Voice Recording

Stream chat allows you to exchange Voice Recordings in your channels. Those Voice Recordings are a built-in Attachment types (as are being defined here). We are going to take a look in the main parts of Voice Recording feature below.

note

Voice Recordings are available since version 4.32.0.

VoiceRecording creation and playback is by default disabled. If you would like to enable it, you can do so by setting the Components isVoiceRecordingEnabled property to true like below:

Components.default.isVoiceRecordingEnabled = true

note

When the VoiceRecording feature is disabled, any message containing VoiceRecordings will render them as regular audio attachments.

UI Flows

Recording

The recording flow is being presented on the ComposerContentView of the ComposerVC and it's being managed by VoiceRecordingVC. The VoiceRecordingVC takes care of the following:

• Update the UI based on the recordings state (for example recording, locked or in preview)
• Update the UI with recording related information (for example duration, current playback time)
• Coordinate information flow from audio recording and player.
• Communicate with ComposerVC in order to add a VoiceRecording as attachment or send it.

The main UI components that VoiceRecordingVC manages are the following:

recordingTipView

A view that is being used to display a tip to user when the tap duration on the recording button wasn't long enough (the display of this view relates to RecordButton.minimumPressDuration).

slideToCancelView

A view that during an unlocked recording, displays information on how to cancel the active recording.

recordingIndicatorView

A view that indicates to the user that we are currently recording audio.

lockIndicatorView

A view that indicates to the user if the currently active recording is locked or not.

liveRecordingView

A view that during a locked recording, displays information about the active recording or its preview.

bidirectionalPanGestureRecogniser

The gesture recogniser used to identify touch movements in the horizontal or vertical axis.

Presentation

VoiceRecordings are being presented using the voiceRecordingAttachmentView that is defined and configurable in Components:

Components.default.voiceRecordingAttachmentView = ChatMessageVoiceRecordingAttachmentListView.ItemView.self

The view manages the visibility and state of all UI components (e.g play/pause button, waveform visualization, playback rate button and name label).

Properties

content

The content of the VoiceRecordingVC.

slideToCancelDistance

The distance (in pixels) the user needs to slide to the leading side of the view to cancel a recording.

note

Default value: 75

lockDistance

The distance (in pixels) the user needs to slide to the top side of the view to lock the recording view.

note

Default value: 50

waveformTargetSamples

The number of samples (data points) we want to extract from a recording in order to render a waveform visualisation of it.

note

Default value: 100

delegate

The delegate which the VoiceRecordingVC will ask for support on presenting views and communicating the state of the recording flow to its parent controller.

audioAnalysisFactory

An object responsible for extracting the required number of samples from an audio file, that can be used to render a waveform visualisation.

Customization

Components configuration

isVoiceRecordingConfirmationRequiredEnabled

When set to true recorded messages can be grouped together and send as part of one message. When set to false, recorded messages will be sent instantly.

Components.default.isVoiceRecordingConfirmationRequiredEnabled = false

By default the property is set to true.

voiceRecordingViewController

A ViewController that manages all aspects of the voice recording from the ComposerVC.

tip

The VoiceRecordingVC even though it's a ViewController (as it manages a view) it doesn't manage its own view. Instead, the view property of the ViewController has been overridden and returns the ComposerView.

Avoid adding the ViewController's view in view hierarchy. ::::

Components.default.voiceRecordingViewController = VoiceRecordingVC.self

audioPlayer

The AudioPlayer that will be used for the voice recording playback.

Components.default.audioPlayer = StreamRemoteAudioQueuePlayer.self

Default value is StreamRemoteAudioQueuePlayer.self. In case you want to disable voice recording messages to play automatically once the previous one finishes, you can set this one to StreamRemoteAudioPlayer.self :::

audioRecorder

The AudioRecorder that will be used to record new voice recordings.

Components.default.audioRecorder = StreamAudioRecorder.self

audioSessionFeedbackGenerator

A feedbackGenerator that will be used to provide haptic feedback during the recording flow. It can be customized in cases where you want to provide different or none haptic feedback for actions during the recording flow.

Components.default.audioSessionFeedbackGenerator = StreamAudioSessionFeedbackGenerator.self

audioQueuePlayerNextItemProvider

ComposerVC parent view controllers (ChatChannelVC and ChatThreadVC) will become the data source of the audio player if it's an instance of StreamRemoteAudioQueuePlayer. The default implementation of the AudioQueuePlayerDatasource in those view controllers is using the audioQueuePlayerNextItemProvider instance defined here to find the next (if any) voice recording to play.

Components.default.audioQueuePlayerNextItemProvider = AudioQueuePlayerNextItemProvider.self

When calling findNextItem on the AudioQueuePlayerNextItemProvider instance, we can specify a lookup scope. This will inform the AudioQueuePlayerNextItemProvider to know where to look for the next voice recording. The available values for the lookUpScope are:

sameMessage

The provider will look for the next VoiceRecording in the attachments of the message containing the currently playing URL.

subsequentMessagesFromUser

The provider will look for the next VoiceRecording in the attachments of the of the message containing the currently playing URL and if not found will apply the same logic in all subsequent messages that have the same author.

You can create your own lookupScope instances and override the implementation AudioQueuePlayerNextItemProvider in order to fit its functionality to your use-cases.