School of Electrical and Electronic Engineering
Quick and convenient
Allow unconstrained continuous head movement with direction recorded automatically.
Accurate HRTF
Measured HRTFs with low error and consistent with conventional static method
Fast HRTF Measurement System with Unconstrained Head Movements for 3D Audio in Virtual and Augmented Reality Applications Nguyen Duy Hai, Nitesh Kumar Chaudhary, Santi Peksi, Rishabh Ranjan, Jianjun He, and Woon-Seng Gan
Immersive/Interactive 3D sound
Personalized HRTFs create the most immersive and interactive 3D sound used in multimedia, VR/AR applications
Well suited for VR/AR
Measured HRTFs reflect natural head/torso movements in VR/AR applications
Dr. Gan Woon Seng
[email protected] http://www3.ntu.edu.sg/home/ewsgan/ Digital Signal Processing Laboratory School of Electrical and Electronic Engineering Nanyang Technological University 50 Nanyang Avenue, Singapore 639798
Simple setup
Binaural microphone, Loud speaker, Recorder, Head tracker or VR/AR headset
High resolution HRTF
The proposed system can obtain HRTF at a very fine resolution down to every 1 degree.
Introduction • Head-related transfer function (HRTF) is essential to realize an immersive listening experience over headphones in multimedia, virtual and augmented reality applications. • To extract the HRTFs from dynamic binaural measurements with random head movements, an improved adaptive filtering algorithm is proposed by integrating direction quantization, variable step size, and optimal HRTF selection using the progressive based normalized leastmean-squares algorithm.
Results
Static vs Dynamic comparison
Specification
Excitation signal x(n) yaw
Direction signal {θ(n),φ(n)}
h[θ(n), φ(n)]
-
pitch
Binaural signal y(n)
HRIR h(θ,φ)
+
e(n)
• • • • • •
MOTU UltraLite-mk3 Hybrid soundcard Sound-professional Binaural microphone (20-20kHz) Genelec speaker (8320A) Sampling rate:48kHz Quantization & Grid resolution for rendering: 5x10 degree (Azimuth & elevation) Interpolation for finer resolutions and roll variations limit : ±5 degrees
Adaptive signal processing
System overview of the proposed fast, continuous and
Systemunconstrained overview of theHRTF proposed fast, continuous and measurement system. unconstrained HRTF measurement system.
Subjective Performance ABX test further confirmed our objective results. 95% identification accuracy for Generic vs Personalized whereas 50-70% identification accuracy for Static vs Dynamic, which indicates that our measured HRTF is almost indifferentiable from conventional static methods.
Static vs Dynamic comparison
Subjective Evaluation ABX Effect of head and torso movement
1- Easily differentiate 0.5 - Guessing level
Effect of Rift
References 1. 2.
3. 4. 5.
6.
J. He, R. Ranjan, and W. S. Gan, “Fast continuous HRTF acquisition with unconstrained movements of human subjects,” in Proc. ICASSP, Shanghai, China, Mar. 2016, pp. 321-325. R. Ranjan, J. He, and W. S. Gan, “Fast Continuous Acquisition of HRTF for Human Subjects with Unconstrained Random Head Movements in Azimuth and Elevation ” AES Conference on Headphone Technology, Aalborg, Denmark, 2016 Aug 24–26. K. Sunder, J. He, E. L. Tan, and W. S. Gan, “Natural sound rendering for headphones: Integration of signal processing techniques,” IEEE Sig. Process. Mag., vol. 32, no. 2, Mar 2015, pp. 100-113. K. Sunder, E. L. Tan, and W. S. Gan, "Individualization of binaural synthesis using frontal projection headphones," J. Audio Eng. Soc., vol. 61, no. 12, pp. 989-1000, Dec. 2013. J. He, E. L. Tan, and W. S. Gan, “Primary-ambient extraction using ambient spectrum estimation for immersive spatial audio reproduction,” IEEE/ACM Trans. Audio, Speech, Lang. Process., vol. 23, no. 9, pp. 1430-1443, Sept. 2015. Ranjan, Rishabh, and Woon-Seng Gan. "Natural listening over headphones in augmented reality using adaptive filtering techniques." IEEE/ACM Transactions on Audio, Speech and Language Processing (TASLP) 23.11 (2015): 1988-2002.
