/R3VIVAL

A Repository of Room Responses and 360 Videos of a Variable Acoustics Lab

Primary LanguageHTMLCreative Commons Attribution 4.0 InternationalCC-BY-4.0

LabOverview

R3VIVAL - Repository of Room Responses and 360 Videos of a Variable Acoustics Lab

SUMMARY OF THE DATA

This datset contains Spatial Room Impulse Responses (SRIR) and 360 stereoscopic video recordings of a variable acoustics lab. The dataset contains:

  • Spatial Room Impulse Responses (total of 272 SRIRs).

    • 30 source positions arranged in 3 concentric circles (1.5m, 2m, 3m) every 30 degrees.
      • The largest circle contains only 6 positions due to room dimensions.
      • Directional sound source (Genelec 8320).
    • 4 source positions next to room corners.
      • Omnidirectional source (B&K 4295).
    • 1 receiver position (7 microphone open array from 1).
    • 8 acoustic panel configuration.
      • 5 uniform conditions (0%, 25%, 50%, 75%, 100% absorption).
      • 3 asymmetric configurations (right, right/back, right/back/left absorptive).

  • Video captures

    • Stereoscopic 360 videos (resolution 4096 x 2160).
    • Captured with LivePlanetVR camera.
    • 2 rooms (Lab and RoomA - shoebox room).
    • 4 seconds duration.
    • Static scenes showing the empty room.

drawing

VARIABLE ACOUSTICS LAB

drawing drawing drawing

  • Shoebox room
  • Dimensions: 9.7 × 5.5 × 2.7 m (L × W × H)
  • RT60 at mid frequencies (500-1000 Hz) ranging from ~0.5 to 0.75 seconds.

drawing

Microphone array

  • 7 Microphone array
    • Earthworks M50 (centre)
    • 6x DPA 4060
    • 10 cm diameter

drawing

Channel Position Azimuth Elevation Microphone
1 Front 0 0 DPA 4060
2 Back 180 0 DPA 4060
3 Right Top 90 45 DPA 4060
4 Left Top 90 45 DPA 4060
5 Right Bottom 270 -45 DPA 4060
6 Left Bottom 270 -45 DPA 4060
7 Center N/A N/A Earthworks M50

360 Stereoscopic Videos

LabOverview LabOverview

How to use this data? (Work in Progress)

  • Render Binaural Room Impulse Responses

    • Set up the Binaural SDM Toolbox and its dependencies2
    • Run the example file

  • Generate video for a specific loudspeaker arrangement

  • Combine the data with the Audiovisual Speech Corpus3 to generate synthetic multi-talker scenes. Note that the link on the original publication is no longer available. A mirror source is provided in4

LabOverview

Contributing

See the CONTRIBUTING file for how to help out.

License

R3VIVAL is CC-BY-4.0 licensed, as found in the LICENSE file.

Contact

Sebastia V. Amengual (samengual@meta.com)

References

Footnotes

  1. Amengual Garí, S. V., Arend, J. M., Calamia, P. T., & Robinson, P. W. (2021). Optimizations of the spatial decomposition method for binaural reproduction. Journal of the Audio Engineering Society, 68(12), 959-976. https://doi.org/10.17743/jaes.2020.0063 ↩

  2. https://github.com/facebookresearch/BinauralSDM/ ↩

  3. Kishline, L. R., Colburn, S. W., & Robinson, P. W. (2020). A multimedia speech corpus for audio visual research in virtual reality (L). The Journal of the Acoustical Society of America, 148(2), 492-495. https://doi.org/10.1121/10.0001670 ↩

  4. https://drive.google.com/drive/u/1/folders/1E_sC5SfjZCJOnkLkIN7MJ38-bwkAHQkY ↩