Voice Support for the Linux Client

I couldn't get Wine to work at all under 64-bit. I switched to 32-bit and Wine works great.

Adamas

I was able to get it working in a 32-bit enviroment... but only after pointing ESD into my laptop's main card and not the Logitech headphones I have. I guess it's because there's no master volume control, and thus it's shutting out the mixer itself. I'll find out more.

are voice binaries still missing in first look viewer for linux?
i find it quite interresting that it works with wine but i think thats not an option as a cure from a technical POV. and even if it works its exactly what i undertand as yak shaving.
will voice be available natively soon? i dont think about talking - to hear anything would do the job for me.

After some work, i got wine sollution to work here... I can hear (with jump and lag, but i can) and I can talk (and the other person told me that heard me very well, without fails).

There is still a problem of connecting and disconnection all the time from server...

Is the actual linux blob worse than that? If it is not, pleeeease Tofu release it!

Sorry I got this out so late. I've had big trouble with aRts and had to replace it. I apologize for the length, but I wanted to give enough details to minimize problems.

Please let me know if I'm missing anything.

*Damn, damn, damn, damn, damn, damn, damn, damn, damn, damn!*

Won't let me post! Probably too big. (You'd think they'd tell you that!) Well, I uploaded it to one of my websites (one I don't use much) and here is the URL:

http://www.paganstead.com/extras/sl_voice_linux.txt

Please let me know if it works (or more important, doesn't work!).

Adamas
P.S. Ignore those url tags. That's Linden Lab's software putting those in, not me.
P.P.S. Tofu, I'd like to try this with the Linux blob. Please?

Isn't it possible to put the linux blob just on the test version (Windlight)? Who is using Tst version, knows that not all should work fine...

And I guess we could help a trying it, and fedbacking where and how it works and how it doenst

How about a testing program for those of us who put out the effort, did the work, and actually managed to get the Windows version working? I imagine the feedback from that would be very useful to LL, Vivox, and Users and would shed better light on what it will take to make it work for everyone.

Adamas

I heard from Tofu Linden in Jira that linux voice will not make it in the next version. It might make the version after that though.

quoting Tofu Linden-

"
Linux Voice support is in QA right now. It might miss the 1.19.0 boat, but should arrive subsequently.
"

I just wanted to post an addendum to my instructions.

I had a problem with my on-motherboard sound hardware where it would feedback anything I was hearing in the headset. If someone else tried to talk while my talk button was down, it would feed it back to them. Very annoying. My fix was to replace it with a new soundcard I found at MicroCenter for about 60 bucks, a Turtle Beach somethingorother. It works perfectly and all I had to do was play with the controls in kmix to get it to voice out.

So, if your system is feeding back to SL, the problem is very likely bad/crappy sound hardware. (Like mine.) Find something (it doesn't have to be really expensive) to replace it with.

After installation/configuration you may need to play with the buttons and sliders in kmix (or whatever your system uses) to get it to play nice.

Adamas

Did anyone found any sollution for the coppy voice problem, when using wine?

As I can see googling around, that is not a problem that only happens with SLVoice.exe, but also with many other software that uses sound on wine... Usualy is related to the frequenciy of 44Khz, and chang that on winecfg to 22 or 8khz do the trick, but with SLVoice that does not helped.

I've noticed that when there's choppiness on voice for me, everyone else complains about it too. It appears that LL's voice server(s) are overloading at those times. Also, sometimes one person will be choppy and everyone else is clear, suggesting that person has a computer or network problem. Everyone says I sound clear.

I seem to be getting the same service everyone else does. PulseAudio, for some reason, defaults to 48000 when it starts up, but Wine is set to 44100 and everything works great.

Adamas

I am able to hear people int he Linux client with wine and SLVoice.exe but the only way people can hear me is if I yell into my mic and also I am experiencing the same problem with losing connection to the voice server. I have a Realtek HD audio device which is built onto my mother and part of the nforce chipset. It works fine for skype and and everything else just doesn't seem to get loud enough for people to hear on SL. I checked all my mixer settings and my mic is even up all the way. Any suggestions of how to fix this would be grealy appreciated.

Try checking all the volume settings in the SL client. There are three settings groups: the main volume list, accessed from the control at the lower-right, the pop-up box that shows the speakers in your vicinity (you can adjust these separately for each speaker) and the volume control in Preferences / voice chat / device settings.

Adamas

As an added note, onboard audio is crappy. There is something about the SL client and voice that makes it a bad match for SL. Motherboard mfgrs don't spend a whole lot of money on stuff they add to a motherboard. This built-in crap is always the first thing to go.

I replaced my onboard audio with a (relatively) inexpensive Turtle Beach somethingorother card (about $60US) and it works great. The only problem I had with it was fixed when I finally figured out how to disable the onboard audio from the BIOS, and did so.

Adamas

For al the fedora core 8 users with pulse audio enabled:
update your wine
switch in winecfg to OSS audio only!!!
Then just run c:\program files\secondlive\slvoice.exe .... (rest of the arguments)
Start secondlife
enable voice
And you have voice, as clear as windows.

The wine update did it, alsa was not working and choppy sound in esd.

Cheers,

Hilby

Audio Capture fixed in release 0.9.51 of Wine:
What's new in this release:
- A bunch of WinHelp improvements.
- Better Japanese font support.
- A ton of rpcrt4 fixes.
- Several Alsa capture fixes.
- Improved support for screen resolution changes.
- Lots of bug fixes.

SLVoice.exe should capture audio now.
Upgrade wine to 0.9.52 or higher.

Winehq just released 0.9.53 this morning. OpenSUSE has it available now.

I just updated and tested it with sl. It works perfectly.

Adamas Carter

Ok trying the new wine 0.9.53, I have voice working under 6.x crossover, prefer to run it outside of crossover. Plus it tends to get choppy a lot, hoping this new wine would fix that.

The 0.9.53 does not pick up the esound / esd driver under the audio tab, the version of wine in crossover 6.x finds it just fine.

Any ideas ???????

As far as I can tell, Wine only tells you it sees ALSA, but works with ESD (and PulseAudio) anyway.

I don't know why, it just works.

Adamas

Using 0.9.53 wine slvoice loads fine connects everything is good there.

I have confirmed that esd is running. But under edit, preferences, voice chat, device settings only the alsa stuff shows up, no esd

and of course the mic is not working as well

still works ok in the crossover 6.x, wine 0.9.34

Ideas ?????

Are you using PulseAudio?

I found that ALSA and ESD alone don't work, you have to run PulseAudio also.

I posted a detailed procedure that I worked out to make voice happen:

http://www.paganstead.com/extras/sl_voice_linux.txt

It's not formatted, but you can just download it and open it with kwrite or something.

Adamas

P.S. Ignore the url tags in the line, LL puts them there, not me.

My setup
pulseaudio 0.9.7
wine 0.9.52
with winecfg configured to use only oss

This is the first setup without choppy sound

Following Adamas post, it's working great now, no choppiness in incoming or outgoing. Any choppiness now I believe is just the slvoice being slvoice, and in general just typical voice over IP loss.

My final config for voice / audio is esound (I have to use this for music and other sounds to play write in sl client), pulseaudio, wine 0.9.53, wine drivers set to alsa. I've not tried the oss, I'll experiment with it and post back in time

I've liked the pulseaudio so much if installed the plugins for the over audio / video apps I use as well.

Now if only there was away for the microphone to be shared with slvoice and IDJC

Nice to see there is a way...shame its needing to use WINE....but its a start

Nice one Adam.

I suspect the Linux voice binary will require this same kind of setup (when Tofu finally releases it). If so, it should be easy to use it in place of the Wine stuff. When I can get ahold of it, I'll set it up and let everyone know. But, at least we have something WE can use!

Also, I check with windows and mac users I'm talking to, and they confirm they get the same choppiness from the same people at the same time as I do. I think maybe sl's voice servers are sometimes not up to the task of handling that many voice users. Maybe they'll upgrade or add more servers.

Adamas