This is a guide to the process of publishing talk recordings
- 1 Step 1: gather and organize the files
- 2 Step 2: preflight check
- 3 Step 3: encoding
- 4 Step 4: inspection
- 5 Step 5: upload to the NAS
- 6 Step 6: upload to youtube
- 7 Step 7: promote the new video
Step 1: gather and organize the files
The first step is to gather the video footage from the different cameras and organize them.
The encoding script expect all the footage for one talk (can be multiple files) to be in a dedicated directory. The suggested layout is the following:
. |-- first-talk | |-- XXXX.MVI | |-- YYYY.MVI | |-- title.svg | `-- metadata.mk |-- second-talk | |-- AAAA.MVI | |-- BBBB.MVI | |-- CCCC.MVI | |-- title.svg | `-- metadata.mk |...
The name of the directory will be used as the file name for the final video, so it should be as meaningfull as possible.
The following convention has been used previously:
12SAT011515-session-title └┤└─┤└┤└┤└┴─ minutes │ │ │ └─ hours │ │ └─ room (numbered from 1 in the order in which they appeared on the schedule) │ └─ The three first characters of the day └─ the last two digit of the year
For instance this particular file was recorded in 2012, on the Saturday in the first room at 15.15.
Generate extra files
To generate the final version of a video, the script will need the following files for each talk (in the same directory as the video footage.)
a title screen with the name of the video and the speakers using this template: File:Title-template.svg
A makefile with the video's metadata. Here is an example:
TITLE="Making Ourselves Scarce" SPEAKERS="Dougald Hine" TRACK="Welcome to Dystopia " DATE="2013-11-09T15:15:00"
Step 2: preflight check
- Look at the video files
- is the quality ok?
- are both the beginning and the end of the talk present?
- do you have all the pieces? (the end of the first segment should correspond to the beginning of the second ect)
- Look at the title file
- is all the text in the limits of the page?
- is the text correct?
- Look at the metadata file
- there should not be any new line in the strings (replace by \n)
Step 3: encoding
Step 4: inspection
There should be 2 new files in the current directory: A mpeg video and an ogv video. Open the two files and check that the quality is ok.
Step 5: upload to the NAS
scp ./XXXXXXXXXXX-my-video.ogv nas.ffkp.se:/fscons/Public/<year>/<trackname>
scp ./12SAT011515-open-animation-projects-state-art-problems-and-per.ogv nas.ffkp.se:/fscons/Public/2012/miscellaneous
Step 6: upload to youtube
The mpeg file is used on youtube because it seems that ogv files do not work as well (and it's in higher quality.) The upload can be done automatically using the video script:
This requires you to have the [googlecl] tool installed (
yum install googlecl) and access to fscons' youtube account.
The above command adds a very limited amount of metadata, it will only set the title of the video to [speakers]: [title] and the category to Nonprofits & Activism
To fix the metadata, the following have to be done manually:
- copy the abstact from the schedule
- add tags
- optionally change the category (some video may be better categorized as Science & Technology for instance)
- set the license to creative common
- set the recording location to
- set the recording date
In the description, it is nice to add the link to the ogv url, for instance:
Download video: [ogv url] [abstract]
Step 7: promote the new video
New video: [speaker names/@account] on [subject/short title] [short url] #FSCONS[year]
Example: New video: @jvelkova & @k_dmitriev on Open animation projects #FSCONS12
A talk by [speaker names/+accounts] at FSCONS [year]
and include the youtube video
Example: A talk by +Konstantin Dmitriev & Julia Velkova at FSCONS 2012