Zencoder 07/23/12

This week’s Tools Spotlight interview is with John Riske, Director of Business Development and Marketing of Zencoder. Zencoder is an API-based online video encoding service. They convert videos from your website, application, or video library into formats that are compatible with web playback, mobile phone, or any other device you need to support. Arguably the fastest web encoding platform, Zencoder has an interesting origin story and a unique set of open source tools that make it worthy of a closer look.

 

The majority of video editors encode on their computer and network and are used to having local control of their media. Explain how encoding in the Cloud works and what the benefits are.

Video encoding is a processor-intensive activity, and scaling operations to meet a growing business – let alone satisfying big spikes in demand – can be expensive and a serious distraction. Zencoder provides video encoding as a service. Through a simple API, we provide high-performance, high-quality video encoding for web and mobile, at any scale, small or large.

The “Cloud” is sort of over-hyped these days, and the fact that we are based on Amazon Web Services is almost beside the point.  What it means for our customers is that they can throw any amount of video at us (and not worry about it) and pay only for what they use.  We charge based on the number of minutes of output video encoded.

The most common use cases boil down to:

• Converting large libraries to a new file type
• Handling day-to-day encoding volume

When I think about content-based Cloud services, the first thing that comes to mind is the constant bandwidth battle.  It seems like this would be an even bigger concern when doing Cloud-based video encoding.  What are the common challenges and misconceptions of this workflow?

This is a great question.  File transfer is a concern for customers dealing with big files (AVCHD, ProRes, etc.)  For some scenarios, like one-off encoding jobs, bandwidth constraints might mean that it’s better to handle the job in house.

However, using the Cloud becomes favorable when encoding to multiple outputs, and especially if there’s good connectivity and an accelerated file transfer tool like Aspera available.  Moreover, once the file is stored in the Cloud it becomes much easier to pump out additional versions as needed, and you won’t face that transfer obstacle again.

If you’re producing lots of renditions over time it’s also advantageous to deliver those renditions from the Cloud, rather than bring them back in house and redistribute from there.

We’ve published a white paper on this very topic, and it presents a some practical tips on gaining efficiencies when dealing with large files in the Cloud: Rethinking Large Video Files in the Cloud: Strategies for Eliminating Bandwidth Bottlenecks.

Zencoder has an interesting origin story.  Talk about how the company came about.

Zencoder’s three founders (Brandon Arbini, Jon Dahl and Steve Heffernan) have been working on video encoding for some time, and much like other areas of innovation, market demand and necessity pushed them to develop a better method for encoding video.  Lots of people had great ideas for online video products and services, but the expense and hassle associated with scaling encoding operations were major obstacles.

Initially, they worked in conjunction with On2 (the codec creator) and launched one of the very first Cloud encoding services, called FlixCloud.  Google bought On2 and closed down FlixCloud, but the founders were able to re-launch the service as Zencoder.  They subsequently went through the startup bootcamp Y Combinator, which gave them a chance to further refine the product and business model.

Today we have over a thousand customers, with users ranging from web video publishers like Funny or Die to broadcasters like PBS and prosumer video service providers like SmugMug.

There are key areas where Zencoder feels that it stands apart from the competition like integration and particularly encoding speed.  Talk more about this.

Zencoder is a developer-centric organization and our main focus is on performance, which we measure in terms of video encoding speed and encoding success rate.  A lot of our performance advantages are achieved through how we’ve designed our application, and how quickly it’s able to scale to meet customer demand.  We encode millions of videos per month, and system-wide we are encoding video at 1.8x faster than real time, at a success rate of 99.98%

Almost all of our customers integrate with us via API, meaning that some sort of application (a CMS, workflow tool, etc.) is talking to our application.  We’re focused on making that as easy as possible.  Aside from having a well-constructed and thoroughly documented API, we provide integration libraries in different programming languages, and tools like our API Request builder make it easy to get up and running

What are some well-known companies that use Zencoder and how do they use the service?

We’re focused on encoding video for web and mobile distribution and the range of companies that use us underscores the flexibility of an API-based service.

A few examples:

• TwitVid is the largest service for sharing video on Twitter, and they use Zencoder to encode user-submitted video
• PBS uses Zencoder to prep video for delivery to iOS devices.
• IGN encodes video into several output files, so they can go to consumers on whatever device they’re on.
• Video collaboration platform ScreenLight uses Zencoder to encode videos so that their customers can view video on any device.

You have a lot of interesting open-source tools, particularly your HTML5 video player (video.js). Talk about some of these tools and how they are useful to companies.

As mentioned above, we use and create open-source tools and really believe in giving back to the open-source community. Zencoder open source tools is a more thorough rundown of the software that we’ve created and use.

There’s a big (but certainly not complete) shift underway in how video is played in the browser and on devices. The movement is away from third party plugins like Flash, Silverlight and Quicktime, and toward direct support for video in the browser.

All web pages are made up of HTML, and the latest version, HTML5, allows developers to add video to a webpage just like they would a picture.  This will allow for easier deployment of video, for standardization of video across browsers and devices, and will improve performance.

Video.js is the most popular open-source HTML5 video player, and is a fantastic complement to Zencoder.  We want to make it easy to encode video, but it’s equally as essential that it’s easy for developers to catch this next wave of video playout.

Video.js is a JavaScript and CSS library that makes it easier to work with and build on HTML5 video.  Video.js provides a common controls skin, fixes inconsistencies between internet browsers, adds additional features like fullscreen and subtitles, and manages the fallback to Flash or other playback technologies when HTML5 video isn’t supported, and also provides a consistent JavaScript API for interacting with the video.

As more and more video services are sent out of house, there is a potential for people to eventually forget how to do these procedures all together.  That’s why it is good to see that your FAQ has a lot of video compression fundamentals.  How do you balance the need to relieve clients of the technical grunt work while still helping them stay media literate?

We strike that balance by making it easy for a novice to use presets and get great-looking video, while simultaneously exposing granular settings via the API for the encoding expert.  You can really have the best of both worlds:  the expert has the control they need, while the novice can be successful and become better acquainted with video technology if they so choose.

Are they any future partnerships (that you can talk about) that integrate Zencoder technology?

A big part of my job is finding technology partners that complement our service and make it easier to use or better.  A couple of examples are our partnerships with Aspera and BuyDRM.

Aspera makes products to accelerate file transfer to the Cloud.  BuyDRM is a partner that helps with certain aspects of content security (DRM).  We’re always looking for good ideas, so send any thoughts my way.

What question(s) do you have for content creators and specifically video editors?

As a video editor, what are the most common output file types from your processes, and what will they be in three years?

What are the biggest challenges that you face when it comes to web and mobile video?

Is there anything that I didn’t ask you that you want to mention?

Not that I can think of!  Thanks for the opportunity!