PlayAudio - Amazon Chime SDK

PlayAudio

Play an audio file on any leg of a call. The audio can be repeated any number of times. The in-progress audio can be terminated using the DTMF digits set in the PlaybackTerminators.

Currently, Amazon Chime SDK only supports playing audio files from the Amazon Simple Storage Service (Amazon S3) bucket. The S3 bucket must belong to the same AWS account as the SIP media application. In addition, you must give the s3:GetObject permission to the Amazon Chime SDK Voice Connector service principal. You can do that by using the S3 console or the command-line interface (CLI).

The following code example shows a typical bucket policy.

{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Sid": "SMARead",
            "Effect": "Allow",
            "Principal": {
                "Service": "voiceconnector.chime.amazonaws.com"
            },
            "Action": [
                "s3:GetObject"
            ],
            "Resource": "arn:aws:s3:::bucket-name/*",
                "Condition": {
                "StringEquals": {
                    "aws:SourceAccount": "aws-account-id"
                }
            }
        }
    ]
}

The Audio Service reads and writes to your S3 bucket on behalf of your Sip Media Application. To avoid the confused deputy problem, you can restrict S3 bucket access to a single SIP media application.

{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Sid": "SMARead",
            "Effect": "Allow",
            "Principal": {
                "Service": "voiceconnector.chime.amazonaws.com"
            },
            "Action": [
                "s3:GetObject"
            ],
            "Resource": "arn:aws:s3:::bucket-name/*",
                "Condition": {
                "StringEquals": {
                    "aws:SourceAccount": "aws-account-id",
                    "aws:SourceArn": "arn:aws:chime:region:aws-account-id:sma/sip-media-application-id"
                }
            }
        }
    ]
}

The following code example shows a typical action.

{
    "Type": "PlayAudio",
    "Parameters": {
        "CallId": "call-id-1",
        "ParticipantTag": "LEG-A",
        "PlaybackTerminators": ["1", "8", "#"],
        "Repeat": "5",
        "AudioSource": {
            "Type": "S3",
            "BucketName": "valid-S3-bucket-name",
            "Key": "wave-file.wav"
        }
    }
}
CallID

Description – CallId of participant in the CallDetails.

Allowed values – A valid call ID.

Required – No, if ParticipantTag is present.

Default value – None.

ParticipantTag

Description – ParticipantTag of one of the connected participants in the CallDetails.

Allowed values – LEG-A or LEG-B.

Required – No, if CallId is present.

Default value – ParticipantTag of the invoked callLeg. Ignored if you specify CallId.

PlaybackTerminator

Description – Terminates in-progress audio by using DTMF input from the user

Allowed values – An array of the following values; “0”, ”1”, ”2”, ”3”, ”4”, ”5”, ”6”, ”7”, ”8”, ”9”, ”#”, ”*”

Required – No

Default value – None

Repeat

Description – Repeats the audio the specified number of times

Allowed values – An integer greater than zero

Required – No

Default value – 1

AudioSource.Type

Description – Type of source for audio file.

Allowed values – S3.

Required – Yes.

Default value – None.

AudioSource.BucketName

Description – For S3 source types, the S3 bucket must belong to the same AWS account as the SIP application. The bucket must have access to the Amazon Chime SDK Voice Connector service principal, which is voiceconnector.chime.amazonaws.com.

Allowed values – A valid S3 bucket for which Amazon Chime SDK has access to the s3:GetObject action.

Required – Yes.

Default value – None.

AudioSource.key

Description – For S3 source types, the file name from the S3 bucket specified in the AudioSource.BucketName attribute.

Allowed values – A valid audio file.

Required – Yes.

Default value – None.

The SIP media application tries to play the audio from the source URL. You can use raw, uncompressed PCM .wav files of no more than 50 MB in size. Amazon Chime SDK recommends 8 KHz mono.

When the last instruction in a dialplan is PlayAudio and the file finishes playback, or if a user stops playback with a key press, the application invokes the AWS Lambda function with the event shown in the following example.

{
    "SchemaVersion": "1.0",
    "Sequence": INTEGER,
    "InvocationEventType": "ACTION_SUCCESSFUL",
    "ActionData": {
        "Type": "PlayAudio",
        "Parameters" : {
            "CallId": "call-id-1",
            "AudioSource": {
                "Type": "S3",
                "BucketName": "valid-S3-bucket-name",
                "Key": "wave-file.wav",
         }           
     }
}

After a terminating digit stops the audio, it won't be repeated.

Error handling

When the validation file contains errors, or an error occurs while running an action, the SIP media application calls an AWS Lambda function with the appropriate error code.

Error Message Reason

InvalidAudioSource

The audio source parameter is invalid.

This error can happen for multiple reasons. For example, the SIP media application cannot access the file due to permission issues, or issues with the URL. Or, the audio file may fail validation due to format, duration, size, and so on.

SystemException

System error while running action.

Another system error occurred while running the action.

InvalidActionParameter

CallId or ParticipantTag parameter for action is invalid.

The action contains an invalid parameter.

The following code example shows a typical invocation failure.

{
    "SchemaVersion": "1.0",
    "Sequence": 2,
    "InvocationEventType": "ACTION_FAILED",
    "ActionData": {
        "Type": "PlayAudio",
        "Parameters" : {
            "CallId": "call-id-1",
            "AudioSource": {
                "Type": "S3",
                "BucketName": "bucket-name",
                "Key": "audio-file.wav"
            },
        },
        "ErrorType": "InvalidAudioSource",
        "ErrorMessage": "Audio Source parameter value is invalid."
    }
    "CallDetails": {
        ...
    }
}

See working examples on GitHub: