A method for head-externalization of virtual sound images of headphones

Abstract

This invention relates to a method for externalizing a virtual audio image in headphones, the method comprising: generating multiple sets of random time delays; the resulting signal is denoted as BRIR. Lt and BRIR Rt Multiple sets of random time delays are added to the BRIR signal after removing the transmission delay from the sound source to both ears. Lt and BRIR Rt The reflected sound portion, respectively obtained multiple sets of BRIR Lt.pro and BRIR Rt.pro The sound source signal was compared with each group of BRIR signals. Lt.pro and BRIR Rt.pro The method involves convolution to obtain multiple sets of binaural signals; dividing each set of binaural signals into N narrowband signals; calculating the ILD TSD of each narrowband signal group derived from each set of binaural signals; constructing an objective function E based on the ILD TSD; determining the optimized random delay based on the maximum objective function value; adding the optimized random delay to the reflected sound part of the BRIR to obtain the optimized BRIR; and convolving the sound source signal with the optimized BRIR. This method enhances the external sound imaging effect during binaural playback using headphones.

Description

Technical Field

[0001] This invention relates to the field of spatial audio technology, and in particular to a method for externalizing a virtual sound image in headphones. Background Technology

[0002] Headphone virtual sound technology aims to reproduce surround sound or stereo sound using standard stereo. To create a more immersive audio experience, a sense of space is created by designing an appropriate binaural room impulse response (BRIR). BRIR characterizes the impulse response of an electroacoustic system from a sound source in a given room to the eardrum. Generally, BRIR is divided into three parts: the first part is the direct sound component, representing the impulse response from the sound source to the eardrum in an anechoic chamber, typically lasting 5 ms or less; the second part is the early reflection component, consisting of a series of discrete reflections from walls, floors, ceilings, etc., which can be observed within milliseconds after the direct sound; the third part is the diffuse reflection component, observed approximately 80 ms after the direct sound, consisting of high-density reflections, which determines the user's overall auditory impression of the room. Under ideal BRIR measurement and headphone listening conditions, binaural audio rendered with BRIR based on physical room measurements can completely confuse the user into not being aware that they are wearing headphones. However, based on physical room BRIR measurements, even without considering individualization, it is impractical to manually measure the BRIR at all possible locations within a physical room. Therefore, in practical applications, artificially synthesized BRIRs are usually used for binaural reproduction. However, the currently used artificially synthesized BRIRs cannot accurately reproduce the spatial information required for binaural reproduction, which greatly affects the external sound image effect during binaural reproduction. Summary of the Invention

[0003] In view of this, the main objective of the present invention is to provide a method for externalizing virtual sound image in headphones, which constructs an objective function using ILD fluctuations, determines a set of random delays that maximize the objective function as optimized random delays and adds them to the BRIR reflected sound part, thereby obtaining an optimized BRIR for rendering the sound source, which enhances the external sound image effect when the headphones reproduce sound in both ears.

[0004] To achieve the above objectives, this application provides a method for externalizing a virtual audio-visual head for headphones, comprising:

[0005] Generate multiple sets of random time delays;

[0006] The signal obtained by removing the propagation delay from the sound source to both ears in the binaural impulse response (BRIR) is denoted as BRIR. Lt and BRIR Rt Multiple sets of random time delays are added to the BRIR separately. Lt and BRIR RtThe reflected sound portion was used to obtain multiple sets of signals, each set of signals denoted as BRIR. Lt.pro and BRIR Rt.pro ;

[0007] The sound source signal is compared with each group of BRIRs. Lt.pro and BRIR Rt.pro Convolution is performed to obtain multiple sets of binaural signals; each set of binaural signals is divided into frequency bands by an analysis filter bank, and each set of binaural signals is divided into N narrowband signals; the binaural sound level difference fluctuation (ILD) TSD of each set of narrowband signals is calculated for each set of binaural signals.

[0008] Construct the objective function E based on ILD TSD:

[0009]

[0010] in, The ILD TSD of the i-th narrowband signal in a set of binaural signals;

[0011] Substitute each ILD TSD into the objective function to calculate the objective function value corresponding to each group of binaural signals, and take the maximum value among the calculated objective function values;

[0012] The BRIR used when obtaining the binaural signal corresponding to the maximum value Lt.pro and BRIR Rt.pro In this process, a set of random delays is added as the optimized random delay;

[0013] An optimized random time delay is added to the reflected sound portion of the BRIR to obtain an optimized BRIR;

[0014] Convolve the sound source signal with an optimized BRIR.

[0015] In one possible implementation, multiple sets of random delays are added to the BRIR separately. Lt and BRIR Rt The reflected sound part also includes:

[0016] Extract BRIR separately Lt and BRIR Rt The reflected sound portion, denoted as BRIR. Lt,reflect and BRIR Rt.reflect ;

[0017] By analyzing the filter bank, BRIR Lt.reflect Decomposed into 24 BRIRs Lt,reflect Subband signal, BRIR Rt.reflect Decomposed into 24 BRIRs Rt.reflect Sub-band signal.

[0018] In another possible implementation, the step of adding any set of random delays to the BRIR is... Lt and BRIR Rt The reflected sound portion yields a set of signals including:

[0019] The random delay contained in each BRIR is respectively Lt.reflect The random time delay corresponding to the sub-band signal is added to each BRIR. Lt.reflect Subband signal, and will have 24 BRIRs with random time delay added. Lt.reflect The subband signal is recovered to the full-band left ear reflection portion through a synthesis filter bank, denoted as BRIR′. Lt.reflect ; splicing BRIR Lt.dir and BRIR' Lt.reflect , obtain BRIR Lt.pro ;

[0020] The random delay contained in each BRIR is respectively Rt.reflect The random time delay corresponding to the sub-band signal is added to each BRIR. Rt.reflect Subband signal, and will have 24 BRIRs with random time delay added. Rt.reflect The subband signal is recovered to the full-band right ear reflection portion through a synthesis filter bank, denoted as BRIR′. Rt.reflect ; splicing BRIR Rt.dir and BRIR' Rt.reflect , obtain BRIR Rt.pro ;

[0021] Among them, BRIR Lt.dir For BRIR Lt The direct sound section; BRIR Rt.dir For BRIR Rt The direct sound section.

[0022] In another possible implementation, dividing any set of binaural signals into N narrowband signals using an analytical filter bank specifically involves:

[0023] The left ear signal is divided into frequency bands by analyzing the filter bank, and the resulting signals are then rectified by half-wave and low-pass filtered to obtain N left ear narrowband signals.

[0024] The right ear signal is divided into frequency bands by analyzing the filter bank, and the resulting signals are then rectified by half-wave and low-pass filtered to obtain N right ear narrowband signals.

[0025] By taking any corresponding set of left and right ear narrowband signals as a set of narrowband signals, N sets of narrowband signals are obtained.

[0026] In another possible implementation, calculating the ILD TSD of any set of narrowband signals includes:

[0027] Calculate the envelopes of the narrowband signals for the left and right ears of this set of narrowband signals respectively;

[0028] Subtract the calculated narrowband signal envelopes of the left and right ears to obtain the ILD of the narrowband signals of the left and right ears;

[0029] The standard deviation of the ILD of the narrowband signals of the left and right ears was calculated and used as the ILD TSD of the narrowband signals.

[0030] In another possible implementation, the optimized BRIR includes:

[0031] Extract the reflected sound components of the BRIR from the left and right ears respectively, and denot them as BRIR. L,reflect and BRIR R.reflect ;

[0032] By analyzing the filter bank, BRIR L.reflect Decomposed into 24 BRIRs L,reflect Subband signal, BRIR R.reflect Decomposed into 24 BRIRs R.reflect Sub-band signal;

[0033] Incorporate optimized random delays into each BRIR L.reflect The random time delay corresponding to the sub-band signal is added to each BRIR. L.reflect Subband signal, and will have 24 BRIRs with random time delay added. L.reflect The subband signal is recovered to the full-band left ear reflection portion through a synthesis filter bank, denoted as BRIR′. L.reflect ; splicing BRIR L.dir and BRIR' L.reflect , obtain BRIR L.pro ;

[0034] Incorporate optimized random delays into each BRIR Rt.reflect The random time delay corresponding to the sub-band signal is added to each BRIR. R.reflect Subband signal, and will have 24 BRIRs with random time delay added. R.reflect The subband signal is recovered to the full-band right ear reflection portion through a synthesis filter bank, denoted as BRIR′. R.reflect ; splicing BRIR R.dir and BRIR' R.reflect , obtain BRIR R.pro ;

[0035] Among them, BRIR L.dir This refers to the direct sound portion of the left ear's BRIR; BRIR R.dir This is the direct sound portion of the BRIR signal in the right ear.

[0036] In another possible implementation, the sound source signal is compared with each group of BRIRs. Lt.pro and BRIR Rt.pro Before convolution, it also includes:

[0037] For each group of BRIR Lt.pro and BRIR Rt.pro Reflection suppression is applied to the reflected sound portion 10ms after the direct sound. Attached Figure Description

[0038] Figure 1 This is a flowchart illustrating a method for externalizing a virtual audio-visual head in an earphone according to an embodiment of the present invention;

[0039] Figure 2 This is a schematic diagram illustrating the maximum delay limit of the critical frequency band. Detailed Implementation

[0040] This application refers to the early reflected sound portion and diffuse reflected sound portion of a BRIR as the reflected sound portion. The reflections that make up the reflected sound portion of a BRIR reach the human ear at different times. When artificially synthesizing a BRIR, random time delays are added to each reflection of the reflected sound portion of the BRIR to simulate the delay time of the reflection reaching the human ear relative to the direct sound.

[0041] The BRIR includes binaural cues related to external sound imaging effects, such as binaural correlation (IC) and interaural level differences (ILD). The applicant has found that ILD fluctuation (ILD Temporal Standard Deviation), i.e., the fluctuation of ILD over time, is a good indicator of external sound imaging effects; high ILD fluctuation corresponds to good virtual sound imaging externalization. Therefore, this invention constructs an objective function using ILD fluctuation, generates multiple sets of random time delays, and determines the set of random time delays that maximizes the objective function as the optimized random time delay added to the reflected sound part of the BRIR. This results in an optimized BRIR used to render the sound source, thereby enhancing the external sound imaging effect during binaural playback.

[0042] Specifically, the process of a method for externalizing a virtual audio image in an earphone according to an embodiment of the present invention is as follows: Figure 1 As shown, steps 101 to 108 are included.

[0043] Step 101: Generate multiple sets of random delays.

[0044] Step 102: Remove the propagation delay from the sound source to both ears in the binaural impulse response (BRIR) signal, and denote the resulting signal as BRIR. Lt and BRIR Rt Multiple sets of random time delays are added to the BRIR separately. Ltand BRIR Rt The reflected sound portion was used to obtain multiple sets of signals, each set of signals denoted as BRIR. Lt.pro and BRIR Rt.pro .

[0045] Step 103: Match the sound source signal with each group of BRIR signals. Lt.pro and BRIR Rt.pro Convolution is performed to obtain multiple sets of binaural signals; each set of binaural signals is divided into frequency bands by an analysis filter bank, and each set of binaural signals is divided into N narrowband signals; the binaural sound level difference fluctuation (ILD) TSD of each set of narrowband signals is calculated.

[0046] Step 104: Construct the objective function E based on ILD TSD:

[0047]

[0048] in, ILD TSD is the i-th narrowband signal of a set of binaural signals.

[0049] Step 105: Substitute each ILD TSD into the objective function to calculate the objective function value corresponding to each group of binaural signals, and take the maximum value among the calculated objective function values.

[0050] Step 106: The BRIR used when obtaining the binaural signal corresponding to the maximum value. Lt.pro and BRIR Rt.pro In this process, a set of random delays is added as the random delay for optimization.

[0051] Step 107: Add the optimized random time delay to the reflected sound part of the BRIR to obtain the optimized BRIR.

[0052] Step 108: Convolve the sound source signal with the optimized BRIR.

[0053] Here, in step 101, multiple sets of random delays can be generated by a random delay generator, and the multiple sets can be 500 sets.

[0054] Each group of random delays includes items corresponding to BRIR. Lt,reflect Subband signal and BRIR Rt,reflect The sub-band signal has a total of 48 random time delays, and the random time delay corresponding to each sub-band signal does not exceed the maximum delay limit of that sub-band signal to prevent audible artifacts. The sub-band is a critical frequency band, BRIR. Lt,reflect Subband signal and BRIR Rt,reflect The maximum delay limit for subband signals is as follows: Figure 2 As shown, one frequency band number corresponds to one sub-band signal.

[0055] In one possible implementation, in step 102, the addition of multiple sets of random time delays to the BRIR is... Lt and BRIR Rt The reflected sound part also includes:

[0056] Extract BRIR separately Lt and BRIR Rt The reflected sound portion, denoted as BRIR. Lt,reflect and BRIR Rt.reflect ;

[0057] By analyzing the filter bank, BRIR Lt.reflect Decomposed into 24 BRIRs Lt,reflect Subband signal, BRIR Rt.reflect Decomposed into 24 BRIRs Rt.reflect Sub-band signal.

[0058] Here, the analysis filter bank is implemented using the Gammatone filter bank.

[0059] Accordingly, in step 102, the step of adding any set of random time delays to the BRIR is... Lt and BRIR Rt The reflected sound portion yields a set of signals including:

[0060] The random delay contained in each BRIR is respectively Lt.reflect The random time delay corresponding to the sub-band signal is added to each BRIR. Lt.reflect Subband signal, and will have 24 BRIRs with random time delay added. Lt.reflect The subband signal is recovered to the full-band left ear reflection portion through a synthesis filter bank, denoted as BRIR′. Lt.reflect ; splicing BRIR Lt.dir and BRIR' Lt.reflect , obtain BRIR Lt.pro ;

[0061] The random delay contained in each BRIR is respectively Rt.reflect The random time delay corresponding to the sub-band signal is added to each BRIR. Rt.reflect Subband signal, and will have 24 BRIRs with random time delay added. Rt.reflect The subband signal is recovered to the full-band right ear reflection portion through a synthesis filter bank, denoted as BRIR′. Rt.reflect ; splicing BRIR Rt.dir and BRIR' Rt.reflect , obtain BRIR Rt.pro ;

[0062] Among them, BRIR Lt.dir For BRIRLt The direct sound portion, i.e., the portion of the left ear BRIR after removing the transmission delay from the sound source to the left ear; BRIR Rt.dir For BRIR Rt The direct sound portion, i.e., the direct sound portion of the right ear BRIR after removing the transmission delay from the sound source to the right ear.

[0063] Here, with 500 sets of random delays, we can correspondingly obtain 500 sets of BRIRs. Lt.pro and BRIR Rt.pro .

[0064] The synthesized filter bank is implemented using a Gammatone filter bank.

[0065] In another possible implementation, in step 103, any set of binaural signals is divided into N narrowband signals by an analysis filter bank, specifically as follows:

[0066] The left ear signal is divided into frequency bands by analyzing the filter bank, and the resulting signals are then rectified by half-wave and low-pass filtered to obtain N left ear narrowband signals.

[0067] The right ear signal is divided into frequency bands by analyzing the filter bank, and the resulting signals are then rectified by half-wave and low-pass filtered to obtain N right ear narrowband signals.

[0068] By taking any corresponding set of left and right ear narrowband signals as a set of narrowband signals, N sets of narrowband signals are obtained.

[0069] Here, the analysis filter bank is implemented using a Gammatone filter bank. The analysis filter bank divides the frequency band using either the ERB or bark scale, with N = 24.

[0070] The process of half-wave rectification and low-pass filtering of the resulting signals can simulate the peripheral hearing of humans.

[0071] In another possible implementation, step 103, specifically calculating the ILD TSD of any set of narrowband signals, includes:

[0072] Calculate the envelopes of the narrowband signals for the left and right ears of this set of narrowband signals respectively;

[0073] Subtract the calculated narrowband signal envelopes of the left and right ears to obtain the ILD of the narrowband signals of the left and right ears;

[0074] The standard deviation of the ILD of the narrowband signals of the left and right ears was calculated and used as the ILD TSD of the narrowband signals.

[0075] Here, the envelopes of the left and right ear narrowband signals of the i-th group of narrowband signals can be represented as:

[0076] env i,L (n) = 20log(|w i,L (n)|)

[0077] env i,R (n) = 20log(|w i,R (n)|)

[0078] Among them, w i,L (n), w i,R (n) represent the nth sampled signals of the left and right ear narrowband signals of the i-th group of narrowband signals, respectively. i,L (n), env i,R (n) are the envelopes of the nth sampled signals of the left and right ear narrowband signals of the i-th group of narrowband signals, respectively.

[0079] Correspondingly, the ILD of the i-th narrowband signal is represented as:

[0080] ILD i (n) = env i,L (n)-env i,R (n)

[0081] Correspondingly, the ILD TSD of the i-th narrowband signal is expressed as:

[0082]

[0083] in, Let L be the ILD TSD of the i-th narrowband signal; L is the length of the narrowband signal in the left and right ears. Let be the average ILD of the i-th narrowband signal group.

[0084] In another possible implementation, the optimized BRIR obtained in step 107 specifically includes:

[0085] Extract the reflected sound components of the BRIR from the left and right ears respectively, and denot them as BRIR. L,reflect and BRIR R.reflect ;

[0086] By analyzing the filter bank, BRIR L.reflect Decomposed into 24 BRIRs L,reflect Subband signal, BRIR R.reflect Decomposed into 24 BRIRs R.reflect Sub-band signal;

[0087] Incorporate optimized random delays into each BRIR L.reflect The random time delay corresponding to the sub-band signal is added to each BRIR. L.reflect Subband signal, and will have 24 BRIRs with random time delay added.L.reflect The subband signal is recovered to the full-band left ear reflection portion through a synthesis filter bank, denoted as BRIR′. L.reflect ; splicing BRIR L.dir and BRIR' L.reflect , obtain BRIR L.pro ;

[0088] Incorporate optimized random delays into each BRIR Rt.reflect The random time delay corresponding to the sub-band signal is added to each BRIR. R.reflect Subband signal, and will have 24 BRIRs with random time delay added. R.reflect The subband signal is recovered to the full-band right ear reflection portion through a synthesis filter bank, denoted as BRIR′. R.reflect ; splicing BRIR R.dir and BRIR' R.reflect , obtain BRIR R.pro ;

[0089] Among them, BRIR L.dir This refers to the direct sound portion of the left ear's BRIR; BRIR R.dir This is the direct sound portion of the BRIR signal in the right ear.

[0090] Here, both the analysis filter bank and the synthesis filter bank are implemented using the Gammatone filter bank.

[0091] In another possible implementation, in step 108, the sound source signal is compared with each group of BRIRs. Lt.pro and BRIR Rt.pro Before convolution, the following steps are also included: processing each group of BRIRs separately. Lt.pro and BRIR Rt.pro Reflection suppression is applied to the reflected sound portion 10ms after the direct sound.

[0092] Here, for any BRIR Lt.pro The specific steps for suppressing reflections are as follows:

[0093] The BRIR Lt.pro Multiply by a time window, which has a value of 1 from 0ms to 2.5ms and a value of 0 until 10ms, and implements a transition from 0 to 1 from 10ms to 15ms.

[0094] For any BRIR Rt.pro The specific steps for suppressing reflections are as follows:

[0095] The BRIR Rt.pro Multiply by a time window, which has a value of 1 from 0ms to 2.5ms and a value of 0 until 10ms, and implements a transition from 0 to 1 from 10ms to 15ms.

[0096] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention.

Claims

1. A method for externalizing a virtual sound image in headphones, characterized in that, include: Generate multiple sets of random time delays; The signal obtained by removing the propagation delay from the sound source to both ears in the binaural impulse response (BRIR) is denoted as BRIR. Lt and BRIR Rt Multiple sets of random time delays are added to the BRIR separately. Lt and BRIR Rt The reflected sound portion was used to obtain multiple sets of signals, each set of signals denoted as BRIR. Lt.pro and BRIR Rt.pro ; The sound source signal is compared with each group of BRIRs. Lt.pro and BRIR Rt.pro Convolution is performed to obtain multiple sets of binaural signals; each set of binaural signals is then divided into frequency bands by an analysis filter bank, and each set of binaural signals is divided into N narrowband signals. Calculate the binaural sound level difference fluctuation (ILD TSD) of each narrowband signal group divided from each group of binaural signals; Construct the objective function E based on ILD TSD: ； in, The ILD TSD of the i-th narrowband signal in a set of binaural signals; Substitute each ILD TSD into the objective function to calculate the objective function value corresponding to each group of binaural signals, and take the maximum value among the calculated objective function values; The BRIR used when obtaining the binaural signal corresponding to the maximum value Lt.pro and BRIR Rt.pro In this process, a set of random delays is added as the optimized random delay; The optimized random time delay is added to the reflected sound part of the BRIR to obtain the optimized BRIR; Convolve the sound source signal with an optimized BRIR signal; The step involves adding multiple sets of random time delays to the BRIR. Lt and BRIR Rt The reflected sound part also includes: Extract BRIR separately Lt and BRIR Rt The reflected sound portion, denoted as BRIR. Lt,reflect and BRIR Rt.reflect ; By analyzing the filter bank, BRIR Lt.reflect Decomposed into 24 BRIRs Lt,reflect Subband signal, BRIR Rt.reflect Decomposed into 24 BRIRs Rt.reflect Sub-band signal.

2. The method according to claim 1, characterized in that, Add a set of random time delays to BRIR Lt and BRIR Rt The reflected sound portion yields a set of signals including: The random delay contained in each BRIR is respectively Lt.reflect The random time delay corresponding to the sub-band signal is added to each BRIR. Lt.reflect Subband signal, and will have 24 BRIRs with random time delay added. Lt.reflect The subband signal is recovered to the full-band left ear reflection portion through a synthesis filter bank, denoted as... ; splicing BRIR Lt.dir and , obtain BRIR Lt.pro ; The random delay contained in each BRIR is respectively Rt.reflect The random time delay corresponding to the sub-band signal is added to each BRIR. Rt.reflect Subband signal, and will have 24 BRIRs with random time delay added. Rt.reflect The subband signal is recovered to the full-band right ear reflection portion through a synthesis filter bank, denoted as... ; splicing BRIR Rt.dir and , obtain BRIR Rt.pro ; Among them, BRIR Lt.dir For BRIR Lt The direct sound section; BRIR Rt.dir For BRIR Rt The direct sound section.

3. The method according to claim 1, characterized in that, A set of binaural signals is divided into N narrowband signals using an analytical filter bank, specifically: The left ear signal is divided into frequency bands by analyzing the filter bank, and the resulting signals are then rectified by half-wave and low-pass filtered to obtain N narrowband signals for the left ear. The right ear signal is divided into frequency bands by analyzing the filter bank, and the resulting signals are then rectified by half-wave and low-pass filtered to obtain N right ear narrowband signals. By taking any corresponding set of left and right ear narrowband signals as a set of narrowband signals, N sets of narrowband signals are obtained.

4. The method according to claim 3, characterized in that, Calculating the ILD TSD of a set of narrowband signals includes: Calculate the envelopes of the narrowband signals for the left and right ears of this set of narrowband signals respectively; Subtract the calculated narrowband signal envelopes of the left and right ears to obtain the ILD of the narrowband signals of the left and right ears; The standard deviation of the ILD of the narrowband signals of the left and right ears was calculated and used as the ILD TSD of the narrowband signals.

5. The method according to claim 1, characterized in that, The optimized BRIR includes: Extract the reflected sound components of the BRIR from the left and right ears respectively, and denot them as BRIR. L,reflect and BRIR R.reflect ; By analyzing the filter bank, BRIR L.reflect Decomposed into 24 BRIRs L,reflect Subband signal, BRIR R.reflect Decomposed into 24 BRIRs R.reflect Sub-band signal; Incorporate optimized random delays into each BRIR L.reflect The random time delay corresponding to the sub-band signal is added to each BRIR. L.reflect Subband signal, and will have 24 BRIRs with random time delay added. L.reflect The subband signal is recovered to the full-band left ear reflection portion through a synthesis filter bank, denoted as... ; splicing BRIR L.dir and , obtain BRIR L.pro ; Incorporate optimized random delays into each BRIR Rt.reflect The random time delay corresponding to the sub-band signal is added to each BRIR. R.reflect Subband signal, and will have 24 BRIRs with random time delay added. R.reflect The subband signal is recovered to the full-band right ear reflection portion through a synthesis filter bank, denoted as... ; splicing BRIR R.dir and , obtain BRIR R.pro ; Among them, BRIR L.dir This refers to the direct sound portion of the left ear's BRIR; BRIR R.dir This is the direct sound portion of the BRIR signal in the right ear.

6. The method according to claim 1, characterized in that, The sound source signal is compared with each group of BRIRs. Lt.pro and BRIR Rt.pro Before convolution, it also includes: For each group of BRIR Lt.pro and BRIR Rt.pro Reflection suppression is applied to the reflected sound portion 10ms after the direct sound.

Images