## 5.4\. Audio Recording While some of the requirements outlined in this section are listed as SHOULD since Android 4.3, the Compatibility Definition for future versions are planned to change these to MUST. Existing and new Android devices are **STRONGLY RECOMMENDED** to meet these requirements that are listed as SHOULD, or they will not be able to attain Android compatibility when upgraded to the future version. ### 5.4.1\. Raw Audio Capture and Microphone Information If device implementations declare `android.hardware.microphone`, they: * [C-1-1] MUST allow capture of raw audio content with the following characteristics: * **Format**: Linear PCM, 16-bit * **Sampling rates**: 8000, 11025, 16000, 44100, 48000 Hz * **Channels**: Mono * SHOULD allow capture of raw audio content with the following characteristics: * **Format**: Linear PCM, 16-bit and 24-bit * **Sampling rates**: 8000, 11025, 16000, 22050, 24000, 32000, 44100, 48000 Hz * **Channels**: As many channels as the number of microphones on the device * [C-1-2] MUST capture at above sample rates without up-sampling. * [C-1-3] MUST include an appropriate anti-aliasing filter when the sample rates given above are captured with down-sampling. * SHOULD allow AM radio and DVD quality capture of raw audio content, which means the following characteristics: * **Format**: Linear PCM, 16-bit * **Sampling rates**: 22050, 48000 Hz * **Channels**: Stereo * [C-1-4] MUST honor the [`MicrophoneInfo`]( https://developer.android.com/reference/android/media/MicrophoneInfo) API and properly fill in information for the available microphones on device accessible to the third-party applications via the [`AudioManager.getMicrophones()`]( https://developer.android.com/reference/android/media/AudioManager#getMicrophones%28%29) API, and the currently active microphones which are accessible to the third party applications via the [`AudioRecord.getActiveMicrophones()`]( https://developer.android.com/reference/android/media/AudioRecord#getActiveMicrophones%28%29) and [`MediaRecorder.getActiveMicrophones()`](https://developer.android.com/reference/android/media/MediaRecorder#getActiveMicrophones%28%29) APIs. If device implementations allow AM radio and DVD quality capture of raw audio content, they: * [C-2-1] MUST capture without up-sampling at any ratio higher than 16000:22050 or 44100:48000. * [C-2-2] MUST include an appropriate anti-aliasing filter for any up-sampling or down-sampling. ### 5.4.2\. Capture for Voice Recognition If device implementations declare `android.hardware.microphone`, they: * [C-1-1] MUST capture `android.media.MediaRecorder.AudioSource.VOICE_RECOGNITION` audio source at one of the sampling rates, 44100 and 48000. * [C-1-2] MUST, by default, disable any noise reduction audio processing when recording an audio stream from the `AudioSource.VOICE_RECOGNITION` audio source. * [C-1-3] MUST, by default, disable any automatic gain control when recording an audio stream from the `AudioSource.VOICE_RECOGNITION` audio source. * SHOULD record the voice recognition audio stream with approximately flat amplitude versus frequency characteristics: specifically, ±3 dB, from 100 Hz to 4000 Hz. * SHOULD record the voice recognition audio stream with input sensitivity set such that a 90 dB sound power level (SPL) source at 1000 Hz yields RMS of 2500 for 16-bit samples. * SHOULD record the voice recognition audio stream so that the PCM amplitude levels linearly track input SPL changes over at least a 30 dB range from -18 dB to +12 dB re 90 dB SPL at the microphone. * SHOULD record the voice recognition audio stream with total harmonic distortion (THD) less than 1% for 1 kHz at 90 dB SPL input level at the microphone. If device implementations declare `android.hardware.microphone` and noise suppression (reduction) technologies tuned for speech recognition, they: * [C-2-1] MUST allow this audio effect to be controllable with the `android.media.audiofx.NoiseSuppressor` API. * [C-2-2] MUST uniquely identify each noise suppression technology implementation via the `AudioEffect.Descriptor.uuid` field. ### 5.4.3\. Capture for Rerouting of Playback The `android.media.MediaRecorder.AudioSource` class includes the `REMOTE_SUBMIX` audio source. If device implementations declare both `android.hardware.audio.output` and `android.hardware.microphone`, they: * [C-1-1] MUST properly implement the `REMOTE_SUBMIX` audio source so that when an application uses the `android.media.AudioRecord` API to record from this audio source, it captures a mix of all audio streams except for the following: * `AudioManager.STREAM_RING` * `AudioManager.STREAM_ALARM` * `AudioManager.STREAM_NOTIFICATION` ### 5.4.4\. Acoustic Echo Canceler If device implementations declare `android.hardware.microphone`, they: * SHOULD implement an [Acoustic Echo Canceler](https://en.wikipedia.org/wiki/Echo_suppression_and_cancellation) (AEC) technology tuned for voice communication and applied to the capture path when capturing using `AudioSource.VOICE_COMMUNICATION` If device implementations provides an Acoustic Echo Canceler which is inserted in the capture audio path when `AudioSource.VOICE_COMMUNICATION` is selected, they: * [C-SR] are STRONGLY_RECOMMENDED to declare this via [AcousticEchoCanceler](https://developer.android.com/reference/android/media/audiofx/AcousticEchoCanceler) API method [AcousticEchoCanceler.isAvailable()](https://developer.android.com/reference/android/media/audiofx/AcousticEchoCanceler.html#isAvailable()) * [C-SR] are STRONGLY_RECOMMENDED to allow this audio effect to be controllable with the [AcousticEchoCanceler](https://developer.android.com/reference/android/media/audiofx/AcousticEchoCanceler) API. * [C-SR] are STRONGLY_RECOMMENDED to uniquely identify each AEC technology implementation via the [AudioEffect.Descriptor.uuid](https://developer.android.com/reference/android/media/audiofx/AudioEffect.Descriptor.html#uuid) field. ### 5.4.5\. Concurrent Capture If device implementations declare `android.hardware.microphone`,they MUST implement concurrent capture as described in [this document]( https://developer.android.com/features/sharing-audio-input). Specifically: * [C-1-1] MUST allow concurrent access to microphone by an accessibility service capturing with `AudioSource.VOICE_RECOGNITION` and at least one application capturing with any `AudioSource`. * [C-1-2] MUST allow concurrent access to microphone by a pre-installed application that holds an Assistant role and at least one application capturing with any `AudioSource` except for `AudioSource.VOICE_COMMUNICATION` or `AudioSource.CAMCORDER`. * [C-1-3] MUST silence the audio capture for any other application, except for an accessibility service, while an application is capturing with `AudioSource.VOICE_COMMUNICATION` or `AudioSource.CAMCORDER`. However, when an app is capturing via `AudioSource.VOICE_COMMUNICATION` then another app can capture the voice call if it is a privileged (pre-installed) app with permission `CAPTURE_AUDIO_OUTPUT`. * [C-1-4] If two or more applications are capturing concurrently and if neither app has an UI on top, the one that started capture the most recently receives audio. ### 5.4.6\. Microphone Gain Levels If device implementations declare `android.hardware.microphone`, they: * SHOULD exhibit approximately flat amplitude-versus-frequency characteristics in the mid-frequency range: specifically ±3dB from 100 Hz to 4000 Hz for each and every microphone used to record the voice recognition audio source. * SHOULD set audio input sensitivity such that a 1000 Hz sinusoidal tone source played at 90 dB Sound Pressure Level (SPL) yields a response with RMS of 2500 for 16 bit-samples (or -22.35 dB Full Scale for floating point/double precision samples) for each and every microphone used to record the voice recognition audio source. * [C-SR] are STRONGLY RECOMMENDED to exhibit amplitude levels in the low frequency range: specifically from ±20 dB from 5 Hz to 100 Hz compared to the mid-frequency range for each and every microphone used to record the voice recognition audio source. * [C-SR] are STRONGLY RECOMMENDED to exhibit amplitude levels in the high frequency range: specifically from ±30 dB from 4000 Hz to 22 KHz compared to the mid-frequency range for each and every microphone used to record the voice recognition audio source.