如何通过再流解决方案最大化受众

最大限度地暴露和达到你的流, 将它们同时发送到多个站点通常是非常有益的. While we may not think of YouTube and Facebook the same way we think of CDNs, 它们的目的相似, 我将在本文中对此进行讨论. The key differences are that YouTube and Facebook come pre-populated with thousands, hundreds of thousands, 甚至是数百万的观众, and they offer ways that people can subscribe for notifications when your show is coming up and when it goes live. 如果你有积极的社区参与, then you can count on the audience being aware that you’ve published new content.

但如果你想在不过度使用你的装备的情况下最大化你的覆盖面, you can utilize some cloud technologies to send your one upload to multiple destinations. 理解这一点很重要，尽管有些解决方案, like vMix, TriCaster, or Wirecast, have the capability to send multiple streams to multiple CDNs from your location, you will need both a beefy computer system and a lot of upload bandwidth to ensure the reliability of each of those streams from your system to each of the various CDNs you want to leverage.

Alternatively, cloud solutions offer the ability to accept a single good stream and then send that to multiple destinations, 甚至可以用转码. 但是为什么转码很重要呢? 为什么我们不能发送CDN想要看到的流呢? Two reasons come to mind. First, 不同的cdn可以处理不同的分辨率和质量, 所以你可以把1080p60甚至4K视频放到YouTube上, 你最好用720p30的流媒体到Facebook. 尽管YouTube对实时消息协议(RTMP)推送很好, Facebook想要安全的rtmp. You can manually set each of these from your device or let the cloud service provide the different feeds.

Second, 如果你从远程位置通过手机推送信息, being able to cut the data rate almost in half and utilize HEVC can make your connection better able to get through a potentially crowded or saturated cellular tower. Plus, it will lower the amount of data you are using to send your feed to the cloud. 然后，云服务可以将您的HEVC转换为H.cdn希望看到的内容，并将其发送给多个cdn. 所以不是两个10Mbps的流, 你可以发送一个5Mbps的数据流，减少75%的蜂窝数据使用量.

If you’re covering, say, 全天的体育赛事，每个周末都在做, leveraging HEVC, cloud conversion, 云分布可以为你节省很多钱, while delivering great image quality to multiple destinations on a regular basis. That’s the key to developing an audience: Deliver something good that they can count on, 他们还会回来索取更多.

Multistreaming Solutions

如果你是流媒体的忠实读者, you know there are dozens of CDNs and streaming services to get your broadcast in front of your viewers. As I mentioned, some of them come with built-in audiences, like YouTube, Facebook, and Twitch. 其他的设计让你对谁能看到视频有更多的控制. 它们还能让你更好地控制观看体验, including the ability to white-label the entire viewing page and even create paywalls and integrate copy protection. 但是这些类型的服务不是本文的重点.

I’ve previously written about some iOS solutions for live-streaming production, such as Teradek’s Live:Air该公司已更名为Airmix. What hasn’t changed is the software’s ability to connect directly to Teradek’s own cloud services, called Core. In Core (Figure 1, below)，您可以将单个提要重新流到您想要的任何地方. Core还有其他一些技巧.

Figure 1. Teradek's Core 2.0. 单击图像以查看其全尺寸.

First, 核心可以接收远程摄像头, 这个摄像头在Airmix中可以看到. You can easily cover an event with cameras miles apart; each camera shows up in Airmix, 然后你可以在它们之间切换，就像它们是本地的一样.

其次，Core也能成键. This is when a single source is chopped up and sent across multiple connections to the internet. Teradek的vidu Go硬件可以做到这一点. 然后Core接收数据包, 因为它们走的路径不同而经常出现混乱. It then reassembles the packets in the right order and sends your program on to the next step. 再加上Core处理HEVC的能力, and it’s a nice, 全面的移动生产中心.

其他服务，如Castr (Figure 2, below) and Restream (Figure 3, below Figure 2)真正回到了过程的回放部分, 接受到云端的单个提要，然后将其发送到云端, two, 甚至五个或更多的目的地.

Figure 2. 使用Castr流式传输到多个站点. 单击图像以查看其全尺寸.

Figure 3. 多流与Restream. 单击图像以查看其全尺寸.

Another feature I like about involving an intermediary service such as this is the set of statistics you get on your upload. 你可以在流媒体上传到YouTube或Facebook之前看到它的健康状况, as well as viewers’ comments that the feed is choppy—because someone always comments. You can look at the data that is being received in the cloud and see that it’s solid, with no dropped frames. 因为它已经在云端了, you can be confident that this is what the other streaming services are receiving.

Recently, 我直播了一场慈善音乐会, and Facebook took issue with some of the recorded music that was played in between the live bands’ performances. We would check our upload feed and find an alert that Facebook had stopped the stream. 我们立即重新初始化该连接以保持正常运行. 让这一切发生在云端, 而不是在我们使用的设备上进行实时切换和上传, 意思是换节目的人根本不用想, 其他人可以在另一台电脑上管理Facebook的工作. The crew member monitoring your Facebook or YouTube feed doesn’t even have to be at the event because cloud services can be managed from anywhere.

云服务不仅可以将一个流发送到多个地方, 但它们也可以接受流，而不将其发送到任何地方. This allows you to check your connection to the internet before you need to go live and be sure that everything is working and your connection to the cloud is good. 最糟糕的感觉是需要在下午两点上线.m.，按下设备上的按钮，在下午2点开始播放流媒体.m., and seeing that it doesn’t appear on the receiving end, with no time left to fix it.

Another little-known benefit of cloud services’ ability to accept a stream and not send it is specific to YouTube. 假设你在YouTube上设置了一个活动, 客户端提供的“查看器”链接嵌入到各个页面中. 如果你测试你的连接，它通过YouTube, 你停止测试的那一刻, 视频的嵌入代码是固定的，已经用完了. YouTube automatically creates a new embed code every time you push a test through an event’s RTMP. 为了避免最后一分钟的网站更新, using a service in between and not activating the output lets you test your internet connection without causing any ripples down the line.