[asterisk-biz] Experimental/new VoIP rate search engine.

[Kristian Kielhofner]

On 1/4/09, Nitzan Kon <nk3569 at yahoo.com> wrote:
> --- On Sun, 1/4/09, Kristian Kielhofner <kristian.kielhofner at gmail.com> wrote:
>
>  >   Getting back on topic, will there be any sort of ranking
>  > for quality?
>
>
> I would love to maintain a separate table/search based on
>  quality. The real question though is how do you quantify
>  "quality". Maybe through user reviews, although those tend to
>  be too subjective to really rely on long term. i.e. a user
>  might think their VoIP provider has the best quality or worst
>  quality - but in actuality it's the only provider they've tried.
>
>  Maybe give points for quality of support, time in business,
>  reported call quality, etc. and come up with some type of
>  formula to sum it all up to a ranking..
>
>  Will have to think about this one. :)
>
>  -- Nitzan
>  http://www.comparevoipproviderrates.com
>

Nitzan,

  There are many factors but I would be willing to help you with this.
 For US based providers, for example, my utility recqual could be used
to provide a pretty clear overall picture of actual call quality.  I
know my shameless plug here might seem suspect but  I swear - I'm no
where near that calculating!

  My concern would be various tricks providers could use to fool any
IP only automated tools.  There are classic examples of providers
providing gateways to ping, etc that *clearly* have ICMP optimized and
may even be in different address space than the actual equipment used
to terminate the call (read: RTP endpoint addresses, etc).

  Any provider that uses IAX, for example, is most likely proxying the
media and converting it to SIP/RTP at some other point in the network.
 IP based tools have no visibility into what happens to that media
once it isn't IAX anymore.

  An audio analysis tool like recqual cannot easily be fooled by such
practices because it tests actual audio quality through to the PSTN.
Any underlying problems (network, host, implementation) will be
reflected in the actual audio quality.

  I would be willing to provide some expertise and maybe even hosting
for this effort.  Listing in the directory could require a minimum
number of recqual test calls to determine a baseline of quality.  As
long as the test runs were done with some consistency the results
shouldn't be skewed that much.

  Certainly some providers would score better than others based purely
on geography (proximity to the machine running recqual).  Of course
there are ways around this too but it would require a much larger,
organized effort. :)

-- 
Kristian Kielhofner
http://blog.krisk.org
http://www.submityoursip.com
http://www.astlinux.org
http://www.star2star.com