Skip to main content
Apr 25, 2011 11:23 AM UTC
Answered

What is a common way to create audio chat rooms?

What is an expected way to create audio chatrooms? Either try to mix all published audio streams into single one and republish it under common name? Or just inform all clients about each other and allow each one to connect to a stream, published by each other?

Do I need to write custom webapplication with my own scripts in it, or this is some default functionality?

Thanks.

    This topic is closed to new replies. Start a new post to keep the conversation going.
    Nikhil_Kalyan
    Nikhil_KalyanCorrect answer
    Participating Frequently
    Apr 25, 2011 2:26 PM UTC

    Hi Dmitry,

    I doubt if mixing is possible in FMS. Even for your many-many case, I would still say to have a separate stream for each user (so that he can 'speak') and playback/subscribe all the streams to all the users and record them as well. So you will have as many recordings /streams as the number of users. Then you need to write some scripting to 'sync' all of them (all the recordings).

    Hope it helps.

    Thank you !

    3 replies

    Nikhil_Kalyan
    Participating Frequently
    Apr 25, 2011 11:59 AM UTC

    Hi,

    Since any audio chat application is more or less a one-many or one-one communication, it is better you make different streams and allow each client to either chose to connect or automatically join/subscribe one of the published streams depending on your requirement(s). You do not need to combine them all into one single stream.

    Yes, you need to write your own custom application(s) for your particular use case. We can help as much as we can though !

    Thank you !

    Translate
    Apr 25, 2011 1:51 PM UTC

    I want all visitors to hear each other. User ability to mute himself is required, but user ability to mute other sides separately is not required. I.e. user is just hearing all sound in the room.

    Also recording of room "ambience", which is consisting of voices of all visitors is required.

    So, I wonder if mixing is possible at all? Does API have appropriate functions? Won't it load a server hard?

    And if mixing is impossible or not recommended, then how is it possible to record entire room sounding? Is it practical to put individual visitor sessions separately? But how to place them in time correctly?

    Thank you in advance!

    Regards,

       Dmitry

    Translate
    Nikhil_Kalyan
    Nikhil_KalyanCorrect answer
    Participating Frequently
    Apr 25, 2011 2:26 PM UTC

    Hi Dmitry,

    I doubt if mixing is possible in FMS. Even for your many-many case, I would still say to have a separate stream for each user (so that he can 'speak') and playback/subscribe all the streams to all the users and record them as well. So you will have as many recordings /streams as the number of users. Then you need to write some scripting to 'sync' all of them (all the recordings).

    Hope it helps.

    Thank you !

    Translate