2018年4月20日
By Tim Siglin 特约编辑
专题文章

编码器设备买家指南

Here’s a list of things that software-only encoding solutions have in their favor: quality and ubiquity. 随便选两个. 但是说到编码和传输的速度, 特别是在低延迟编码方面, 有时，仅使用软件的编码解决方案并不能解决问题. 对于高帧率(HFR)也是如此。, 高动态范围(HDR), 甚至4K超高清(UHD)编码.

This Buyers’ Guide on encoder appliances will explore the tradeoffs between hardware and software encoding solutions, hopefully empowering readers to more easily determine which solution suits their specific needs.

硬件vs. 软件的困境

在流媒体的早期, every format and codec was defined by its capability to deliver on-demand content in one of three ways: file downloads, 渐进式下载, 或流.

File downloads were fairly straightforward: a video file needed to be fully downloaded before playback could begin. 另一方面, progressive download files could start playing the content after a certain percentage of the file was downloaded, on the assumption that the file would download faster than it was played back. This meant the initial wait for the beginning of the file to download gave the patient end user a “head start” in playback that would, ideally, 没有缓冲.

第三种方式, 真正的点播内容流, required hefty hardware assistance in further compressing video content into data rates more akin to dial-up modem speeds than to CD-ROM data rates. 说实话, though, hardware assistance was needed for almost every early video codec, 不管数据速率如何, resolution, 甚至是帧率.

In fact, the only codec in the first year of streaming media (the 1997–1998 timeframe) that had a software-only encoding option was MPEG-1 and its audio sidekick, MP3格式. MP3, which has just reached the out-of-patent-protection stage after 20 years, was called that because it was the third audio coding format in the MPEG-1 video and audio standard ratified by the Moving Picture Experts Group (MPEG).

As more powerful general processors—whether general-purpose processors (GPPs) or central processing units (CPUs)—emerged, bringing with them an opportunity to move the next generation of video (MPEG-2) from hardware-only to software-only encoding solutions, there was often a sacrifice to be made in using software-only encoding: the final output often reduced the frame rates from 24–30 frames per second down to 10–15 fps to lower the overall encoding session length.

Part of the reason these software-only options didn’t cut it was due to the additional need to interweave the MP3 audio and MPEG-2 video elementary streams together in to a multiplexed stream suitable for transmission. 该多路复用, or muxed, 文件通常被简称为M2TS, 用于MPEG-2传输流.

这种MPEG-2传输流技术, 现在已经20多岁了, still forms the basis for the majority of Apple HTTP 在线直播 (HLS), 使用更新的编解码器取代MPEG-2视频编解码器:H.264, also known as MPEG-4 Part 10 or as Advanced Video Coding (AVC).

为什么硬件仍然是必要的

HLS allows “streaming” delivery of a series of small files at a predefined data rate that are downloaded, 组装成一个播放列表背靠背播放, and then played in sequence until no more files (known also as “fragments,“段,或“数据块”).

The lack of additional files for playback either signifies the end of the video being played back, which can be confirmed by comparing the segment number with a predefined numbering sequence in a manifest file, or it signals a need to send subsequent segments at a lower data rate, since the player is not receiving current data rate segments in a timely enough manner for proper playback.

换句话说, 这个版本的流媒体, where the segments are transmitted in a near-real-time fashion, 是否更类似于早期的文件下载方法, 尽管实际上它更像是一个渐进下载, since HLS won’t start playback until at least three segments have been downloaded as a “head start” for continuous playback.

结果是, on-demand content needs to be transformed in faster-than-real-time encoding sessions, where an hour of prerecorded content may need to be converted in less than 30 minutes, 或者大约是实时速度的两倍. 这允许执行各种附加步骤, 例如设置广告插入点和标志, 以及编码多个通道的备用音频(i.e.(外语配音).

实时交付现场内容, though, has to move beyond the multiple-segment scenarios of HLS and its MPEG-ratified equivalent, MPEG-DASH, 或HTTP上的动态自适应流.

克服延迟

流媒体革命已经过去20年了, 2018年, streaming live video in a low-latency scenario often requires hardware assistance, 都是为了遵从老派, real-time streaming protocols (RTP and the more secure RTSP) and for a way to minimize delay or latency.

Inherently, the average frame is 1/25 or 1/30 of a second, so about 33.平均3毫秒. If several frames are needed to analyze changes between frames (aka temporal changes; see our Buyers’ Guide on content- and context-aware encoding), 然后引入了额外的固有延迟, since the encoder will need approximately 100–300 ms worth of video content to accurately assess temporal changes.