Solving The Challenge Of Online Captioning

If you go to CBS.com and watch an episode of The Big Bang Theory, you’ll find on the video player a button right between the volume and screen-size controls that turns on the closed captioning for the hearing impaired.

It’s not there because CBS feels particularly sympathetic toward the hearing impaired.

The 21st Century Communications and Video Accessibility Act of 2010 and the FCC implementing rules mandate that as of Sept. 30 prerecorded, full-length programs that were captioned for broadcast also have to be captioned for IP online and mobile distribution.

That pretty much covers the networks and most of their programming.

Local broadcasters’ turn comes next spring. Starting March 30, according to the same law, they must caption full-length newscasts that they simulcast online live or stream within 24 hours of their original broadcast.

Exempt for now from the captioning requirement are video clips — short news segments that account for much of the video content on stations’ websites.

The live streaming obligation is clear. What isn’t is how to fulfill it.

According to IP video pros, the streaming of live broadcasts with captions is no easy task. Not only do the captions have to be kept in sync with the video, but they have to be streamed so that they can be played back on computers and mobile devices with a variety of operating systems and video formats.

“Until now most of the focus on Web captioning has been video on demand [non-live video], which has pretty broad captioning support across the different formats, but live captioning has mostly been an afterthought, says Jason Livingston, product manager and developer at CPC, a maker of filed-based captioning creation and encoding software.

While it may be difficult, it’s not impossible, say the IP video technology and platform providers. And if it was an afterthought, they say, it no longer is.

WorldNow, a leading content management provider for broadcasters with its own in-house video platform, is on track to meet the FCC mandate and its clients’ needs, says Joe Sticca, SVP, product development.

The live captioning system is now in the proof-of-concept stage and should be ready for its first round of testing by the end of December, Sticca says. “We are darn confident of meeting the March deadline.”

The big challenge has been coming up with a system that works with all the mobile devices, particularly Apple’s iPhone and iPad, Sticca says.

For the Apple and late-model Android devices, WorldNow has come up with an either-or solution, Sticca says. Rather than just one overlay inviting users to start the video, he explains, the solution employs two — one that will launch the video with captions and one that will launch it without.

WorldNow COO Bob Mischel says the development work is based on research in the deaf community and aimed at coming up with best possible on-screen presenation. “We are really trying to work on the location of the closed captioning [on the screen], the opacity, the colors and the fonts.

“There is a lot to this. We are trying to be thoughtful about it and put in the resources and energy not just to get it right, but to lead the field in providing this service.”

Critical Media, which supplies the Syndicaster video platform used by content management system provider Inergize and others, will be ready to go next spring, says Ion Puspurica, EVP and GM, media services group

“Syndicaster delivered closed captioning for recorded programming ahead of the Sept. 30 deadline and it is committed to delivering closed captioning for live streaming ahead of the March 30 deadline.

“We have been running successful tests of delivering captions to live streams and will add appliance and player support for DFXP and WebVTT caption formats for live and near-live programming in mid-March 2013.”

DFXP and WebVTT are standards developed by the World Web Web Consortium for putting text in video.

When the time comes, Critical Media will implement the live streaming capability with a remote software upgrade of the Syndicaster encoders that are installed at each of its client-stations, Puspurica says. “Based on our massive, geographically distributed broadcast video capture and processing infrastructure, we’re uniquely equipped to enable any U.S. broadcaster to comply with the captioning requirements….”

Internet Broadcasting, the content management system provider to Hearst Television, Post-Newsweek and other top station groups, uses third-party video platforms and so is not developing its own IP captioning solution. But IB is working with several of its clients of finding the right solution, says Dave Michela, VP of business development.

“We will have discussions with our stations to determine what is the best fit based their infrastructures and budgets and then design a solution for them.”

Those discussion may involve any of a number of third parties expert at moving video onto the Web and into mobile devices, including Syndicaster, Ramp and Anvato.

“We are not in the business for forcing our clients down a particular path,” Michela says. “We want to provide them with as much flexbility and control as we can.”

Another company claiming a fix is Uvault, a content delivery network that distributes video on the Internet for a variety of clients, including broadcasters.

According to President Eyal Menin, Uvault, in cooperation with CPC, has developed and is now testing a live captioning solution with Wisconsin Public Television and a major commercial station group, which he declined to identify publicly without its permission.

The partnership was announced last April at the NAB Show.

“The challenge really was to try to address these devices and create a solution where the traditional broadcaster doesn’t need to go out of the way to add more equipment and spend more money.”

To implement the Uvault solution, it’s up to the station to strip the captioning data stream created for the broadcast and then, after running it through the CPC software in a local workstation, send captioning to Uvault along with the related video stream.

“Our solution is to take the video, take the data of the closed captions, merge it together and put it on the Web so it’s available to all these different devices,” Menin says.

Uvault uses Wowza media servers. “It allows you to send in one signal and then on the fly replicate and translate it so that it can be played on the protocols that Apple devices need, that Android devices need and the different computers need, Menin says.” The whole process takes just a few seconds, he adds.

Tom Micksch, digital production specialist at Wisconsin Public Television, says that the public broadcaster has been testing the UVault/CPC solution since last August and has so far been impressed. However, he says, WPT will not be making a final decision on technology until next year.

Micksch says he is using an EEG captioning encoder/decoder, Telestream Wirecast streaming software along with the CPC live streaming software.

Most stations have the basic gear, says Micksch. “If you are going out completely cold and didn’t have any of this material, then it could be kind of expensive to get into It.”

Micksch has looked for alternative solutions. “In a lot of cases, it seems that people don’t even understand that there is a difference between them, let alone supporting it. So, it’s been surprising and frustrating that there hasn’t been more activity on this front.”

If a station goes the UVault route, it becomes wedded to the service, Micksch says. “To use their service for the captioning, you need to use their servers as well. They have to host the video. It’s a turnkey service.”

One option for broadcasters who fail to get their act together by next March is to stream open captions — captions that everyone can see, Micksch says.

But open captions can be annoying for those with normal hearing and they can cover up bottom-third graphics, he says. On the other hand, he adds, they are relatively easy to do, requiring no additional hardware or software, and they would fulfill the FCC requirement.

One thing broadcasters cannot do is use the exemption on video clips to circumvent the requirement on live and near-live programming.

The FCC rules say: “[W]hen substantially all of a full-length program is available via IP, we will not consider that program to be a ‘clip,’ but rather, a ‘full-length program’ subject to the IP closed captioning requirements.”