Voice problems

[Ilan Tochner]

Hi Alexander,

The SDK you referred to discusses the code that needs to be integrated on the client side, which in the case of OpenSim/Second Life means inside the viewer (specifically in the closed-source SLvoice.exe module that is included in all SL-derived viewers that support Vivox voice). It's only connection with the OpenSim server is getting the identity (address/channel) of the Vivox server the viewer needs to connect to as part of the region login process.

The scenario you described where A may not hear B but B can hear A (or vice versa) indicates that OpenSim transferred the correct Vivox server information to both A and B's viewers (otherwise neither one of them would be able to hear the other as they wouldn't be communicating on the same Vivox channel). The fact that one viewer can hear the other while the other can't supports the hypotheses that this is a Vivox issue (either the Vivox server they are connected to or the SLvoice.exe module on one or both these viewers).

I'm asking about location in order to see if the existence of coordinates higher than 256 somehow effects the Vivox solution. Again, this is not on the OpenSim/Second Life simulators side, those servers are not involved with transferring the actual voice data - their role in enabling Vivox voice is limited to providing viewers with the information they need to connect to the proper Vivox channel. If high coordinates are triggering this issue then VarRegions may be indirectly causing it by enabling avatars to be coordinates which the Vivox solution doesn't handle properly. However, if this problem exists in SL and single-region Kitely worlds as well then the high coordinates are not what is triggering this (or at least not the sole reason) and thus VarRegions aren't likely to be the source of this new issue.

[Alexander Zoltai]

Hi Alexander,

The SDK you referred to discusses the code that needs to be integrated on the client side, which in the case of OpenSim/Second Life means inside the viewer (specifically in the closed-source SLvoice.exe module that is included in all SL-derived viewers that support Vivox voice). It's only connection with the OpenSim server is getting the identity (address/channel) of the Vivox server the viewer needs to connect to as part of the region login process.

The scenario you described where A may not hear B but B can hear A (or vice versa) indicates that OpenSim transferred the correct Vivox server information to both A and B's viewers (otherwise neither one of them would be able to hear the other as they wouldn't be communicating on the same Vivox channel). The fact that one viewer can hear the other while the other can't supports the hypotheses that this is a Vivox issue (either the Vivox server they are connected to or the SLvoice.exe module on one or both these viewers).

I'm asking about location in order to see if the existence of coordinates higher than 256 somehow effects the Vivox solution. Again, this is not on the OpenSim/Second Life simulators side, those servers are not involved with transferring the actual voice data - their role in enabling Vivox voice is limited to providing viewers with the information they need to connect to the proper Vivox channel. If high coordinates are triggering this issue then VarRegions may be indirectly causing it by enabling avatars to be coordinates which the Vivox solution doesn't handle properly. However, if this problem exists in SL and single-region Kitely worlds as well then the high coordinates are not what is triggering this (or at least not the sole reason) and thus VarRegions aren't likely to be the source of this new issue.

So, if i read you correctly, there is still the possibility that that Var is, in part, the culprit?

[Ilan Tochner]

Only to the extent that it allows users to be in region coordinates that are greater than 256, which is conceivably something that the Vivox solution may not properly support. Again, please keep in mind that this is assuming that that is indeed what is causing the issues you're reporting, which doesn't seem to be the case given that similar issues exist in Second Life where there are no regions bigger than 256x256.

[Alexander Zoltai]

Only to the extent that it allows users to be in region coordinates that are greater than 256, which is conceivably something that the Vivox solution may not properly support. Again, please keep in mind that this is assuming that that is indeed what is causing the issues you're reporting, which doesn't seem to be the case given that similar issues exist in Second Life where there are no regions bigger than 256x256.

Well...

This never happened in Second Life...

And, it never happened in Kitely, until Var appeared...

[Ilan Tochner]

Graham reported that it did happen in Second Life, see:

It's no consolation but I was watching a VWBPE session on SL this Friday and they had a 5 min Vivox outage across all their venues. The OpenSim sessions used Teamspeak (which caused some initial confusion but worked OK afaik).

Granted this comment doesn't include enough detail to conclude it is the same issue but it sounds similar.

[Alexander Zoltai]

Graham reported that it did happen in Second Life, see:

It's no consolation but I was watching a VWBPE session on SL this Friday and they had a 5 min Vivox outage across all their venues. The OpenSim sessions used Teamspeak (which caused some initial confusion but worked OK afaik).

Granted this comment doesn't include enough detail to conclude it is the same issue but it sounds similar.

I'm very sorry but I don't consider what Graham speaks of as the issue that Arton began this thread for and what I've been trying to understand...

There are "normal" Vivox failures---I experienced many of them in my years over in Second Life...

What Arton and I are talking about is distinctly different.........

[Graham Mills]

Apologies if I have added a layer of misdirection and feel free to ignore this as another in similar vein. However, I note that the Lab has a new Release Candidate viewer with modified Vivox code "that corrects a number of Voice quality and connection issues on both Windows and the Mac". Perhaps this will improve matters when it finds its way into TPVs... one can but hope.

http://wiki.secondlife.com/wiki/Release ... 0.2.311302

(I haven't attempted to discover the original impetus behind the production of the Vivox RC but presume it went beyond the 2-3 issues identified on this page)

[Ilan Tochner]

Thank you Graham for that additional information.

Linden Lab is obviously aware of some issues with the current Vivox SLvoice.exe implementation. It remains to be seen whether the new viewer release candidate will address the issues that were reported in this forums thread as well. We'll have to wait for a new Firestorm (or other TPV that supports OpenSim) to include this updated SLvoice.exe module to see how it affects the reported problems.

[Graham Mills]

Might be worth trying the new Firestorm. Fixes include:

changeset: 48429:6098aa2dc165
date: 2015-09-10 16:16 +0300
user: andreykproductengine akleshchev@productengine.com
summary: MAINT-5584 FIXED On voice disabled parcel or region, speak button remains disabled in ad-hoc and group voice calls - you can hear but not speak.

I appreciate that's not entirely the case here but maybe worth a try?

http://www.firestormviewer.org/firestor ... 7-7-48706/

[Alexander Zoltai]

At the risk of incurring displeasure, I feel I need to repeat something I've already said:

"I'm very sorry but I don't consider what Graham speaks of as the issue that Arton began this thread for and what I've been trying to understand...

"There are "normal" Vivox failures---I experienced many of them in my years over in Second Life...

"What Arton and I are talking about is distinctly different........."

Also, another issue is that I'm unable to use Firestorm, in fact unable to use any viewer but Singularity---it's an issue with the limited resources on my Macbook Air---only 2 Gig memory...

Yet, it isn't my computer causing these problems (which, again, I've never experienced in Second Life and never in Kitely, until Var-regions were put in place), no not my computer since Arton with a top of the line computer is also suffering from them.........