Legal notice | Copyright
Joint Interpreting and Conference Service
EUROPA > Languages > Interpretation > Speech repository
 
Login
The Speech Repository
What is it?
How it works?
Search speeches
iPods and the like
System description
Web portal
SCICrec
Videos
About this project
Contact

Web portal

For full access to the multilingualspeeches.tv system, you need a password. Once logged in, you get a page where you can search for speeches using a few simple criteria:

  • language (all official EU languages and candidate-country languages)
  • level of difficulty (4)
  • domain of interest divided into EU policy areas (about 30)
  • intended use of the speech (simultaneous or consecutive interpretation)
  • numerical identifier (to immediately retrieve a previously identified speech).

To make the tool easy and quick to use, the only search criteria are “language”, “difficulty” “domain” and “use” (criteria are put in “AND”). You can also search using a speech's numerical identifier if you know it (including several at once, by separating each identifier by a space).

If you don't enter any criteria, the search will return all the videos in the system.

A 20-second trailer (audio only) lets you quickly decide whether the speech's pace, accent or other characteristics meet your needs. For more information on a specific video, you can click "Details".

You can label and save your search results. Labelled search results can be used with SCICrecTM.

Streaming and downloading

On the "Details" webpage, you can:

  • view all metadata associated with a video
  • stream a video (i.e. watch immediately, without having to wait for the entire file to be downloaded)
  • view a transcript (if available).
  • download the whole video for further use (e.g. uploading to a handheld device such as a smartphone or mp3 player, etc. – see the online help section).

Video clips and audio file formats

An average video takes 30 seconds to download over a 3 Mbps ADSL line and takes up about 2.8 MB of disk space per minute.

Most videos are good quality even in full-screen mode, so they can be projected in a classroom.

Audio recordings of mock interpretations are audio-only AAC files recorded at 44.1 kHz.

Video and audio recordings are stored locally on your personal computer, together with the metadata for each video (in XML files). All videos are watermarked to identify their origin (multilingualspeeches.tv).

Video format details:

  • MPEG-4.10, a.k.a. ITU-T H.264
  • 320x240 pixels
  • 25 frames per second
  • video compression ratio = 320 kbps
  • AAC (Advanced Audio Coding, which replaces the older MP3 and is available in all personal computers and all new (portable) electronic devices
  • audio data rate and resolution: 64 kbps, mono, 48 kHz, 16 bit.

More…

Legal notice | Copyright