Voice transfer demo

[中文]

1、Function overview

The two-way voice call between the board and the mobile phone apk is realized through wifi.

2、Code path

demo path: \components\demos\media\doorbell\src\doorbell.c

For the detailed description of the Aud_Intf API interface, please refer to the same webpage: /api-reference/multi_media/bk_aud_intf.html

3、Cli command introduction

The commands supported by the demo are as follows:

Command	Description
video_transfer -a test 12345678	Initialize the voice channel

The macro configuration that the demo runs depends on:

Name	Description	File	value
CONFIG_AUDIO	Configure the audio function	\middleware\soc\bk7237\bk7237.defconfig	y
CONFIG_AUDIO_RISCV_IP_V1_0	Configure audio ip	\middleware\soc\bk7237\bk7237.defconfig	y
CONFIG_AUD_INTF	Configure aud_intf enable	\middleware\soc\bk7237\bk7237.defconfig	y
CONFIG_AUD_INTF_TEST	Configure demo enable	\middleware\soc\bk7237\bk7237.defconfig	y
CONFIG_AUDIO_TRANSFER	Configure voice transmission enable	\middleware\soc\bk7237\bk7237.defconfig	y
CONFIG_AUD_TRAS_MODE_CPU0	Configure Voice Transmission Mode	\middleware\soc\bk7237\bk7237.defconfig	y

Voice call debugging macro configuration:

Name	Description	File
CONFIG_AUD_TRAS_AEC_DUMP_DEBUG	Configure aec dump	\middleware\soc\bk7237\bk7237.defconfig
CONFIG_AUD_TRAS_LOST_COUNT_DEBUG	Configuring Tx Packet Loss Statistics	\middleware\soc\bk7237\bk7237.defconfig
CONFIG_AUD_TRAS_AEC_MIC_DELAY_DEBUG	Configure mic delay debug	\middleware\soc\bk7237\bk7237.defconfig
CONFIG_AUD_TRAS_AEC_MIC_DELAY_POINTS	Configure mic delay points	\middleware\soc\bk7237\bk7237.defconfig

Demo runs dependent libraries and drivers:

• DMA DMA driver

• AUD audio module driver

• Sdcard module driver

• WIFI module

4、Demo introduction

The steps performed by the demo are as follows:

1.Open the audio streaming app

• Uart sends AT command video_transfer -a test 12345678 to initialize the voice call application

2.establish a wifi connection

• The mobile phone wifi is connected to the route of the test name, the password is: 12345678

3.setup mobile apk

• After the connection is successful, open the video transfer app, and set Settings->Demos Solution to Video transfer mode.

4.Start a voice call

• Click the voice call button on the apk interface to start a voice call

5.stop voice call

• Click the voice call button on the apk interface again to stop the voice call

5、Voice call function adaptation instructions

The debugging process of the voice call function is as follows:

1. Functional debugging

   • 1、Turn off all debug macros

   • 2、Refer to the sample code of the bk_aud_intf_voc_init interface, call this interface to initialize the voice call;

   • 3、Call the bk_aud_intf_write_spk_data interface to write the voice data sent by the apk to the board for playback;

2. mic delay point debugging

   • 1、Set the debug macros CONFIG_AUD_TRAS_AEC_DUMP_DEBUG and CONFIG_AUD_TRAS_AEC_MIC_DELAY_DEBUG to y, compile bin for testing, talk for a minute, and dump the data of the AEC algorithm to the tf card;

   • 2、According to the dump data, calculate the delay points of mic, and set the point value to the value of the macro CONFIG_AUD_TRAS_AEC_MIC_DELAY_POINTS (normally, the default value of 55 can be used);

3. aec echo cancellation effect debugging

   • 1、open voice call;

   • 2、According to the echo effect heard, use Realtime Video apk (downloadable from the open source armino platform on github) to debug the parameters of AEC and NS online, and record the final optimal value;

   • 3、First adjust echo depth , the larger the echo, the larger the value will be. When you continue to increase the value, but the echo cancellation effect cannot be improved, stop setting the value;

   • 4、Secondly, set the value of max amplitude and min amplitude according to the size and range of the echo to optimize the echo cancellation effect;

   • 5、Then set the noise level according to the size of the noise floor you hear. The larger the noise floor, the larger the set value, to optimize the noise floor removal effect;

   • 4、Finally, set the value of noise param, set to 0, 1, 2 respectively, and choose the best value;

   • 5、Record the online debugging results of all parameters and set them as the default parameters when the voice is initialized.

Note

• 1.You can also use AT command to debug online, send serial command aud_intf_set_aec_param_test {param value} through uart to set the value of each parameter. The specific debugging process is the same as the above method. For details, please refer to the example in \components\demos\media\aud\aud_intf\demo\aud_intf_demo.c.

• 2.The AEC online debugging parameter setting interface is shown in the figure below.

Figure 1. AEC parameter setting entry interface

Figure 1. AEC parameter setting interface

3. Turn off all debug macros

   • Close all debug macros when debugging work is done

Note

• 1.The current voice call is based on a single mic scene, 8K sampling rate;