mosbah.abdelkader
2010-Mar-08 10:10 UTC
[asterisk-users] Calculating R Factor and MOS metrics for VoIP
Hello All, MOS and R factor are the two QoS parameters used to estimate VoIP call quality. I have found that they are calculated from other metrics like jitter, latency, packet loss,...etc. But, haven't found any formula or arithmetic rule to calculate them. Do you have an idea about their formulas or an open source that calculates them. Is it possible to interpret them from wireshark. Thanks. -- Please discover scientific miracles of CORAN http://www.55a.net/ -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.digium.com/pipermail/asterisk-users/attachments/20100308/56217ab7/attachment.htm
Marc LEURENT
2010-Mar-08 13:03 UTC
[asterisk-users] Calculating R Factor and MOS metrics for VoIP
Take a look at http://dev.leurent.eu/voip/MOS/ I'v done this a long time ago, hope it will help! ++ Le 08.03.2010 11:10, mosbah.abdelkader a ?crit :> Hello All, > > > > MOS and R factor are the two QoS parameters used to estimate VoIP call > quality. > > > I have found that they are calculated from other metrics like jitter, > latency, packet loss,...etc. But, haven't found any formula or > arithmetic rule to calculate them. > > > > Do you have an idea about their formulas or an open source that > calculates them. Is it possible to interpret them from wireshark. > > > > Thanks. > > > > > > -- > Please discover scientific miracles of CORAN > > http://www.55a.net/
Klaus Darilion
2010-Mar-08 13:27 UTC
[asterisk-users] Calculating R Factor and MOS metrics for VoIP
Am 08.03.2010 11:10, schrieb mosbah.abdelkader:> Hello All, > > > > MOS and R factor are the two QoS parameters used to estimate VoIP call > quality. > > > I have found that they are calculated from other metrics like jitter, > latency, packet loss,...etc. But, haven't found any formula or > arithmetic rule to calculate them.IMO most of these calculations are crap, because you can not tell the voice quality just based on network statistics. The perceived quality also heavily depends on the playback algorithms, e.g. jitter buffer (static? dynamic? size? FFT for faster playback ...) and error correction and error concealment. AFAIK all the available VoIP MOS calculations operate on packet layer, thus they do not include the playback characteristics. regards klaus> > > > Do you have an idea about their formulas or an open source that > calculates them. Is it possible to interpret them from wireshark. > > > > Thanks. > > > > > > -- > Please discover scientific miracles of CORAN > > http://www.55a.net/ >
Steve Underwood
2010-Mar-08 15:54 UTC
[asterisk-users] Calculating R Factor and MOS metrics for VoIP
On 03/08/2010 06:10 PM, mosbah.abdelkader wrote:> Hello All, > > > > MOS and R factor are the two QoS parameters used to estimate VoIP call > quality.You can't calculate MOS. Its an assessment based on a lot of human hearing. Nonetheless, there is a profitable industry in pretending to calculate MOS. The definition of the R factor can be found in one of the ITU G series specs. I can't remember which one, but it should be easy to find.> > I have found that they are calculated from other metrics like jitter, > latency, packet loss,...etc. But, haven't found any formula or > arithmetic rule to calculate them. > > > > Do you have an idea about their formulas or an open source that > calculates them. Is it possible to interpret them from wireshark. >Steve