Whether it‘s a corporate event, a press conference, a product launch or a sales meeting with thousands of employees - digital livestream events are becoming increasingly popular. To ensure that the livestream from the studio to the audience arrives smoothly, without delay and no matter what, we have developed Tournet Streaming.
With Tournet Streaming, you get a geo-redundant infrastructure that continues to run stably at any moment, even in the event of widespread Internet or power outages. The principle behind it: The stream is transported from the production studio to the audience on two completely separate channels. Two encoders in the studio, two Internet uplinks and several servers in different data centers on different continents guarantee fail-safe live streams. If one pipeline fails, the second one takes over without interruption - the viewer doesn‘t notice a thing. Almost as if the person at home had two setup boxes with two satellite dishes on the roof.
The highlight: our watermark is encoded into the stream on the server side. So even experienced users don‘t have the possibility to remove it.
Finally, an add-on for the transmission of confidential and sensitive video streams: Our team has developed the Tournet Watermarked Streaming for you. It protects your customers from data theft by encoding a personalized watermark into the stream for each guest. This way, shared screenshots or recordings can be traced back perfectly.
We almost forgot to tell you: Adaptive transfer rates, multi-language support and subtitles of course are also possible.
Detailed monitoring & reporting on audience behavior
When the largest AWS data center "us-east-1" went down on 12/08/21 (report on Bloomberg), we were in the middle of an event production that ran partly through this data center. Thanks to Tournet Streaming's redundant infrastructure, the major data center outage at the world's largest cloud provider went unnoticed by the event's guests.
In a lot of cases redundant livestreaming works by streaming against two ingest endpoints, for example via RTMP or, even better, via SRT. If one signal path fails, the provider takes over the other signal. Only the path from the streaming device on site to the provider is redundant. If the provider's central server fails, the stream is over.
Our solution encodes both Ingest-signals at all times. Failover in case of an error then takes place automatically in the viewers browser. This means that the entire path from the studio to the viewer is redundant, not just the path from the studio to the provider. It's like having two SAT receivers with two satellite dishes at home.
The hurdles are not particularly high. On the technical side, it is only important that we have each language as a separate audio source. This can be done via embedded audio channels in the HDMI/SDI/NDI signal (e.g. the second language is on SDI channels 3 and 4) or separately via DANTE, AES or if necessary, analog.
Of course, interpreters are needed for this and the sound engineers have to produce a further mix accordingly. We can help you with that as well.
To prevent leaks of confidential information, we have developed the possibility for each viewer to have their own name and/or email address embeded in the form of a large watermark in the live image. Accordingly, the costs for such an event increase proportionally to the number of participants - all viewers de facto get their own stream. Therefore, it is necessary to work with a clearly defined guest list to ensure that each viewer is getting their personal stream. The computational and thus cost expenditure for this procedure is higher, which is why we have so far only used it for events with up to 1000 participants. However, since the processing of the stream is cloud-based, the system also scales beyond that.