This forum includes tips for maintaining the best audio quality possible with the Ooma System. If your Ooma system is having issues with dropped calls, static audio or echo, look here for assistance.
#43445 by Mike123
Fri Jan 22, 2010 5:34 pm
I bought and connected one of the Ooma Telo devices the other day. The first call that I made with it resulted in the person on the other end not being able to understand me due to my voice dropping out so much, and we terminated the call because of that. After that, I came here and looked through the forum to see what I could find out.

I ended up running the test at http://www.whichvoip.com/voip/speed_test/ppspeed.html a number of times. What I found out was that apparently my QoS stinks. I tried the test a number of times, without any other internet activity running, and with the Ooma Telo not even connected. It didn't make any difference whether the test was run with the Telo connected or not, and either way the QoS results were worse than any of those that I've seen posted here by anyone else. One thing that I found surprising is that a MagicJack I have works pretty well most of the time if I don't have too much other internet activity going on with the computer at the same time.

My internet connection is DSL from AT&T, and is their Express 1.5 Mbps/384 Kbps offering. The modem is the SpeedStream 4100 that was provided by AT&T when I initially signed up for the DSL service. The modem connects directly to a single computer, with no router being used. Is there anything that I can do to increase my low QoS test results? Since the results are so low even with the Ooma Telo not connected, I'm assuming that changing it's QoS setting would make any difference in regards to call sound quality.

Typical of the test results I got, both with and without the Telo being connected:

VoIP test statistics
--------------------
Jitter: you --> server: 0.2 ms
Jitter: server --> you: 3.6 ms
Packet loss: you --> server: 0.0 %
Packet loss: server --> you: 0.0 %
Packet discards: 0.0 %
Packets out of order: 0.0 %
Estimated MOS score: 4.0

Speed test statistics
---------------------
Download speed: 289920 bps
Upload speed: 305944 bps
Download quality of service: 9 %
Upload quality of service: 19 %
Download test type: socket
Upload test type: socket
Maximum TCP delay: 607 ms
Average download pause: 50 ms
Minimum round trip time to server: 74 ms
Average round trip time to server: 75 ms
Estimated download bandwidth: 1280000bps
Route concurrency: 4.415011
Download TCP forced idle: 0 %
Maximum route speed: 7084864bps
#43446 by murphy
Fri Jan 22, 2010 5:38 pm
There is something seriously wrong. Your download speed is lower than your upload speed. Do you have anything connected to the DSL line other than the DSL modem that does not have a DSL filter in place?
#43450 by Mike123
Fri Jan 22, 2010 6:16 pm
murphy wrote:There is something seriously wrong. Your download speed is lower than your upload speed. Do you have anything connected to the DSL line other than the DSL modem that does not have a DSL filter in place?

No, there's nothing else connected. In addition to the DSL modem, there's only one phone connected to the line that the DSL is on, and that's after one of the DSL filter/splitters.

phone line > DSL filter/splitter > phone on the "phone" outlet of the splitter, modem on the "DSL" outlet of the splitter > computer's network port from modem

Results from http://www.speedtest.net are a very consistent 1.30Mb/s down, 0.33Mb/s up, and 21ms ping.

Some other speed test locations give results basically the same as those from speedtest.net , but on other test sites the results are considerably lower. With some of them, it looks like the transfers get interrupted and restarted, and that results in the lower end results that they display. Some of the locations that a particular test allows you to choose from to connect to for the test do that, and other locations available from the same test work well, giving speeds like those I mentioned above from speedtest.net. Good locations are consistently good when tried, and bad ones are consistently bad. The distance from here of the city selected for the test doesn't seem to be a factor. Some are close and give poor results, and some are half a world away and give good results.
#43464 by davidm
Fri Jan 22, 2010 7:09 pm
Yes that download speed (as measured by the QoS test) is the big problem.

By default I believe the telo does not do any QoS on the download portion but even if it did ~ 210 kbps DL really isn't enough. I would contact your ISP and treat it as a speed issue. Keep a list of the speed tests which are giving you low results. Sometimes connection issues can be intermittent which might explain the on again off again thing. Also keep in mind that I have heard of ISPs setting things up so that speeds to certain hosts (such as speedtests) are higher than other hosts. This might give the customer the impression that they are getting a higher speed.
#43594 by tommies
Sat Jan 23, 2010 4:33 pm
Maximum TCP delay: 607 ms
This should be less than 100ms, There must be some thing that causes this.
One thing that I found surprising is that a MagicJack I have works pretty well most of the time if I don't have too much other internet activity going on with the computer at the same time.
Are you sure that MagicJack software is not running in the background?
#43633 by Mike123
Sun Jan 24, 2010 1:11 am
tommies wrote:
One thing that I found surprising is that a MagicJack I have works pretty well most of the time if I don't have too much other internet activity going on with the computer at the same time.
Are you sure that MagicJack software is not running in the background?


I was wondering the same thing. Since my last post I've done a lot of testing using different combinations of computers, modems, and operating systems to run the VoIP speed test. Having the software that the MagicJack installs on the computer didn't make any difference at all.

There is something in the computer that was used to get the results I posted previously that's causing part of the problem though. It's an old installation of Windows XP that's had a lot of stuff installed in it over the years. When I ran the test with a fresher install of XP, Windows 7, or Linux Ubuntu I always got consistent upload QoS results of 99%, regardless of what the particular hardware configuration was. The download QoS when trying those different operating systems gave different, and pretty consistent results for each of them though. The fresh XP installation had download QoS readings that were the same as the initial old XP installation. Windows 7 was considerably better, and the Linux Ubuntu was better than the Windows 7. I didn't save any of the test results done when trying the Ubuntu, but here's a typical Windows 7 one. All hardware was the same as for the initial test results that I posted above.

VoIP test statistics
--------------------
Jitter: you --> server: 0.1 ms
Jitter: server --> you: 3.6 ms
Packet loss: you --> server: 0.0 %
Packet loss: server --> you: 0.0 %
Packet discards: 0.0 %
Packets out of order: 0.0 %
Estimated MOS score: 4.0

Speed test statistics
---------------------
Download speed: 1001624 bps
Upload speed: 329664 bps
Download quality of service: 58 %
Upload quality of service: 99 %
Download test type: socket
Upload test type: socket
Maximum TCP delay: 576 ms
Average download pause: 13 ms
Minimum round trip time to server: 77 ms
Average round trip time to server: 78 ms
Estimated download bandwidth: 1280000bps
Route concurrency: 1.2779247
Download TCP forced idle: 0 %
Maximum route speed: 6808824bps

I did try the Ooma Telo without the computer connected to it, and the person on the other end said that the sound quality was good. Based on what I found out, part of my problem was something in the old XP install, and part of it the quality of the DSL service. The phone like that the DSL comes in on tends to be noisier then the other line in the house, due to something in the outside wiring somewhere in the neighborhood. I'll be calling AT&T this week to see about moving the DSL service to the other line anyway, so hopefully that improves the VoIP test results I get, and gives me consistent call quality with the Telo.
#43675 by tommies
Sun Jan 24, 2010 6:39 pm
Maximum TCP delay: 576 ms

This is still over the limit(<100ms)

You should call ATT to have a line test on your DSL line(portion outside of the house.)

Your house wiring maybe/maybe not part of the problem, and you could eliminate this by connect your modem directly to the 'test jack' in the DEMARC box outside the house. Also look for some bad connector in your DEMARC box(i.e. rusted wires+screws)
#44009 by Mike123
Tue Jan 26, 2010 11:01 pm
tommies wrote:Your house wiring maybe/maybe not part of the problem, and you could eliminate this by connect your modem directly to the 'test jack' in the DEMARC box outside the house. Also look for some bad connector in your DEMARC box(i.e. rusted wires+screws)

I did try it with the modem connected directly to the box outside. It didn't make any difference in the results. I'll see how it goes after the DSL service gets transfered to the other phone line in the house.
#44360 by rlewis
Fri Jan 29, 2010 4:41 am
I am dealing with similar issues - phone conversation sounds OK at home unit but caller hears only broken sentences. Road Runner - Time Warner is carrier.
Below are my stats - what am I dealing with? Do I need to get a new cable segment run from house to cable provider box at street?

VoIP test statistics
--------------------
Jitter: you --> server: 1.3 ms
Jitter: server --> you: 5.1 ms
Packet loss: you --> server: 0.0 %
Packet loss: server --> you: 0.0 %
Packet discards: 0.0 %
Packets out of order: 0.0 %
Estimated MOS score: 4.0

Speed test statistics
---------------------
Download speed: 14127400 bps
Upload speed: 968392 bps
Download quality of service: 43 %
Upload quality of service: 99 %
Download test type: socket
Upload test type: socket
Maximum TCP delay: 21 ms
Average download pause: 1 ms
Minimum round trip time to server: 19 ms
Average round trip time to server: 21 ms
Estimated download bandwidth: 28800000bps
Route concurrency: 2.0385916
Download TCP forced idle: 14 %
Maximum route speed: 27593680bps
#44364 by rlewis
Fri Jan 29, 2010 5:11 am
Also - as a follow on question. If I redo the cable box to my house connection should I be using RG6 or RG11 shielded cable? I'm seeing both mentioned and my run is pretty long - probably 200 feet at least.

Who is online

Users browsing this forum: No registered users and 10 guests