Announcement

Collapse
No announcement yet.

Mega-Size installation and I'm fried from info-overload

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

    #16
    Thanks Brian,

    One thing that seems to be lacking in the HA scene (so far in my research) is audio over IP (including telephony or VoIp).

    Playing voicemail over IP is neat, but not that amazing...I mean it's just playing a wave file that the voice modem saved.
    It seems that if using PC's in "kiosk" mode, that it would be nice to be able to answer the phone at them. Ironically, it appears there are caller ID plugins for most HA. There are USB phones, and IP phones up the yang, but it seems there is no software to remotely transmit signal to and from a modem in real time. The codec should be similar to any "internet phone"...just internal to the network. This is one of thos "I wish I were a programmer" things.

    The next in line is intercoms. There are hardware IP intercoms for use over ethenet, and software intercoms that do the same...but no plugins that I've found for say...MainLobby. Seems a natural.

    And lastly, syncronized streaming audio. If you have a soundcard at every station, it seems a waste to not use it and run physical wire throughout a building from a hardware solution. I don't think syncing the audio is as tricky as people think. Short of timecode or digital resolving, I think it could be done by delaying the source by at least the amount of latency that the slowest station has, and then programing the delay of the slave station to match up with the source. If you don't use the source (server's) audio directly, then most of the slaves should be experiencing very similar latency and should be easy to sync together I did pro-audio for music and film for over a decade so I'm very curious about this and why it's not being done.

    As I said...looking around, there are already software solutions for most of this, but maybe they haven't made it to HA yet.
    Phones, intercoms, and "joining" the whole house audio stream really seem like they would be priority tasks...over things like checking the weather anyway.

    Thoughts?
    Sean

    Comment


      #17
      VoIp has been brought up several times. Way2Call does support this But currently Homeseer Phone don't take advantage of this (Feature Request?). It has been discussed several times but nothing ever came of it. Lack of request I assume. I think everyone is headed that way so it would be a good feature. There should be a way of doing this But I'm not aware of anything. Way2Call is very powerful but its features are hardly used with Homeseer.

      Intercoms I really know nothing about it but if a request is posted on Mainlobby site someone might be willing to write a plugin. I know Mario does write custom stuff (of course for a fee).

      In syncronized streaming audio, I'm really not sure what you are planning on but check this link out. http://www.smarthome.com/7666.html It may be what your looking for.

      Comment


        #18
        Well I know this is going to sound the wrong way but...(and I really am a nice guy)...

        I find it hard to believe that these basic features aren't "stock" in any modern HA solution.

        I've looked at that Extreamer before...200 bux.

        Have a look at "ToneCast" on this page...

        http://www.vypress.com/

        Free download and I think it's ultra-cheap.
        I've already been streaming uncompressed 16 bit stereo audio from PC to PC across the LAN with this and it works great. There is latency between the server and the clients, but I'm not sure how much there would be (if any) between clients. Since it plays what is playing on the server sound card, it has the delay of having to encode and then transmit the data through the network. It seems like if you didn't listen to the server's sound card that the clients would be darned close to sync'd as they would all experience a similar latency.

        This same tech should be able (with some mod's) to do "answer over ethernet" much the same way that there are "Speakerphone" apps for the PC that broadcast the modem speaker through the soundcard.

        Intercoms via this tech' should easily follow.

        A little more fiddling and you could control the volume of any "tonecast-like" plugin on any client from any client without changing the master output volume of that computer (which is doing things like ATT voices or other HA audio).

        I guess I should point the developer of Tonecast this way or the developers of Homeseer and MainLobby that way. (maybe someone from Homeseer is actually even lurking on this thread?).

        Thanks for listening
        Sean

        Comment


          #19
          I think one of the problems developers struggle with in voice applications is the way audio is done on the PC and the traditional Ethernet network. All streaming technologies have buffering to some extent, which adds delays to the actual signal path, or at least to the playback end.

          In addition, the fact that the Ethernet network is a collision-based domain with packets that may be transmitted and retransmitted (due to collisions) in non-sequential order, obviates the need to reassemble them in the proper order before they can be played, adding delay.

          Historically, voice systems have been basically audio switches with dedicated lines to/from the handset (PBX), thereby avoiding all these problems as it is not a shared medium. The handset wasn't a directional device - the mic and the earpiece were part of the same circuit and there wasn't a distinct transmit channel and receive channel. It was all done on one pair of wires, eliminating feedback and echo problems.

          That factor is the lack of handset at each machine. An open-air mic and speaker will cause some sort of feedback loop based on delay factors of the transmitted and received signals, and echo cancellation causes more delay. We're talking about the kind of stuff you run into during video teleconferencing sessions.

          Finally, if you have Cat5 into a room, and you are only using 2 pair of the 4 pair available (for 10 or 100baseT), why not just hook a normal analog phone to the third pair and run that back to your central PBX? Then you have a phone right by the computer, and integration can occur at the server end, between the PBX system and your main HA or other control systems. This also makes the majority of the parts easily replaceable and relatively cheap...

          It's hard to beat straight wire-speed audio switching vs. networking with all the delays in conversions and reassembly.
          |
          | - Gordon

          "I'm a Man, but I can change, if I have to, I guess." - Man's Prayer, Possum Lodge, The Red Green Show
          HiddenGemStudio.com - MaineMusicians.org - CunninghamCreativeMaine.website

          Comment


            #20
            Thanks Gordon for your reply...interesting...

            In my opinion, the only thing time critical is sync'd audio. As I mentioned above, you could set the server audio back (or not use it at all). By nature, digital audio doesn't need traditional sync stratagies as it is embedded in the audio itself (meaning you probably wouldn't have to maintain any timeclock, workclock or time code as long as you could assure that the start time of the playback or joining of the playback was referenced to the source. It shouldn't really matter if your audio starts 300ms later or joins 300ms...or even a full few seconds after you push the button as long as it is in time once started.
            The buffering is working with faster than real-time data transmission, so again...the delay shouldn't matter as long as the "join point" is the same. Simply put... if you had 4 people start the same digital song at precisely the same time in 4 different countries, it would sync if you could listen to them all at once. The buffer takes care of the collisions assuming the data rate is fast enough to keep the buffer full. The trick is really to get the 4 computers to start or join at exactly the right time and a reference would be needed for this.

            For Telephony and intercom...It really doesn't matter if there are delays. Compared to the stampede of internet phones, the delay over LAN should be a marked improvement.
            I've had people talk to me over the internet on the PC speakerphone (using PC mic and speakers) and it was acceptable...and that was hundreds of miles away. Again...over the LAN should show quite an improvement. Add a USB phone, or even figure out how to use the modem of the client as the handset connection and we then eliminate the annoying speakerphone-ness as well.
            More than 10 years ago I was using voice modems and even then you could use the handset connected to the modem to record your outgoing wave files, so this must not be new tech'.

            Further evidence is all the little shareware programs that are already sending audio over LAN (IP). There already are IP intercoms and I tried a couple...they work great.
            As I mentioned above...that "tonecast" app sends really super quality audio over my LAN. I checked and the shareware price is $49. While sync may be an issue...less demanding tasks such as intercom and telephony over IP should be pretty easy and high quality.

            Of course I'm expressing my humble opinon here. I don't mean to be the guy with 5 posts who thinks he knows everything, but being that this stuff is available on the low-end shareware sites, you would think it could be integrated.

            Thoughts? (besides "Sean S is a jerk")

            Sean

            Comment

            Working...
            X