As young consumers move away from Live TV shows and into Video-on-Demand (VOD), the traditional TV broadcasters must find an effective method to convert their Live assets into VOD.
Historically, TV broadcasters had the function of creating content, and to distribute it to a multitude of home TV viewers. Even in a large country there were typically just two to three large networks, and the whole population would end up watching a small number of the same Live TV shows.
The digital era enabled three changes that challenged that old order:
- First, digital recording created assets that were reproducible, and did not need to be consumed on the spot.
- Second, the digital highways (IP, Internet) made it possible to connect to each house, and later to each portable device directly.
- Third, the smartphone created a portable screen that is suitable to display TV content anytime, anywhere, and on demand.
We are now in the Multi-screen era, where every meaningful TV (or Radio) content needs to be repurposed to a number of screens in order to reach the whole population of viewers. These screens include TV, computers, smartphones, tablets, and other.
All these screens are hungry for more and more content.
On the other side of the network, there are TV producers that create more and more content but have less and less viewers on their traditional TV screens.
The challenge is to distribute this large amount of content to the whole viewers population and to all screens, as fast as possible.
Given that each of these screens is in a different size, given that each of the player devices are connected using a different mode and speed of communication, and given that they do not all play the same file formats, it becomes necessary to convert (or adapt, repurpose) the original TV content into a multitude of bitrates, resolutions and file formats.
In addition, there is a growing use of adaptive bit-rate (meaning that one device may decide to increase or decrease the bit-rate it is playing, to adapt to changes in the communications speed) and this means that any system preparing TV clips for VOD may have to prepare that content in several bit rates in parallel.
Need for Speed
As soon as we began speaking with TV networks about the idea of developing an automated system for converting Live TV shows into on-demand clips of TV shows, one of the first requirements was to have these clips available for streaming “within seconds after their Live broadcast”.
The idea was that Social network users should be able to distribute to their friends any “Fresh from the Oven” segment, such as a sports action (Goallllll!!!!!!!!!) or a hilarious interview. If that event is not available as a clip within seconds, then it becomes old news.
We implemented a solution that incorporates real-time encoding. In typical installation we would have a server, capturing a number of TV channels (from any source, IP, SDI, ASI, Analog etc.) and perform real-time encode of TV content into the most efficient format possible, such as H.264 or HEVC. The recorded TV show is then available as a digital chunk file within seconds.
Cut to cut Video Editing
The next step in the workflow must involve a video editor that would cut video clips into the precise segment. The mark-in and mark-out must be very precise, and frame accurate.
We developed a powerful but minimal video editor (that we call Clip-Factory) that allows the editing operator to review the segment and its timeline, to mark a number of frames on-screen, and to save the clip they created, all in a matter of seconds.
In the case of commercial TV networks, it is necessary to remove some TV commercial spots that may be included in or around the clip build. For this purpose, we included in Clip-Factory the ability to cut the segment into several parts, and to remove any part with a single click. This can be done by a human operator.
Alternatively, we can use our automatic Ads detection program (that we call AdWatch) which has the capability to “learn” all commercial spots using fingerprinting, and then to send the marks (mark-in, mark-out, per each detected spot) to the Clip-Factory, for removal. This step is fully automatic.
In the early days of creating video-on-demand for over-the-top systems, the actual editing was done by the TV broadcasters, using top of the line professional Video editing systems such as Adobe Premier Pro, Apple’s Final Cut Pro, Avid’s Media Composer, etc. These are all excellent systems, with a multitude of editing functions.
The issue was that each new workstation running these VOD producer software represents a non-negligible investment, while at the same time the number of functions that is actually required for OTT or VOD clips is very limited compared to what these systems have to offer.
At the suggestion of our customers, we were able to identify just a handful of functions that we added to the Clip-Factory. For example, we added the ability to blur a face, remove a logo, add a new logo instead, add a pre-roll and post-roll text, or an overlay.
Thus we are now offering a simple but professional VOD creating and clip editing software. A software that is fast and powerful enough to complete the whole workflow for content clipping into OTT, Web and VOD, at a reasonable budget. Not to mention the fact that Clip-Factory is intuitive and can be learned by any operator in less than 1 hour (not requiring a dedicated employee who is a professional Video editor).
As already mentioned above, each newly built clip can now be transcoded into a multitude of resolutions, bitrates, and file formats (we support all file formats supported by FFMPEG, which practically means: ‘any’).
And we have reached the final step in our workflow; exporting. We allow each broadcast clip, and each format of the clip, to be exported to more than one destination. For that purpose we have created a number of profiles that support exports directly into an FTP address, a CDN, social Media (Facebook, Youtube, ..), any computer in a Domain, eMail, and more.
API for automated workflow
In some cases, there is a possibility to replace all the above manual work by using automation data that is available within the TV station operational systems. For example, if we receive access to the AsRunLog (the actual log of what TV programs were aired) then we can use that data to mark the segments with frame accuracy.
For this purpose, the Clip-Factory comes with a complete API whereas every function described above can be accessed from a 3rd party computer system, and using a simple script language and hot folders.
The system integrated with a number of software modules. You may refer to the following modules for more information: