Exploration of an Inegrated Media Tool

This was a great talk by Greyson Purcell outlining a simple but powerful video uploading and playback system integrated into D2L.

YouTube convinced our instructors that uploads should be easy, so...

[demo of UI video uploader site]

Why?

Process before the tool:
- faculty contacts us directly
- we train faculty to encode video
- we set up password rights in ActiveDirectory)
- we sent sample HTML instructions
- faculty encodes files, uploads files, enters HTML into ICON (the local D2L installation)

Problem: encoding
A standard encoding of a lecture is about 500 kbps, which means a 50 minute lecture is 180 mb. But faculty would encode at 5000 mbps, resulting in files that were far too large to manage

Problem: HTML
Complex HTML encoding "It was a complete failure"

Plus: we basically had no way to get student material to the web. We basically had students give it to faculty members.


What?

Simple video file conversion.

ffmpeg - is a simple server-based encoding tool. http://ffmpeg.mplayerhq.hu/
Sample command: ffmpeg -i video.avi -ar 22050 -ab 32 -f flv -s 320x240 video.flv

On the web there's tons of materials on how to use this software. Eg. at vexxblog (How to convert/encode files to FLV and how to install ffmpeg )

So we set this up, attached it to some D2L code that captured the student's name, course, role, etc. This created simple upload and display screens for students.

We also wanted to have a nice palyer for the videos. Flash player with features for things ike full screen, popup menu, etc. We also added a comment system (because it uses the D2L login, you can't spoof comments).

Also: we can convert videos to podcasts - ffmpeg supports a variety of formats. We can convert the web based player into a podcast manager too and distribute videos via iTunes.


Pedagogy

Now it's in the hands of faculty. We're going out into this beta period, and we'll see what happens.

We're gtting rid of a hurdle - us - and opening the floodgates. How many videos will people upload? What will the time frame be, will there be a massive rush? How many servers do we need? Do we make this on by default? What about copyright issues?

More Stuff

Also - direct recording from video camera... could be done with Flash media server.

Ability to hide the video, or any way to mix and match rights to edit, access, etc.

Files can be stored as files on a normal web server, or can be streamed using a Flash media server.

Comments could be tied to video time - so when you click on the comment, the video advances to the time.