KR20040027015A - New Down-Mixing Technique to Reduce Audio Bandwidth using Immersive Audio for Streaming - Google Patents

Abstract

PURPOSE: A new down-mixing technique using the immersion typed audio to reduce an audio bandwidth for streaming is provided to simultaneously offer high quality video and surround sound by reducing a sound bandwidth on entire bandwidth and allotting more bandwidths to the video when the video is streamed in a restricted bandwidth. CONSTITUTION: After applying a direction to each channel through the respective binaural synthesis of 5 channels by using the HRFT(Head Related Transfer Function), 5 channels are down-mixed to 2 channels. Thus, a user feels stereo when the user listens to the sound through a headphone. The user feels the stereo through two speakers by making a down-mixing result pass a cross-talk filter.

Description

New Down-Mixing Technique to Reduce Audio Bandwidth using Immersive Audio for Streaming}

When streaming, all five channels are transmitted or simply down mixed to two channels for streaming. In the case of transmitting five channels, a large amount of data needs to be transmitted, so real-time transmission may not be possible, and in the case of simple downmixing with two channels, it is difficult to maintain the three-dimensionality of sound.

High-definition video such as DVD uses 5.1 channels provided by Dolby Digital, DTS, etc. to enhance the stereoscopic sound. In general, 5.1 channels have a bandwidth of 448 Kbps for Dolby Digital and 1411 Kbps for DTS. To stream high-quality DVD-quality video, the bandwidth of Dolby Digital or DTS occupies a considerable amount of bandwidth within a limited bandwidth, even on high-speed Internet such as ADSL, and relatively erodes video bandwidth. In the present invention, by maintaining the direction of each channel using the Binaural Synthesis, by downmixing the five channels into two channels, while maintaining the original three-dimensional sense of the five channels can be expected to reduce the bandwidth due to the reduction of the channel. As a result, if video is streamed in a limited bandwidth, the bandwidth of the sound at the full bandwidth is reduced and relatively more bandwidth can be allocated to the video, thereby providing high quality video and sound with stereoscopic effect.

When streaming, multi-channel such as 5.1 channel is maintained in a limited bandwidth, down-mixing to 2 channel by maintaining the three-dimensional effect to allocate more bandwidth to the video so that the video can also be streamed with the audio.

(FIG. 1) The whole process of downmixing five channels into two channels

(Figure 2) Detailed process of downmixing with Binaural Synthesis and 2-channel

In general, in order to listen to stereoscopic sound using 5.1 channels (left, right, center, surround left, surround right and low frequency effect channels), the position of speakers corresponding to 5 channels except the non-directional low frequency effect channel is important. . That is, the stereoscopic feeling is formed by the positions of the five speakers. In the present invention, each of the five channels is binaural synthesis using HRTF to give directionality to each channel, and then downmix the five channels to two channels so that a stereoscopic feeling can be felt when listening with headphones. The two-channel downmixed results are then passed through a cross-talk cancellation filter, allowing listeners to feel three-dimensional through two speakers.

In the first step, in order to perform binaural synthesis on five channels such as left, right, center, surround left, and surround right, the compressed five channels are in the form of pulse coded modulation (PCM). Unzip it. The second stage performs Binaural Synthesis on each of the five channels. Binaural synthesis is the conversion of a non-directional mono channel into a directional signal by passing it through a HRTF filter with a specific angle. The second step is shown in detail in FIG. 2. As shown in FIG. 2, in the case of the left channel, Binaural Synthesis is performed using an azimuth left 30 degrees latitude 0 degrees HRTF. In the case of the right channel, the azimuth right 30 degrees latitude 0 degrees HRTF is used to perform the Binaural Synthesis. The central channel is Binaural Synthesis using azimuth 0 degree latitude 0 degree HRTF. The surround left is Binaural Synthesis using azimuth left 120 degrees latitude 0 degrees HRTF. The surround rain is Binaural Synthesis using azimuth right 120 degrees latitude and 0 degrees HRTF. In the third step, five channels passed through Binaural Synthesis are downmixed into two channels as shown in FIG. 2. When listening to headphones, downmixing for the headphones is done by compressing the PCM signal downmixed into two channels into MP3 or the like. If you compress it to 128 Kbps using MP3 for compression, 448 Kbps will be reduced to 128 Kbps for Dolby digital.

When the listener wants to listen through two speakers, the left channel should be in the left ear and the right channel in the right ear of the two channels downmixed after Binaural Synthesis in order to feel 3D. In the case of speakers, however, a part of the left channel enters the right ear and a part of the right channel enters the left ear. This is called cross-talk. If cross-talk is present, the listener cannot feel three-dimensional effects in the Binaural Synthesis signal. When listening through headphones, the left channel is in the left ear and the right channel is in the right ear, so no crosstalk between channels occurs. However, the case of the speaker is different. In order to listen through two speakers, the channels downmixed into two channels after Binaural Synthesis must pass through a cross-talk cancellation filter, which is indicated by the dotted line in FIG. 1. In other words, when listening through two speakers, two channels through the cross-talk cancellation filter are compressed to MP3. In this way, you can enjoy a three-dimensional sound even with two speakers.

The present invention relates to a method of maintaining the stereoscopic sound of sound using headphones or two speakers while reducing the bandwidth required for audio streaming. Since it is possible to transmit a three-dimensional sound with a small bandwidth, even in a data transmission environment such as a mobile phone, you can enjoy the sound such as playing the orchestra using headphones. In addition, if the proposed technique is applied to a video such as DVD, the stereoscopic effect of the audio is maintained and the bandwidth is reduced. This means that better video can be expected with limited transmission bandwidth, and smaller bandwidth is required (more subscriber service is possible) with the same video quality.

Claims (1)

Multi-channel provided by Dolby Digital or DTS, etc.Binaural Synthesis based on HRTF and then downmixed into 2 channels for streaming