Announcement

Collapse
No announcement yet.

Am i the only one interested in mumble....

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

  • Am i the only one interested in mumble....

    One of the standard applications for in-game voice communication seems to be ventrilo http://www.ventrilo.com/ (at least for the cs:s masses). Despite a promise of a linux client, Flagship Industries will not be delivering any time soon (see http://www.petitiononline.com/vent4Lin/petition.html for more info).

    Personally, I think the linux client is a lost cause (most people are making do with wine + ventrilo http://appdb.winehq.org/appview.php?iAppId=2169) or the teamspeak linux client. Both of these are proprietary binaries though...

    insert Mumble here http://mumble.sourceforge.net/

    Mumble has been progressing reasonably well of late http://mumble.svn.sourceforge.net/vi...y=rev&view=log (although not at the breakneck speed of PTS). Overlays work with most games, positional sound support has been implemented for battlefield 2 and WoW, it offers cross platform support and it is licensed under GPL...

    Hardy seems to have picked up Mumble in its repositories too - Main and PPA (coinciding with Ubuntu defaulting to Pulse Audio). Perhaps it's time open source developers started integrating positional voice chat through mumble... openarena, nexuiz and warsow spring to my mind. I'm not sure how simple/difficult this integration would be but it would be the best way to further mumble development/adoption/awareness.

    A brief overview of Mumble on Phoronix would help too

    Has anyone else been thinking along these lines?

  • #2
    It is indeed refreshing to see this. Last time I checked out Mumble it was fairly early on, and upon the failure to put out a Linux client compatible with ALSA (for TeamSpeak) or a Linux client at all (speaking about Ventrilo), the Linux gaming community does deserve to have an up-to-date VoIP program for our gaming fix. Personally TS has worked really well for me, but being it only OSS and ts lack of ALSA (and dmix) support has caused that only people with hardware mixing sound chips can use it effectively. Now with PA things will be much different, as it matures (PA still doesn't play very nice with capture). It's been a while since I last checked out Muble, I must confess (about a year ago?) but from the sound of it, it has been making real strides. Will have to check it out soon.

    Thanks for the heads up.

    Comment


    • #3
      Good to know about a free software alternative to teamspeak.

      Comment


      • #4
        resurrecting this thread with an interesting post courtesy of icculus, i.e. VoIP support for ioquake3... http://lists.ioquake.org/pipermail/i...ay/002722.html

        Comment


        • #5
          My only concern about Mumble is the current state of affairs with Pulse Audio and software mixing, especially input. This a critical feature to have in PA before it could be really useful. As it is right now, and being most sound chips on the market single stream DSPs, PA is a very welcome feature by offering its software mixing capability plus many other features (merging together more than two sound cards, and per-application volume settings, etc). Its capture support still is flaky (and I have not been able to get it working on any computer running with it). Without software mixing using VoIP apps is difficult or even useless, especially if you have to choose whether the game or the VoIP app will have sound... yes I'm looking at you TeamSpeak! (and what is the point of having a VoIP app, if you don't actually hear your friends?). I'm sure that being Mumble native ALSA it is susceptible to work with dmix, however I have not tried it. Very few things work with dmix (in my experience) even if they have native ALSA support.

          Comment


          • #6
            Originally posted by Thetargos View Post
            My only concern about Mumble is the current state of affairs with Pulse Audio and software mixing, especially input. This a critical feature to have in PA before it could be really useful. As it is right now, and being most sound chips on the market single stream DSPs, PA is a very welcome feature by offering its software mixing capability plus many other features (merging together more than two sound cards, and per-application volume settings, etc). Its capture support still is flaky (and I have not been able to get it working on any computer running with it). Without software mixing using VoIP apps is difficult or even useless, especially if you have to choose whether the game or the VoIP app will have sound... yes I'm looking at you TeamSpeak! (and what is the point of having a VoIP app, if you don't actually hear your friends?). I'm sure that being Mumble native ALSA it is susceptible to work with dmix, however I have not tried it. Very few things work with dmix (in my experience) even if they have native ALSA support.
            Pulseaudio is still under heavy development and I'm sure they will bring the input up to standard as soon as they can. With games I'm more annoyed about the fact many still use OSS and refuse to start unless they can grab the sound output. That and openal's lack of pulseaudio support.

            Comment


            • #7
              Originally posted by Aradreth View Post
              Pulseaudio is still under heavy development and I'm sure they will bring the input up to standard as soon as they can. With games I'm more annoyed about the fact many still use OSS and refuse to start unless they can grab the sound output. That and openal's lack of pulseaudio support.
              The first sort of works okay under PA's OSS emulation. The second can be easily sorted out by making OpenAL think it's talking to OSS or ALSA depending on what else the title's doing.

              Comment


              • #8
                Originally posted by Svartalf View Post
                The first sort of works okay under PA's OSS emulation. The second can be easily sorted out by making OpenAL think it's talking to OSS or ALSA depending on what else the title's doing.
                Ah I didn't know about PA's OSS emulation thanks for pointing that one out I'll give it a go at some point. OpenaAl on the other hand I've battled with trying to get it to use SDL which can use PA but OpenAl just throws up errors (it is compiled with SDL support).

                Comment


                • #9
                  If ALSA is so cumbersome to work directly with, why not use some sort of HAL (SDL making use of OpenAL or PA directly)?

                  Now, I know OpenAL itself is undergoing heavy development as well, especially with its new Sampling Implementation (or whatever its called), which is more efficient and what not. IIRC one of the ultimate goals of PA is to serve as a sort of HAL for sound devices so developers code for it instead of the underlying "bare metal" API (OSS/ALSA) and ensure compatibility, however to achieve this, PA has to first become the defacto standard Sound backend for Linux (as X is for graphics, yes I know I must seem like a scratched record by now), to really work this way.

                  Comment


                  • #10
                    Mumble is indeed a great alternative to TS and Ventrilo and since the latest release (1.4) they also reworked the PTT (push to talk) progress to work without xevie. (afaik xevie is bugged in Hardy for some reason).

                    I'm also running my own Murmur Server for quite some time now.

                    Comment


                    • #11
                      Originally posted by Svartalf View Post
                      The first sort of works okay under PA's OSS emulation. The second can be easily sorted out by making OpenAL think it's talking to OSS or ALSA depending on what else the title's doing.
                      I tried PA's OSS emulation today and it's not really a viable solution as I got a lot of popping/crackling noise with it. Hopefully it'll improve and become a viable solution soon enough.

                      If ALSA is so cumbersome to work directly with, why not use some sort of HAL (SDL making use of OpenAL or PA directly)?

                      Now, I know OpenAL itself is undergoing heavy development as well, especially with its new Sampling Implementation (or whatever its called), which is more efficient and what not. IIRC one of the ultimate goals of PA is to serve as a sort of HAL for sound devices so developers code for it instead of the underlying "bare metal" API (OSS/ALSA) and ensure compatibility, however to achieve this, PA has to first become the defacto standard Sound backend for Linux (as X is for graphics, yes I know I must seem like a scratched record by now), to really work this way.
                      PA is slowly becoming the defacto standard because with both Fedora and Ubuntu using it PA has a fair percentage of the Linux user base. So given some more time I think more distro's will follow suit and start using it.

                      Comment


                      • #12
                        interesting - changelog for mumble 1.1.4 says os x support has been introduced and positional audio on all audio backends, in addition to a bump upto qt 4.4 and improved pulseaudio support.

                        you might want to check out this old link detailing whats coming up with pulseaudio 0.9.11...

                        http://0pointer.de/blog/projects/pulse-glitch-free.html

                        Its a post by a pa dev that details the traditional sound playback model n a glitch-free model, along with the changes to pa so that it adopts a glitch free model.

                        also read the comments, definitely an interesting read.

                        bluetooth support under pa is also a gsoc project this year http://code.google.com/soc/2008/blue...18999748B418AE

                        along with the openal sampling work, I think this year might be one of the more interesting wrt linux audio.
                        Last edited by hmmm; 06-03-2008, 10:57 PM.

                        Comment


                        • #13
                          Any one know how well does Mumble and dmix work together?, so I can disable PulseAudio while playing a game with Mumble as the VoIP app. I'd like to give it a try, I have already setup the server and am able to connect with the client, but have not tested it with other applications running on this ATi Azalia card on my laptop (on my desktop, since I've got a Live! card, that is not an issue, and I'm worried about this non hwmixing capable and how dmix might work (or not) with Mumble.

                          Comment


                          • #14
                            Originally posted by hmmm View Post
                            you might want to check out this old link detailing whats coming up with pulseaudio 0.9.11...

                            http://0pointer.de/blog/projects/pulse-glitch-free.html

                            Its a post by a pa dev that details the traditional sound playback model n a glitch-free model, along with the changes to pa so that it adopts a glitch free model.

                            also read the comments, definitely an interesting read.
                            I hadn't seen that yet, very interesting. Thanks for the heads up.

                            Comment


                            • #15
                              I would love to have an application like mumble on linux. TeamSpeak3 seems to take forever and no idea what the status is for Ventrilo. One big issue for me is that QT is required for the server and I don't want to install X on the server for just one application Any idea if they plan to remove the QT dependency for the server?

                              Comment

                              Working...
                              X