.

VPOs must determine which content is subject to regulation and produce SMPTE-TT for regulated content, in the same or better quality than the captioning that was originally broadcast. VPOs must keep a record of all content subject to regulation and certify that information by document to the VPD on a program-by-program basis. The document must state that the program is in compliance or thoroughly explain the reason for non-compliance. Archive content is also subject to the new regulations once it is rebroadcast on TV.

VPDs must maintain the certification documents from VPOs for one year. VPDs must pass through SMPTE-TT when it is provided by the VPO, and they must ensure devices can render the SMPTE-TT data. By passing through SMPTE-TT and maintaining VPO content certifications, broadcasters can find Safe Harbor from trouble. If there is a complaint or inquiry about a particular program, the broadcaster or VPD can produce documentation from the VPO confirming compliance. Consumers have 30 days to complain, and VPDs and VPOs have 30 days to respond. Certifications from VPO will determine where the complaint goes next. Complaint processing begins in October 2013.

The author would like to thank Andrew Sachs, product management VP at Volicon, for the information he provided for this tutorial.