- 问题背景
晓龙相机录像的时候，回播视频发现外放声音非常小，但如果用耳机来听则正常。
- 产品和参考设计MIC的差别
- 录音时抓log和分析
录像的时候用logcat | grep snd_device抓log，如下
01-03 06:15:41.267 704 1035 D audio_hw_primary: select_devices: out_snd_device(2: speaker)in_snd_device(0: )
01-03 06:15:41.267 704 1035 D msm8916_platform: platform_can_split_snd_device: snd_device(2)num devices(1) new_snd_devices(5)
01-03 06:15:41.267 704 1035 I msm8916_platform: platform_check_and_set_codec_backend_cfg: becf:new_snd_devices[0] is speaker
01-03 06:15:41.267 704 1035 E msm8916_platform: platform_check_backends_match: Invalidsnd_device =
01-03 06:15:41.267 704 1035 D msm8916_platform: platform_can_split_snd_device: snd_device(2)num devices(0) new_snd_devices(0)
01-03 06:15:41.267 704 1035 D audio_hw_primary: enable_snd_device: snd_device(2: speaker)
01-03 06:15:41.309 704 1035 D msm8916_platform: platform_can_split_snd_device: snd_device(2)num devices(1) new_snd_devices(-299367304)
01-03 06:15:44.810 704 1035 D audio_hw_primary: disable_snd_device: snd_device(2: speaker)
01-03 06:15:44.810 704 1035 D msm8916_platform: platform_can_split_snd_device: snd_device(2)num devices(0) new_snd_devices(0)
01-03 06:15:53.127 704 1035 D audio_hw_primary: select_devices: out_snd_device(2: speaker)in_snd_device(0: )
01-03 06:15:53.127 704 1035 D msm8916_platform: platform_can_split_snd_device: snd_device(2)num devices(1) new_snd_devices(6)
01-03 06:15:53.127 704 1035 I msm8916_platform: platform_check_and_set_codec_backend_cfg: becf:new_snd_devices[0] is speaker
01-03 06:15:53.127 704 1035 E msm8916_platform: platform_check_backends_match: Invalidsnd_device =
01-03 06:15:53.128 704 1035 D msm8916_platform: platform_can_split_snd_device: snd_device(2)num devices(0) new_snd_devices(0)
01-03 06:15:53.128 704 1035 D audio_hw_primary: enable_snd_device: snd_device(2: speaker)
01-03 06:15:53.152 704 1035 D msm8916_platform: platform_can_split_snd_device: snd_device(2)num devices(1) new_snd_devices(-299367304)
01-03 06:15:57.753 704 1035 D audio_hw_primary: disable_snd_device: snd_device(2: speaker)
01-03 06:15:57.753 704 1035 D msm8916_platform: platform_can_split_snd_device: snd_device(2)num devices(0) new_snd_devices(0)
01-03 06:18:01.804 704 1035 D audio_hw_primary: select_devices: out_snd_device(2: speaker)in_snd_device(0: )
01-03 06:18:01.805 704 1035 D msm8916_platform: platform_can_split_snd_device: snd_device(2)num devices(1) new_snd_devices(8)
01-03 06:18:01.805 704 1035 I msm8916_platform: platform_check_and_set_codec_backend_cfg: becf:new_snd_devices[0] is speaker
01-03 06:18:01.805 704 1035 E msm8916_platform: platform_check_backends_match: Invalidsnd_device =
01-03 06:18:01.805 704 1035 D msm8916_platform: platform_can_split_snd_device: snd_device(2)num devices(0) new_snd_devices(0)
01-03 06:18:01.805 704 1035 D audio_hw_primary: enable_snd_device: snd_device(2: speaker)
01-03 06:18:01.838 704 1035 D msm8916_platform: platform_can_split_snd_device: snd_device(2)num devices(1) new_snd_devices(-299367304)
01-03 06:18:10.213 704 1035 D audio_hw_primary: disable_snd_device: snd_device(2: speaker)
01-03 06:18:10.213 704 1035 D msm8916_platform: platform_can_split_snd_device: snd_device(2)num devices(0) new_snd_devices(0)
01-03 06:18:10.294 704 3960 Daudio_hw_primary: select_devices: out_snd_device(0: ) in_snd_device(87:handset-stereo-dmic-ef)
01-03 06:18:10.295 704 3960 D msm8916_platform: platform_can_split_snd_device: snd_device(87)num devices(0) new_snd_devices(0)
01-03 06:18:10.295 704 3960 D audio_hw_primary: enable_snd_device: snd_device(87:handset-stereo-dmic-ef)
01-03 06:18:10.331 704 3960 D msm8916_platform: platform_can_split_snd_device: snd_device(87)num devices(1) new_snd_devices(-350227112)
01-03 06:18:10.519 704 1035 D audio_hw_primary: select_devices: out_snd_device(2: speaker)in_snd_device(0: )
01-03 06:18:10.519 704 1035 D msm8916_platform: platform_can_split_snd_device: snd_device(2)num devices(1) new_snd_devices(18)
01-03 06:18:10.519 704 1035 I msm8916_platform: platform_check_and_set_codec_backend_cfg: becf:new_snd_devices[0] is speaker
通过红色部分关键内容D audio_hw_primary:select_devices: out_snd_device(0: ) in_snd_device(87: handset-stereo-dmic-ef)可知录音的时候用到了副MIC(也就是snd_device)，对应的设备名称为handset-stereo-dmic-ef
我们通过系统采用的hardware\qcom\audio\configs\msm8937\mixer_paths_qrd_sku2.xml文件，查看handset-stereo-dmic-ef的内容如下：
<pathname="handset-stereo-dmic-ef">
<path name="speaker-dmic-endfire" />
可知handset-stereo-dmic-ef调用了speaker-dmic-endfire，接着看speaker-dmic-endfire
<path name="adc1" />
<ctl name="DEC2 MUX" value="ADC2" />
<ctl name="MI2S_TX Channels" value="Two" />
<ctl name="ADC2 MUX" value="INP3" />
可知采用的是双升到，要改为单MIC，修改后如下：
<pathname="speaker-dmic-endfire">
<path name="adc1" />
同时新修hardware\qcom\audio\configs\msm8937\msm8937.mk下相关内容，修改前如下：
##fluencetype can be "fluence" or"fluencepro" or "none"
PRODUCT_PROPERTY_OVERRIDES += \
ro.qc.sdk.audio.fluencetype=fluence\
persist.audio.fluence.voicecall=true\
persist.audio.fluence.voicerec=false\
persist.audio.fluence.audiorec=true\
persist.audio.fluence.speaker=true
改后
##fluencetype can be "fluence" or"fluencepro" or "none"
PRODUCT_PROPERTY_OVERRIDES += \
ro.qc.sdk.audio.fluencetype=none\
#ro.qc.sdk.audio.fluencetype=fluence\
persist.audio.fluence.voicecall=true\
persist.audio.fluence.voicerec=false\
#persist.audio.fluence.audiorec=true\
persist.audio.fluence.speaker=true
