Announcement

Collapse
No announcement yet.

Airfoil for wireless speech distribution

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    Airfoil for wireless speech distribution

    Hi All,

    I'm creating this thread to move a discussion on Airfoil out of the Roku forum section where I accidentally detoured a thread on the Soundbridge.

    I've been looking at using a software product called "Airfoil" that allows for audio capture and redirection to an Apple Airport Express, an Apple TV device, or a computer running a free client software called "Airfoil Speakers". The thing that caught my attention the most about this was that it allows you to capture audio from any application or from the audio output driver directly so I'm thinking about trying to use it for wireless speech distribution. they have free clients for Mac, Windows, Linux, and the iPhone/iTouch so this could be quite useful if it works. I'm going to do some playing this weekend and I'll update this thread with my progress.

    Here's the thread I accidentally jumped into: http://board.homeseer.com/showthread.php?t=135714

    Please continue the discussion in this thread.

    Thanks,
    Ken
    "if I have seen further [than others], it is by standing on the shoulders of giants." --Sir Isaac Newton (1675)

    #2
    From Markus in other thread:

    =========
    Hi,
    i tried using airfoil to distribute announcements to my airport expresses, but t the lag was a bit to much. 3-5 sec. On mac i think there is somr kind of api but not on windows unfortunatly.
    ============

    Hi Markus,

    On the Airfoil support forum they talk about a 2 second buffer that's built into the AirTunes protocol and cannot be changed. I'm looking at using it mainly for announcements so the delay should not be a problem unless I end up with a time critical announcement. I could see how the delay could be a problem in some applications but for what I'm looking at doing I think it will be OK.

    I've got some testing to do this weekend and if I find that it works good enough for my application I'll contact the developers and ask them about adding a Windows scripting API in the future.

    Ken
    "if I have seen further [than others], it is by standing on the shoulders of giants." --Sir Isaac Newton (1675)

    Comment


      #3
      3-5 second lag is okay for music and such for me, but not for announcements and notifications like the door is open etc.

      I'm really looking for find a network sound solution, as most of the hypervisors for virtualizing servers do not support sound.

      I can make just about every other device work with the network, why does sound have to be an exception. Funny thing is, I'm almost ready to consider running a very lower power computer just for homeseer speaker client. Unfortunately that just makes for one more PC to maintain

      What we really need is a homeseer speaker client hardware solution
      Joe (zimmer62)

      BLSecurtiy, AC-RF2, RCS Serial Thermostats, RFXCOM SMarthome SwitchLinc, mcsXap, Global Cache GC100, SqueezeBox, TWA_ONKYOINTEGRA, BLLogMonitor, BLPlugins, BLRadar, BLSpeech, BLZLog.aspx, HSTouch (Windows, iPhone, iPod), USB Mimo touchscreens, VMWare Server, Vortexbox, Windows Home Server, MyMovies, Windows Media Center, X10, ZWave, and much much much more.

      Comment


        #4
        Something running on a open WRT 802 router would be great. I had look at this some time back but never got the time to dig in. The idea was a Open WRT router with a USB audio dongle. The setup would be like above with some audio redirection from the main server running multiple HS speaker clients.
        My original intent was to put one in the car and use it for 3 features

        1) Detect the router assosiated to the network for car presence
        2) Use Mic and audio to send voice commands to HS ie. open gararge door and others
        3) Add a keypad for push button comands (ip messages to HS)

        Comment


          #5
          Hi Joe,

          Have you looked at this solution?

          http://www.barix.com/Audio_Distribution/1141/

          It's not cheap but it is very reliable and works well. It was being installed in one of the hotels where my company was installing the network.

          Also, I have to ask about what kind of an announcement would HS be making that couldn't be delayed 3-5 seconds? The reason I'm asking is for my own education as I'm looking at adding announcements as a part of our new intercom system and I can't think of a single thing I'd have HS say that couldn't be delayed. Even something like a door being open seems like it would not really be so critical that the notification couldn't be delayed a few seconds. The few things I could think of that could be critical are related to the security system and they are hardwired to a siren.

          Thanks,
          Ken
          "if I have seen further [than others], it is by standing on the shoulders of giants." --Sir Isaac Newton (1675)

          Comment


            #6
            Originally posted by kenm View Post
            Also, I have to ask about what kind of an announcement would HS be making that couldn't be delayed 3-5 seconds? The reason I'm asking is for my own education as I'm looking at adding announcements as a part of our new intercom system and I can't think of a single thing I'd have HS say that couldn't be delayed. Even something like a door being open seems like it would not really be so critical that the notification couldn't be delayed a few seconds. The few things I could think of that could be critical are related to the security system and they are hardwired to a siren.
            Security is a big one, but announcements for most things don't make much difference.

            I've got different sounds being made for each door, and if it's being opened or closed. If I were to hear one of those in the middle of the night, I'd know right were the intruder entered (not likely to happen, but a 3 second delay would be a big problem there)

            3 seconds for to change tracks or adjust the volume could be annoying, but tolerable.

            Actually the biggest one of all is that I have homeseer respond with sounds to certain button presses to know that actions have been acknowledged. I have several switches that are RF that control lights that are not in the same room.. If I press the button on my X10 keypad, HS plays a sound effect of a button or a switch flipping on or off based on which direction it's toggling. (turning on or off lights in another room) Or adjusting the thermostat from bed with the keypad. Getting that instant feedback is important, otherwise the button would get pressed 4 or 5 times me thinking it's not working. I'm not sure what you call this, probably some sort of audible feedback confirmation.

            One last one, is reminders when I'm leaving the house. I press a button to leave the house on my way out the door. I hear a sound effect that says I'm good to go, or my house speaks things that I'm supposed to remember. Like the back door is still open, or something to that effect. If I had to way 3 to 5 seconds, I'd already be out the door and unable to hear any of the announcements.

            For stuff like "It's 7:00 AM and the coffee pot has been turned on" really doesn't matter about the latency of even 20 seconds I just wouldn't care.

            Hope that helps.

            I'll have to take a look at the product sometime. I'm not sure what I'm willing to pay yet, but I'm open to most options as long as they arn't so expensive that I might as well just build and run a PC with the speaker client on it. (24x7 power plus cost of PC < cost of other solution)
            Joe (zimmer62)

            BLSecurtiy, AC-RF2, RCS Serial Thermostats, RFXCOM SMarthome SwitchLinc, mcsXap, Global Cache GC100, SqueezeBox, TWA_ONKYOINTEGRA, BLLogMonitor, BLPlugins, BLRadar, BLSpeech, BLZLog.aspx, HSTouch (Windows, iPhone, iPod), USB Mimo touchscreens, VMWare Server, Vortexbox, Windows Home Server, MyMovies, Windows Media Center, X10, ZWave, and much much much more.

            Comment


              #7
              Hi Joe,

              Thanks for the detailed explanation. That gives me a lot to think about as I look at adding speech to my system.

              Thanks again,
              Ken
              "if I have seen further [than others], it is by standing on the shoulders of giants." --Sir Isaac Newton (1675)

              Comment


                #8
                Since the only physical windows box in my house is for homeseer, I've been playing with PulseAudio under linux, which includes UPnP multicasting of audio streams among any hosts that subscribe to them. Right now I have linux boxen all over that I can hook speakers to, and I think it's a solution I'm going to go barking at a bit. There happens to be windows binaries of pulseaudio available, or at least home various hacks for putting an audio stream on the network to hosts. This would more or less behave like the HS speaker app... Now if only someone would build the functionality into the speaker app to talk to my PulseAudio clients natively. http://www.pulseaudio.org/wiki/FAQ#H...overthenetwork http://ubuntuforums.org/showthread.php?t=791643 http://gleamynode.net/articles/2228/ -mb

                Comment


                  #9
                  mbutash
                  I did some reading on PulseAudio and am unclear on something. Can PulseAudio be run on a windows box and remotly access linux box microphone and speaker? Or said a diffrent way, could the HS speaker client(running on HS server) Microphone and speakers be on my linux box?

                  Comment


                    #10
                    Originally posted by mbutash View Post
                    There happens to be windows binaries of pulseaudio available, or at least home various hacks for putting an audio stream on the network to hosts. This would more or less behave like the HS speaker app... Now if only someone would build the functionality into the speaker app to talk to my PulseAudio clients natively.
                    I've looked at that same webpage, and got as far as setting up Virtual Audio Cable (VAC) on a PC with no sound card, this lets windows see a virtual sound card. I havn't made it to the point of streaming audio to the linux box yet though. It's by far from a nice installation, but might make a good solution for me trying to run homeseer in a virtual machine that doesn't support sound passthrough. I'd really like someone to just write a virtual sound card driver for windows that would stream to a pulse audio server that you specified the IP. I'd even be willing to pay for it if a windows pulse audio service was written that could accept and route audio from any windows PC in the house. Example If I'm playing a video on my laptop I'd like the audio to come out of my desktop PC if I'm in one room, and out of HTPC if I'm in the other room etc. (The service would be running on each machine, and a virtual sound card would be installed on each machine, then you just tell it which machine you want the audio to come from)

                    Originally posted by Gerard View Post
                    mbutash
                    I did some reading on PulseAudio and am unclear on something. Can PulseAudio be run on a windows box and remotly access linux box microphone and speaker? Or said a diffrent way, could the HS speaker client(running on HS server) Microphone and speakers be on my linux box?
                    PulseAudio can be run on windows or linux. using some goofy command line switches and a combination of virtual audio cable or if your sound card supports capturing the audio stream out of applications you should be able to "access" linux box speaker, don't know about microphone (but I think the answer is yes)

                    This is not a very well documented path, it also seems to have a very rough implementation for windows. There's a compiled binary you can get for windows, but there is no front end, no easy way to stream remotely.
                    Joe (zimmer62)

                    BLSecurtiy, AC-RF2, RCS Serial Thermostats, RFXCOM SMarthome SwitchLinc, mcsXap, Global Cache GC100, SqueezeBox, TWA_ONKYOINTEGRA, BLLogMonitor, BLPlugins, BLRadar, BLSpeech, BLZLog.aspx, HSTouch (Windows, iPhone, iPod), USB Mimo touchscreens, VMWare Server, Vortexbox, Windows Home Server, MyMovies, Windows Media Center, X10, ZWave, and much much much more.

                    Comment


                      #11
                      I tested streaming audio from a windows machine (with no sound card) using virtual cable, lineco.exe and putty's plink.exe

                      As outlined here:
                      http://gleamynode.net/articles/2228/
                      using plink instead of the ssh.exe they talked about.

                      For starters it worked very easily.

                      The latency seemed like it was around a half a second, but I didn't have the tools or time to measure it. I'd say it's probably acceptable latency in the homeseer realm for most things, but not as ideal as < 100ms I'll do some more testing to see if that can be reduced or eliminated. I also need to figure out if that's introduced because of virtual audio cable, ssh, pulseaudio, or network.

                      Watching videos seemed unacceptable as the sound didn't match the video at all, however playing music this solution should work just fine.

                      If the latency could be eliminated I'll probably be using this method with my virtual machines so I can pull the sound card out of the loop and actually run my homeseer in a VM on a baremetal, or linux host

                      Hopefully I get some more time to play with this method seems like a good solution if the latency can be dropped.
                      Joe (zimmer62)

                      BLSecurtiy, AC-RF2, RCS Serial Thermostats, RFXCOM SMarthome SwitchLinc, mcsXap, Global Cache GC100, SqueezeBox, TWA_ONKYOINTEGRA, BLLogMonitor, BLPlugins, BLRadar, BLSpeech, BLZLog.aspx, HSTouch (Windows, iPhone, iPod), USB Mimo touchscreens, VMWare Server, Vortexbox, Windows Home Server, MyMovies, Windows Media Center, X10, ZWave, and much much much more.

                      Comment


                        #12
                        Originally posted by zimmer62 View Post
                        3-5 second lag is okay for music and such for me, but not for announcements and notifications like the door is open etc.

                        I'm really looking for find a network sound solution, as most of the hypervisors for virtualizing servers do not support sound.

                        I can make just about every other device work with the network, why does sound have to be an exception. Funny thing is, I'm almost ready to consider running a very lower power computer just for homeseer speaker client. Unfortunately that just makes for one more PC to maintain

                        What we really need is a homeseer speaker client hardware solution
                        I'm running my HS on a virtualized XP using free VMServer. I have Airfoil running on the underlying W2K3 OS. The HS Speaker.exe client runs on the W2K3 server and connects to the VM. Airfoil uses the speaker.exe client as the audio source.

                        I only see a slight delay, and have Airfoil sending announcements to my Ipod Touch, the host computer, an Airport Express and a Linux machine all at the same time. Cool stuff. I don't really notice the delay. In addition, the Airport Express also acts as a wireless access point and a print server. I picked it up on eBay for under $50.

                        Comment

                        Working...
                        X