Metadata
Creators:
- Pausch, Florian
, Acoustics Research Institute
- Feldner, Flora
, Acoustics Research Institute
- Mihocic, Michael
, Acoustics Research Institute
- Majdak, Piotr
, Acoustics Research Institute
- Lucchi, Michele , USound GmbH
- Neacsu, Adina , USound GmbH
Publishers:
Rightsholders:
Keywords:
- Multi-driver multi-receiver headphones
- Head-related transfer functions (HRTFs)
- Pinna reflections
- Adult anthropometrics
- Machine learning
- Acoustically measured
Relations:
- Selfone was created with the Tool Selfone (2022-2024).
Other:
- DOI: not assigned yet
- Uploaded by: Florian Pausch
- Date (created): 2025-09-25 11:15:46 (GMT)
- Date (updated): 2025-12-03 11:33:14 (GMT)
- Production Year: 2025
- Resource Type: Dataset (SONICOM Ecosystem)
- Rights: CC BY-SA 4.0 Attribution-ShareAlike
- Subject Areas: Life Science , Other SONICOM Ecosystem
- General Description: Selfone is a headphone prototype with 20 loudspeakers and 24 microphones and serves to investigate methods for personalized audio reproduction via headphones. This database includes the Selfone impulse responses (SIRs) reflected at the listener's left ear and additional impulse responses measured from a microphone inserted into the listener's left ear canal. Combined with HRTFs measured from the same listener, this database facilitates future work on algorithms that aim to estimate personalized HRTFs from SIRs.
- Methods: Selfone impulse responses (SIRs) were measured with Selfone between spatially distributed MEMS loudspeakers to an in-ear microphone and to spatially distributed electret microphones. A 20881-ms exponential frequency sweep with a frequency range of 500 to 38000 Hz and a sampling frequency of 96 kHz was used. Impulse responses were time windowed to have a length of 256 samples. Eight measurements were conducted for different wearing positions of Selfone. HRTFs were measured with in-the-ear microphones and blocked-ear canal method. The participant was seated in the center of a loudspeaker sphere and wore in-ear microphones. A 6000-ms exponential frequency sweep (20 to 20000 Hz) at a sampling frequency of 48 kHz was used. Raw data were equalized with the loudspeaker-microphone reference measurement in the frequency range of 20 to 18000 Hz, then temporally windowed. The SIRs were measured at the left ear only.
- Technical Remarks: For SIRs; R=1: In-ear microphone; R>1: Selfone microphones.
For HRTFs; R=1: Left-ear microphone (the same as used for measuring SIRs); R=2: Right-ear microphone (not available in SIR datasets). - Data Source: Acoustically measured
Comments
No comments found.
Uploaded by: Florian Pausch
Created: 2025-09-25 11:15:46
Updated: 2025-12-03 11:33:14