VoIP Bandwidth
-
We will soon be allowing more work-from-anywhere folk to make calls using the company PBX (FreePBX hosted on Vultr with Twilio SIP trunks). I'd like to create some guidelines for users along the lines of "you must have X bandwidth available with your Internet connection to have a chance at making calls."
According to this , for a g711 call 87.2 Kbps is needed, and since traffic moves in both direction, you'd need 175 Kbps for bandwidth for a single call. Obviously, if I tell user the minimum bandwidth they need is truly the minimum bandwidth, then nothing will work, since users will inevitably be using their network for all sorts of things while trying to make calls.
What do you folks tell these users? I'm thinking "broadband with at least 1 Mbps downstream and upstream."
-
Don't be a jackass, turn off any streaming services, podcast, youtube, pandora/spotify and in most cases "you'll be find little user".
-
The bandwidth requirements for VoIP is so trivial that unless they are getting substantially slower than dial up speeds for an remote site, that this shouldn't even be a concern.
-
It doesn't really work that way. It's all about what else they are doing. You can do solid VoIP at 50Kb/s. Bandwidth is not at all the issue.
-
@dustinb3403 said in VoIP Bandwidth:
The bandwidth requirements for VoIP is so trivial that unless they are getting substantially slower than dial up speeds for an remote site, that this shouldn't even be a concern.
Well no, you realistically need dial up speeds at least
-
@scottalanmiller said in VoIP Bandwidth:
@dustinb3403 said in VoIP Bandwidth:
The bandwidth requirements for VoIP is so trivial that unless they are getting substantially slower than dial up speeds for an remote site, that this shouldn't even be a concern.
Well no, you realistically need dial up speeds at least
I did say "substantially slower"
-
Aren't latency and jitter the most likely concerns with remote users? Assuming they have the standard minimum speed available?
-
@nashbrydges said in VoIP Bandwidth:
Aren't latency and jitter the most likely concerns with remote users? Assuming they have the standard minimum speed available?
Yes, those are the real concerns but it is not within the users or @EddieJennings control to fix these issues. These are for the ISP to address.
-
@dustinb3403 said in VoIP Bandwidth:
@nashbrydges said in VoIP Bandwidth:
Aren't latency and jitter the most likely concerns with remote users? Assuming they have the standard minimum speed available?
Yes, those are the real concerns but it is not within the users or @EddieJennings control to fix these issues. These are for the ISP to address.
Exactly, which I think should be part of establishing realistic expectations along with communicating minimum requirements for speed. Would hate to have remote users expect to have VoIP available but it turns out to be unusable.
-
@nashbrydges said in VoIP Bandwidth:
Aren't latency and jitter the most likely concerns with remote users? Assuming they have the standard minimum speed available?
Yes, and packet loss like you see on Comcast networks all of the time.
-
What you really want is end users to generate MOS scores from home.
-
@scottalanmiller said in VoIP Bandwidth:
@nashbrydges said in VoIP Bandwidth:
Aren't latency and jitter the most likely concerns with remote users? Assuming they have the standard minimum speed available?
Yes, and packet loss like you see on Comcast networks all of the time.
Packet loss or packet theft. . . with Comcast you can't be certain. . .
-
@scottalanmiller said in VoIP Bandwidth:
What you really want is end users to generate MOS scores from home.
What do you recommend for that?
-
@nashbrydges said in VoIP Bandwidth:
@scottalanmiller said in VoIP Bandwidth:
What you really want is end users to generate MOS scores from home.
What do you recommend for that?
Pingtest.net used to be great but they got rid of it
-
I don't have a good tool for that right now, maybe @JaredBusch has one that he likes?
-
Here is some interesting information.
-
Back from lunch. I agree with the idea of finding a tool that users can run to generate a MOS.
-
This seems interesting: https://sourceforge.net/speedtest/ (EDITED)
-
@eddiejennings The test you linked is from sourceforge
-
@dustinb3403 So it is