top of page
PingKang Chen's Blog
Project Name: 5.1 Virtual Surround Sound Representation over Headphone
1: Introduction of 5.1 Surround Sound System
5.1 surround sound has 6 sound channels, and so 6 speakers. It includes a center speaker, subwoofer (for low frequency effects, such as explosions), left and right front speakers, and left and right surround speakers surrounding the listener, creating a 360-degree soundstage on a horizontal plane, which can bring more sense of immersion.
2: Technical details
Note: I discard the subwoofer channel, which is the 0.1 channel also called the LFE or Low Frequency Effects channel. Because this channel carries low frequency information that is of little use to stereo or mono speakers, Besides, the LFE channel bandwidth is limited to 120Hz, which is a bit too low for most headphones to reproduce accurately.
The basic method to reproduce this effect is to convolve the BRIRs (Binaural Room Impulse Responses) captured from 5 angles corresponding the 5.1 surround sound playback system with the input signal.
1: The BRIRs were recorded using head and torso simulators in a variety of reverberant rooms with distinct acoustical properties. There are many publicly available BRIR datasets out there, stored in formats such as SOFA, audio, or mat files. I sourced the BRIRs used in this project from the following repository which all have been equalized to remove undesired spectral colouration and to make the BRIRs compatible with diffuse-field equalized headphones : https://github.com/ShanonPearce/ASH-IR-Dataset/tree/master/BRIRs.
Center (0 Deg.):
00:00 / 00:01
Surround Right (110 Deg.):
00:00 / 00:01
Front Left (-30 Deg.):
00:00 / 00:01
Front Right (30 Deg.):
00:00 / 00:01
Surround Left (-110 Deg.):
00:00 / 00:01
2: Since the Binaural Room Impulse Response (BRIR) is the binaural signal with two channels (left and right), we need to convolve each channel of it separately with the input signal.
-
Right ear signal = Input Signal * Right Ear BRIR
-
Left ear signal = Input Signal * Left Ear BRIR
3: Following the convolution, there will be 10 channels in the end – 5 for the left ear and 5 for the right ear. So, sum 5 left ear channels to one left channel, and 5 right ear channels to one right channel.
4: Combine the left and right channel into stereo signal, followed by final normalization.
3: Final Output Demonstration
Original input signal(vocal):
00:00 / 00:09
Output Signal:
00:00 / 00:09
4: Download the Code
click here: From Github!
bottom of page