云端编码
As streaming media enters its second decade, it’s not all that often that we get to explore a totally new topic. But now we do, as I recently took my first look at cloud encoding. 什么是云编码? 在高水平上, it’s an internet-based service that encodes your uploaded source video files and delivers them back to you or your distribution partners.
为什么是云编码，为什么是今天? We like to talk about delivering to three screens—computer, TV, and mobile—but if you take a moment to do the math, it can easily jump to 20 or 30 screens. For delivery to computers, you’ve got services such as Hulu, YouTube, Yahoo! 视频信号.tv, Veeple, Vimeo, Brightcove, Ooyala, and VMIX. 在客厅里, you’ve got different cable providers and different devices, 对于移动设备, it’s challenging to even estimate the total number of available devices and services. Understanding and economically delivering compatible video files and metadata to these screens is an increasingly technical and logistical challenge.
Then there’s the cost of building and maintaining rendering farms. An eight-core Nehalem-based rendering station costs about $10,000, plus software and staffing costs. Unless you’re in the business of providing encoding services for third parties, these costs are all ancillary to your core business (e.g.(开销)，必须最小化.
谁应该考虑云编码? Well, I’ll detail that later in the article. 现在, if you’re considering buying a workstation-class computer solely for encoding (or you just did), cloud encoding should be on your radar. You should also read on if you’re even casually interested in cloud computing in general, as the infrastructure used by most of the cloud encoding companies is pretty fascinating stuff.
什么是云编码?
Cloud encoding services are internet-based companies to which you upload your source video files and then choose encoding and delivery parameters. They encode and deliver your files to the designated locations and/or third-party services.
很简单的东西，嗯? What’s not so transparent is the proper acronym. Candidates are Software as a Service (SaaS), 基础设施即服务(IaaS), 以及平台即服务(PaaS), and the ideal cloud encoding service should provide components of all three.
云编码是如何工作的?
There are two aspects to this question. 操作上, 让你的文件编码, you upload your source video files to the cloud encoding service, typically through some type of XML or REST-based API. (其他, which stands for REpresentational State Transfer, is a protocol popular for web-based services.) Some services also enable a more casual interface for less-automated activities, such as creating FTP-based watch folders or even direct file upload similar to YouTube. 一次上传, you create and designate encoding templates and delivery destinations, just like you would with a desktop program, 然后开始编码工作.
同样，非常简单的东西. It’s what happens behind the scenes that’s the fascinating part. Let’s start by considering the problem.
Unless you’re a 24-hour video producer—such as CNN, YouTube, and the like—most of your video encoding is done in army mode, 或者抓紧时间等待. You work hard to produce your source video files as quickly as possible (with encoding resources completely idle) and then need to encode the video into your delivery formats as quickly as possible (with encoding resources pedal to the metal). If your encoding stations are idle most of the time, it’s obviously wasteful. 在所有条件都相同的情况下, you’d prefer to be able to buy (or rent) an encoding station only when you need it, and then rent a server or servers that can produce your encoded files as quickly as possible.
For most cloud encoding services, Amazon Web Services is that encoding station. Amazon has created a huge global infrastructure to support its retail empire and rents access to that infrastructure on a pay-as-you-go, 无消费合约的基础上. The company offers a range of services, including Amazon Elastic Compute Cloud (or Amazon EC2), which provides scalable computing capacity, and Amazon Simple Storage Service (Amazon S3), which provides online storage capability.
相关文章
It's the great transcoding debate: Do it in-house, or enlist a cloud-based service to do it for you? This 流媒体 East panel addresses the pros and cons of both approaches.
5月25日
SaaS company's relationship with Akamai brings deep data features to customers.
四月十二日
提及的公司及供应商