Audio and Video Digitization

Overseeing the digitization process of more than 5’000 concerts, 11’000 hours of video and 6’000 hours of audio from the Montreux Jazz Festival archive.

Digitizing the Montreux Jazz Festival archive requires certain standards to be met:

  • Digitized media must be as identical as possible to the original for preservation purposes. The resulting uncompressed media files will make up the primary archive.
  • Digitized media must be readable, duplicable and identifiable for accessibility purposes. The resulting compressed media files will make up the secondary archive.
  • Digitized media must be contained and streamed on secure networks and data storage systems for security purposes.

The digitization project took 5 years and is finished since 2016. Due to the scale of the Montreux Jazz Festival archive (14’000 tapes and 18 media formats); digitization of analogue content is outsourced to Vectracom in Paris, France. Vectracom possesses the equipment, resources and expertise to run several processing chains in parallel thus saving EPFL time and unnecessary expenditure. Thanks to recent developments at EPFL, HD capture of up to a third of the archive is taking place on campus.

Every month until December 2014, approximately 200 hours of video and audio were digitized and about 30 LTO tapes were provided to EPFL by Vectracom. LTO tapes contain both primary and secondary archive formats and are stored in two different geographic locations. LTO tapes were shipped at different intervals so they were never in the same location at the same time. All data from the original digitized tapes is now stored on an object based 3 x 4.7PB geo-localised Active Archive System by HGST.

At the start of the Montreux Jazz Digital Project, two EPFL laboratories conducted a thorough investigation to provide a digitization strategy based on their expertise of the audio-visual sector. The labs are the Multimedia Signal Processing Group (MMSPG) and Laboratory of Electromagnetism and Acoustics (LEMA). The following digital media formats were selected:

Standard-Definition Video uncompressed format for the primary archive:

  • 10-bit 4:2:2 video (AVI, v210 codec) with 96kHz and 48kHz 24-bit PCM audio
  • SD: 720 x 576 @ 25 fps (50i) => 210 Mbit/s (100 GB/h)

High Definition Video uncompressed format for the primary archive:

  • 10-bit 4:2:2 video (MOV, v210 codec) with 48kHz 24-bit PCM audio
  • HD: 1920 x 1080 @ 25 fps (50i and 60i) => 1040 Mbit/s (500 GB/h)

Video broadcast format for the secondary archive:

  • SD: 8-bit 4:2:2 MXF (50 Mbit/s MPEG-2) with 48kHz 24-bit PCM audio (compression rate: 4)
  • HD: 10-bit 4:2:2 MXF (100 Mbit/s AVC-intra) with 48kHz 24-bit PCM audio (compression rate: 5)

Video preview format for the secondary archive:

  • WMV 8 Mbit/s (VC-1), with 256 kbit/s WMA (48 kHz, 24 bits)
  • MP4 300 kbit/s AVC for database preview (64 kbit/s, mono)

Stereo audio format for the primary and secondary archive:

  • Digitized in Broadcast Wave Format (BWF)
  • Analogue digitization uncompressed 96kHz 24-bit PCM audio
  • Digital capture at native format of 16-bit/24-bit; 44.1kHz/48kHz PCM audio

Multichannel audio format for the primary and secondary archive:

  • Digitized in Broadcast Wave Format (BWF) with .PTF (Pro Tools session)
  • Analogue digitization uncompressed 96kHz 24-bit PCM audio
  • Digital capture at native format of 16-bit/24-bit; 44.1kHz/48kHz PCM audio

Audio and Video Digitization Process Overview

[epfl:myEpflGalleryBox]