Conference Recording

[subhra_saranyu]

Hi Max, Can you please help us with the following queries -

1. Does Flashphoner recording feature has any API or Webhooks which can tell us when a recording is ending ?

2. Is there any way to identify each recording with the participant's name along with some other parameters ?

 

[Max]

Good day.

1. Does Flashphoner recording feature has any API or Webhooks which can tell us when a recording is ending ?

There is a script which is launched after recording of a stream is finished, by default /usr/local/FlashphonerWebCallServer/bin/on_record_hook.sh. Please read detais here.

2. Is there any way to identify each recording with the participant's name along with some other parameters ?

By default, Room.publish() method publishes a stream named like roomName-participantName-randomSuffix, for example

room-4a4788-test-ff76

where room-4a4788 is room name and test is participant name.
You can set recording file name template as

stream_record_policy_template={streamName}-{startTime}-{endTime}

In this case, stream name will be used in recording name, so you can identify each recording with participant name.

 

[subhra_saranyu]

Hi Max, is enable_empty_shift_writer flag removed from the latest version of Flashphoner ?

We have a use case where we are sending each participant's recorded mp4 file to AWS for Transcribe, to get properly synced speech to text we need to sync video of partipant1 & participant2, is there any other way to achieve that then please suggest.

 

[Max]

Hi Max, is enable_empty_shift_writer flag removed from the latest version of Flashphoner ?

Yes, this feature is obsoleted and removed since build 5.2.142.

We have a use case where we are sending each participant's recorded mp4 file to AWS for Transcribe, to get properly synced speech to text we need to sync video of partipant1 & participant2, is there any other way to achieve that then please suggest.

You can use a new feature to record multiple streams from a chat room with mixing them after conference ending. Please read details here.

 

[subhra_saranyu]

yes Max this I checked, but for Transcribe we need each participant's name as well, so that the end result can be like this for ex:

Savvy: Hi, how are you?
Max: fine, what about you?

Previously what we thought we will record a participant with his name for ex. Savvy-{uniqueKey}.mp4, and same name will be passed to AWS Transcribe to get output like Savvy-{uniqueKey}.json.

any other way is there to achieve this use case?

 

[Max]

any other way is there to achieve this use case?

You can combine each participant stream recording with multiple streams recording:
1. Set recoding file name template to

stream_record_policy_template={streamName}

2. Enable room streams recording to one file

connection.join({
    name: getRoomName(),
    record: true
}).on(ROOM_EVENT.STATE, function(room){
    ...
});

3. Enable participants stream recording

function publishLocalMedia(room) {
    ...
    room.publish({
        display: display,
        constraints: constraints,
        record: false,
        receiveVideo: false,
        receiveAudio: false,
        record: true
        ...
    });
}

The mixed room stream will be recorded to file named roomName_multiRecorded_mixed, for example room1_multiRecorded_mixed.mp4
Each participants stream will be recorded to file named roomName-participantName-randomSuffix[/ICODE[, for example [ICODE]room1-Savvy-1234.mp4
So you can pass the mixed file to AWS Transcribe and each participants file to AWS Transcribe. You'll receive JSON files per each participant and one JSON file for mixed stream.
Is this enough?