Category Archives: Video

6 Reasons You Should Consider HTTP Dynamic Streaming

For years, we’ve been able to deliver video content on web servers using the HTTP protocol as progressive downloadable video. Problem is, if a user wanted to jump to a specific point in a video they would need to wait until the video had downloaded from the beginning to the point where they wanted to watch.
Real-time streaming severs solve this problem by just delivering the bits that are needed to display the video on the screen at the moment but they have their own set of challenges. Specific ports needed to be open so getting through firewalls could be problematic. Real-time streams needed constant, dedicated bandwidth so having multiple, concurrent streams could seriously impact the network.
Enter HTTP Dynamic Streaming (HDS). The concept for this technology has been around for some time. I remember dabbling with it in QuickTime even before HTTP Live Streaming was announced by Apple. The idea is to break up a long video file into smaller segments and fragments. When a user plays a video online, the video player loads a manifest or xml file that tells the player what video segments and fragments to load and plays those files back seamlessly in sequence. Here are some reasons why you should consider HTTP Dynamic Videos to deliver your content:
  1. HDS runs on Apache using a free Apache module called Origin . No expensive streaming server needed to deliver video on demand (still need a streaming server to use HDS for live broadcasts, online video recording, DVR functionality, and DRM).
  2. HDS is delivered through port 80 so all the firewall issues we’ve had in the past are gone. Hurray!
  3. Multi-bitrate support means better network efficiency. Bitrate switching works “mid-stream” as mobile users move from wi-fi to cellular networks.
  4. Better bandwidth control. Real-time streams demand dedicated bandwidth. HDS is managed as downloaded files on the network so latency can be a little more forgiving.
  5. Seeking within the video. HDS allows the user to seek to any point of the video without having to download the entire clip.
  6. A future release of Adobe’s HDS will include iOS support. This will allows us to serve a single set of video files to all platforms. This one is huge! Who wants to maintain a 2nd sever and an extra set of media to serve to iPhones and iPads? Not me!
I’m happy to see Adobe build tools to support HTTP Streaming. Hopefully we’ll see additional features added to the Creative Suite that will help make packaging HDS content easier.

H.264 Licensing for Education

I’ve had the opportunity to read the H.264/AVC licensing summary (PDF file), and then request the official AVC Patent Portfolio Licensing document. It arrived a few days later and I’ve had the opportunity to read the whole text when I’ve had free time during the last couple of months. The document outlines the official terms and conditions for using the H.264 video codec. It’s been a very interesting read (the whole set of documents is about 118 pages). Here’s what I’ve gathered in reading the document and doing a little research online:

  1. H.264 is different than most of the video codecs we use today in that in many instances there’s a royalty fee attached to the viewing of the video.
  1. I asked the MPEG-LA about any exemptions for non-profit organizations. This was their reply: “All Licensees sign the same AVC License and are subject the same terms, including non-profit organizations.”
  1. The MPEG-LA (governing body of this license) wants all content producers using H.264 whether you receive remuneration or not from your video to sign the licensing agreement. (See this post by Jan Ozer )
  1. This is one serious legal document. You should probably have your legal counsel view every detail of this document before you or your institutional representative sign it.
  1. You may be taking your chances if you’re thinking you can use X.264 without making any kind of royalty payments to the MPEG-LA (See http://www.x264licensing.com/faq and the comments of this post by Jan Ozer ).
  1. There are a few instances where no royalties are required. They include:
    1. Making your content freely available on the Internet.
    2. Videos sold under a Title-by-Title structure that are 12 minutes and under in length.
    3. Videos sold by a Subscription with under 100,000 subscribers.

Now where does educational content fit in to all of this? Students pay tuition in order to take classes and indirectly pay to view content specific to that class. If that content is in H.264 format and is not freely available on the Internet, which licensing model does it fall under?

In Article 1.43 of the AVC Patent Portfolio License it gives the definition of of Title-by-Title AVC Video as: “Commercial AVC Video which is Sold to an End User in connection with the End User’s request for the specific content represented by the Title of AVC Video and for which and End User is obligated to pay any form of remuneration.”

By-Title could fit if you’re thinking a course is a title and you’re interpreting the collection of videos used in the course as a title.

In Article 1.41 of the same document it gives the definition of Subscription AVC Video: “Commercial AVC Video which is not Title-by-Title AVC Video and for which an End User is obligated to pay any form of remuneration.”

Subscription could work just by default if Title-by-Title doesn’t apply to you. Students could be seen as subscribers to a particular course, just with a set time limit.

Now, I’m no lawyer, and I would strongly recommend you talk to one before signing any legal document, but you can see there’s no clear model that fits for education. I suppose you could pick a model that best serves your particular setup.

Whether the videos you are delivering in the Title-by-Title or Subscription models fit within the royalty-free structure or not, be prepared to do a little record keeping. Keep track of your viewers so you can provide records if needed. If you exceed the criteria for royalty free video, be prepared to pay the following royalties:

    • Title-by-Title longer than 12 minutes:
      • 2.0% of the remuneration paid in Licensee’s first Arm’s-Length Sale or $0.02 per Title, whichever is lower. (This could get a little tricky trying to determine the actually cost of the video from tuition).
    • Subscription greater than 100,000 subscribers:
      • 100,001 to 250,000 = $25,000
      • 250,001 to 500,000 = $50,000
      • 500,001 to 1,000,000 = $75,000
      • more than 1,000,000 = $100,000

H.264/AVC Video is fantastic video codec. You can get higher quality video with lower bit-rates and smaller file sizes. With all the electronic desktops, laptops, tablets and smart phones that use AVC hardware decoding, educational institutions need to be aware of the terms and licensing issues that come with the technology.

Oh, and in closing, Article 6.1 of the AVC Patent Portfolio Licensing document says: “This Agreement shall expire on December 31, 2015.” So be prepared to take another close look at this in a couple of years. These terms are subject (most likely) to change.

Side note: Always, always, always have a master file of your video projects so when the time comes, whether it’s an updated H.264 codec or some new technology, its easier to re-encode your videos!

Importing Video in Adobe Flash Catalyst

We recently finished an overhaul to a product called Methods of Language Teaching that combines instructional text and video examples to help teachers learn various techniques in teaching foreign languages.

 

The project is delivered via DVD-ROM and as a website. The final product was sent off to a publisher. This time around we used Flash Catalyst to create the project. We learned quite a few things about the development environment. Today I’d like share what we found out about video. It’s really easy to import video into your Catalyst project. Just choose File > Import > Video/Sound File…, select your FLV or F4V file and position the video on the page. The imported video is integrated into a nice looking video player that you can swap between two different visual styles, wireframe and standard.

One thing we learned quickly is that there’s a limit to the size of video you can import. Catalyst limits the file size of video to 150MB.

For the Methods project we had 10 videos ranging from 70MB to 140MB in size. After importing all those videos, we discovered that the size our .fxp project file had ballooned to 1.2GB! Catalyst stores those videos within the project. When you export however, those videos are saved out separately and linked to your main .swf file to keep load times down (phew!). We found that, as the size of our .fxp file went up, the performance in using Catalyst went down. If you do plan on importing lots of video into Catalyst make sure you have lots of RAM on your system.

There was another quirky thing we found about the playback of the video. The moment between displaying a page and having the video load, the video player displays at a smaller size then pops to the size of the video as it loads. When viewing the file locally, you could barely see it. Viewing it on the web however, with a little latency thrown in, it was a lot more distracting. An alternative way to get video into your project would be to import your .fxp file into Flash Builder (don’t forget Flash Builder is free for students and educators) and add the video playback component there. That would keep the size of the development file down and have the video player not re-size on a page. The disadvantage of importing into Flash Builder is that you won’t be able to make any changes again with Flash Catalyst. It’s a one-way import so plan accordingly!

In all, Flash Catalyst is a really slick development tool. It allowed our artists to create a nice looking interface and allowed them to easily wire up that interface to navigate the various pages of the application. As with all development environments, there are limitations. I keep hearing that Flash Catalyst is at version 1.0 and will see many more improvements. I’m confident that will happen. I’ve submitted these issues to the Adobe Bugs Report Form added them to the Flash Catalyst Ideas site. If you agree, please vote for them!

Fix Video Sizing Bug!

Option to Not Store Video in .FXP File .

WebM, A New Era for Web Video

There were quite a few significant announcements in the Google I/O Conference Keynote today. None more important in my mind than Google open-sourcing the VP8 video codec. Yesterday, the web video landscape was pretty one-sided with H.264 video as the only game in town if you wanted to deploy video to desktops and handheld devices. H.264 is a licensed technology that may require royalty payments from content creators in the future.

Today, Google announced WebM, an “open, royalty-free, media file format designed for the web.” VP8 is the video technology in WebM. Most of the web browsers out there, including Chrome, Firefox, Opera and Internet Explorer have announced support for WebM. Hardware vendors including AMD and Nvidia have pledged support to hardware accelerate the technology. Adobe also announced they will support WebM in Flash.

This is a huge win for content creators, especially educators who would prefer not to have to pay future royalty fees for the content they deliver to students. Open WebM make the perfect companion to open HTML5.

What remains to be seen is how well software companies can implement encoders for VP8, how well browsers and other players can decode the video, and if Apple will embrace this new open video standard.

A Codec Conundrum

I really like H.264 video. I love the quality and its efficiency. I love watching video on my iPhone, on my MacBook Pro, and on a Blu-ray player. These experiences are just amazing!That said, there something that worries me about this technology. The H.264 video codec was created by many different companies and is founded on many different proprietary patents. There is a license for using H.264 video and that license is managed by the MPEG-LA. Part of this license contains something we’re not quite used to in the digital video community. That is, in many instances, content producers who distribute their content in H.264 are subject to a royalty fee.According to their current summary for license terms (PDF file), content publishers who sell H.264 video greater than 12 minutes in length on a title-by-title basis are currently subject to paying 2% of the price paid, or $0.02 per title, whichever is less. Currently, there are no royalties for video titles that are 12 minutes or less in length. For subscription-based services, 100,000 subscribers or more require royalties to be paid. Less that 100k subscribers have no royalties required.In instances where remuneration is from other sources (like advertising in free television), there are 2 options for royalty payment: 1) “a one-time payment of $2,500 per AVC transmission encoder”. The MPEG-LA is thinking in terms of broadcasts where the video is broadcasted via encoder to all end-users. 2) A yearly fee per “Broadcast Market”. This fee has a graduated scale that starts at $2,500 per year for 100,000 “television households” and tops out at $10,000 per year for 1 million or more “television households”.For Internet use that isn’t sold by a title-by-title or subscription basis (see above for those terms), the MPEG-LA has announced that they will continue not to charge royalties for H.264 for the second term (until January 1, 2016). Their summary for terms has this to say about after the first term expires: “after the first term the royalty shall be no more than the economic equivalent of royalties payable during the same time for free television.” Since television broadcasts and Internet video have very different processes for publishing video (a transmission encoder makes sense for live broadcasts but doesn’t really fit for online video), this statement leaves a lot of ambiguity as to what is really going to happen in 2016 for Internet video. I do believe that this statement is there to prepare us for the inevitable: All Internet video publishers will need to start paying royalties to publish content in H.264.Now, here’s my conundrum. Do we build repositories of video content using H.264 in hopes the MPEG-LA will extend their no royalties clause in the future? Should I stick with the somewhat older royalty-free codecs out there while waiting for more information from the MPEG-LA? Either way, I think it’s best to make sure we can easily re-encode our archived video to a different codec if the need arises. Suddenly Sorenson Spark and VP6 (older royalty free codecs included in Flash) don’t seem so outdated!