pepi3
join:2013-02-26
Hartford City, IN |
pepi3
Member
2013-Feb-27 9:43 am
Comcast JitterWhat is the usual fix for Comcast Jitter? My 15/2 network is fast but full of Jitter. Causes my IP phone to go nuts. Most Comcast Techs will look at you like your crazy when you tell them there is too much Jitter on the line  |
|
| |
AnonMan
Anon
2013-Feb-27 9:59 am
No such thing as a "usual fix". It could be caused by any number of 500 different things  That said more details will be needed. Modem signal levels and noise level. Pings/trace routes to show ping and hops through your network and comcast etc. Does the issue happen all the time or only sometimes etc. Much more details needed |
|
|
pepi3
join:2013-02-26
Hartford City, IN |
pepi3
Member
2013-Feb-27 11:08 am
Thanks!  » www.pingtest.net/result/ ··· 6230.pngDownstream DCID Freq Power SNR Modulation Octets Correcteds Uncorrectables Downstream 1 1 579.00 MHz 4.55 dBmV 36.17 dB 256QAM 670239166 8 0 Downstream 2 2 561.00 MHz 3.84 dBmV 36.17 dB 256QAM 262305565 26 0 Downstream 3 3 567.00 MHz 4.53 dBmV 36.17 dB 256QAM 225599572 326 832 Downstream 4 4 573.00 MHz 3.72 dBmV 35.97 dB 256QAM 431830172 2409 935 Downstream 5 5 537.00 MHz 3.68 dBmV 35.78 dB 256QAM 404265486 2 0 Downstream 6 6 543.00 MHz 5.06 dBmV 36.17 dB 256QAM 407011167 8 0 Downstream 7 7 549.00 MHz 4.76 dBmV 35.97 dB 256QAM 302197547 15 0 Downstream 8 8 555.00 MHz 3.62 dBmV 35.78 dB 256QAM 327346236 5 0 Reset FEC Counters Upstream UCID Freq Power Channel Type Symbol Rate Modulation Upstream 1 7 31.60 MHz 36.75 dBmV DOCSIS2.0 (ATDMA) 5120 kSym/s 64QAM Upstream 4 1 37.50 MHz 37.50 dBmV DOCSIS1.x (TDMA) 2560 kSym/s 16QAM Status System Uptime: 1 d: 3 h: 49 m Computers Detected: staticCPE(1), dynamicCPE(1) CM Status: Telephony-DHCP Time and Date: Mon 2013-02-25 19:32:00 Interface Parameters Interface Name Provisioned State Speed (Mbps) MAC address LAN Enabled Up 100(Full) CABLE Enabled Up ----- MTA InProgress Down ----- |
|
 pflogBueller? Bueller?
MVM
join:2001-09-01
El Dorado Hills, CA |
pflog
MVM
2013-Feb-27 11:56 am
The signal looks fine. Looks like channel 4 has some minor issues and channel 3 less so, but I don't think those would necessarily cause bad jitter. ~2k uncorrected errors in 24 hours isn't much.
You might want to call support and ask them for your uSNR (upstream signal to noise ratio) value. It should be 30 or higher, the higher the better. |
|
pepi3
join:2013-02-26
Hartford City, IN |
pepi3
Member
2013-Feb-27 12:25 pm
Where I work we have an AT&T DSL line and it is so much cleaner than my home Comcast  |
|
pflogBueller? Bueller?
MVM
join:2001-09-01
El Dorado Hills, CA |
pflog
MVM
2013-Feb-27 12:40 pm
said by pepi3:Where I work we have an AT&T DSL line and it is so much cleaner than my home Comcast
What site is that? And yes, DSL usually has less jitter than HFC connections. It's just the nature of the beast. Then again, good luck getting more than 18Mbps out of DSL in the US |
|
 gar187erI DID this for a living
join:2006-06-24
Seattle, WA |
to pepi3
jitter and ping will always be higher then dsl, as pflog said.
your test also said 100% of packets arrived in the correct order, so i would question your test.
have you run a sustained ping test?
have you trouble shot anything else in your network yet? |
|
 NightfallMy Goal Is To Deny Yours
MVM
join:2001-08-03
Grand Rapids, MI |
said by pepi3:And yes, DSL usually has less jitter than HFC connections. It's just the nature of the beast. Then again, good luck getting more than 18Mbps out of DSL in the US
said by gar187er:jitter and ping will always be higher then dsl, as pflog said.
Sorry, but these two statements are pure bullcrap. I can tell you that it depends on the ISP. DSL can have just as many incidents of "jitter" or varying pings as cable can. It depends on your location, ISP, and so on. To say that "jitter will always be higher" on cable connections really is an uninformed statement. Both DSL and Cable are shared mediums, and I have seen varied pings on both cable and dsl providers. |
|
pflogBueller? Bueller?
MVM
join:2001-09-01
El Dorado Hills, CA |
pflog
MVM
2013-Feb-27 1:41 pm
said by Nightfall:Sorry, but these two statements are pure bullcrap.
I can tell you that it depends on the ISP. DSL can have just as many incidents of "jitter" or varying pings as cable can. It depends on your location, ISP, and so on. To say that "jitter will always be higher" on cable connections really is an uninformed statement. Both DSL and Cable are shared mediums, and I have seen varied pings on both cable and dsl providers.
Obviously jitter can be introduced anywhere, and yes if the DSL aggregation point is overloaded it can cause jitter. I should have clarified that the jitter induced in the "last mile" for HFC will always be worse than the last mile (to the CO/RT) for DSL. The same could be said for FiOS. But in reality and in most of the cases I've seen (granted, my data points are somewhat dated) DSL usually has less jitter. |
|
NightfallMy Goal Is To Deny Yours
MVM
join:2001-08-03
Grand Rapids, MI |
said by pflog:said by Nightfall:Sorry, but these two statements are pure bullcrap.
I can tell you that it depends on the ISP. DSL can have just as many incidents of "jitter" or varying pings as cable can. It depends on your location, ISP, and so on. To say that "jitter will always be higher" on cable connections really is an uninformed statement. Both DSL and Cable are shared mediums, and I have seen varied pings on both cable and dsl providers.
Obviously jitter can be introduced anywhere, and yes if the DSL aggregation point is overloaded it can cause jitter. I should have clarified that the jitter induced in the "last mile" for HFC will always be worse than the last mile (to the CO/RT) for DSL. The same could be said for FiOS. But in reality and in most of the cases I've seen (granted, my data points are somewhat dated) DSL usually has less jitter. I used to do contract work for a couple different ISPs in the past, so you will apologize if I came across harshly. There is this long standing myth that people have about DSL and being "connected to the backbone" which is just false. As you stated, the "last mile" for cable is susceptible to more issues than DSL, but the simple fact of the matter is that they are both shared mediums. I have seen just as many DSL providers have issues as I have cable providers. It really depends on your area. |
|
pflogBueller? Bueller?
MVM
join:2001-09-01
El Dorado Hills, CA |
pflog
MVM
2013-Feb-27 1:55 pm
said by Nightfall:I used to do contract work for a couple different ISPs in the past, so you will apologize if I came across harshly. There is this long standing myth that people have about DSL and being "connected to the backbone" which is just false. As you stated, the "last mile" for cable is susceptible to more issues than DSL, but the simple fact of the matter is that they are both shared mediums. I have seen just as many DSL providers have issues as I have cable providers. It really depends on your area.
No, I understand. I should have qualified my statement. I experienced exactly what you're talking about on SBC DSL back in ~2004-2005. They were bottlenecked at the CO backhaul iirc and it did cause jitter and higher latency. I'm not sure how susceptible Verizon's FiOS is to this, but at least when I had it in Delaware it was rock solid. My smokeping graphs were amazing compared to Comcast. |
|
pepi3
join:2013-02-26
Hartford City, IN |
to pflog
» myspeed.visualware.com/i ··· voip.phpAnd yes guys, I've had a tech at my home twice now. Last time he put in a new ARRIS modem. He said his equipment shows my line is fine? He seems to think if I get close to my 15/2 that all is OK. My VOIP and Video streaming is awful. On my phone I can hear the other person very well but they say I'm breaking up. |
|
pflogBueller? Bueller?
MVM
join:2001-09-01
El Dorado Hills, CA |
pflog
MVM
2013-Feb-27 1:56 pm
said by pepi3:On my phone I can hear the other person very well but they say I'm breaking up.
I would definitely call and ask what your upstream SNR is, given that statement. |
|
pepi3
join:2013-02-26
Hartford City, IN |
pepi3
Member
2013-Feb-27 2:02 pm
I will do this ... Thanks! |
|
 ptrowskiGot Helix?
Premium Member
join:2005-03-14
Woodstock, CT |
to pepi3
Given that it is your voice breaking up I would tend to think you should set up some qos. Typically that is an upload issue. Are you or the family doing anything bandwidth intensive at the time? |
|
kdouglas10
Premium Member
join:2002-10-04
Jacksonville, FL |
Your voice breaking up? What about your router? Are you on WiFi? Cordless phones, microwaves, other things operating on 2.4 GHz? If dual-band router try shifting to 5.0 GHz for your VOIP and see if that helps. If it does, track down WiFi interference, maybe someone using the same channel next door, or across the street?
Have you cold-booted (not just turned it on/off) your phone recently? (It makes a BIG difference on my iPhone, particularly after upgrading any apps). |
|
pepi3
join:2013-02-26
Hartford City, IN |
to ptrowski
I've done this on my wireless router. Doesn't seem to make any difference? |
|
kdouglas10
Premium Member
join:2002-10-04
Jacksonville, FL |
....and as I think about it now, cold boot both router and modem. Pull plug on both; let modem sit at least 15 seconds (Comcast sez 60 seconds), plug-in and allow to establish sync, THEN plug router back in.
You have probably done this already to clear registers in both modem and router, but just in case not. (Thinking of some of my friends who proudly announce they have been running 120+ days without a reboot, even thought they are DS slow - LOL) |
|
ptrowskiGot Helix?
Premium Member
join:2005-03-14
Woodstock, CT |
to pepi3
What VoIP provider are you using? |
|
kdouglas10
Premium Member
join:2002-10-04
Jacksonville, FL |
If none of that helps, then back to what pflog sez re: S/N ratio. My Comcast tech (who rolled out here 3 times to get my S/N above -37db & finally ran an amplified "homerun" from my distribution box) swears that -37db is a **minimum** signal for reasonably reliable speeds and video streaming. I recognize that jitter is the issue, and some of the early on figures you showed were meaningless to me, but testing using the premium tools on this site shows me with extremely low jitter (now) and consistent down of 18-24MHz (I think downstream is dependent on usage on my block) and 4 MHz upstream while paying for the same speed (15/2) that I think you have.
I feel your pain. Stay on it! |
|
pepi3
join:2013-02-26
Hartford City, IN |
to ptrowski
Now, here is something else I forgot. I had Comcast Voice before and bought a OBi unit. I had both hooked up and both were working great. When I called Comcast and had them remove my Voice that's when it all when south. I've tried GoogleVoice and CALLCentric with the same results. I took my OBi box to my nephews home with DSL and it works fine.
And yes, I've reset everything without any luck. I've also tried two different cable modems and two different wireless routers.
Thanks guys! |
|
 EGThe wings of love
Premium Member
join:2006-11-18
Union, NJ |
to kdouglas10
said by kdouglas10:If none of that helps, then back to what pflog sez re: S/N ratio. My Comcast tech (who rolled out here 3 times to get my S/N above -37db & finally ran an amplified "homerun" from my distribution box) swears that -37db is a **minimum** signal for reasonably reliable speeds and video streaming.
FWIW, pflog  was referring to the "uSNR" (upstream Signal to Noise Ratio). Even if the downstream (forward path) SNR looks good and is in spec, there may be noise ingress in only the upstream (return path) carrier/s operating frequency range. |
|
| |
to pepi3
So I opened your docx and found that you have no packet loss but you have "discards" I can only guess the difference is that discards are packets that have been rejected by the network due to congestion. If this is the case then powerboost might be your culprit as I have troubleshot something similar with Skype.
One thing to try is to rate-limit your upstream based on what you know to be your sustained upstream, ie 16/2 or 25/4 50/10 This will stop any application from infringing on the upstream "excessive burst" zone of the powerboost rule and the successive discards to follow after the excess token bucket has been exhausted.
Please let me know if I'm accurate on the test's definition of discards and of course let me know if this helps. |
|
pepi3
join:2013-02-26
Hartford City, IN |
pepi3
Member
2013-Mar-4 7:34 am
"One thing to try is to rate-limit your upstream based on what you know to be your sustained upstream, ie 16/2 or 25/4 50/10 This will stop any application from infringing on the upstream "excessive burst" zone of the powerboost rule and the successive discards to follow after the excess token bucket has been exhausted."
Not sure how to do this ??? Not real up on network stuff. |
|
MikroTik RB750G
Cisco DPC3941
|
to ptrowski
said by ptrowski:Given that it is your voice breaking up I would tend to think you should set up some qos. Typically that is an upload issue. Are you or the family doing anything bandwidth intensive at the time?
I agree, that's where you need to look first. If you go and look at the graphs/test results. It clearly shows "Bandwidth probably too low ....", packet discards and all the jitter is on the upstream. That's a classic indication of the upstream data getting clobbered, the RTP stream is competing for too little bandwidth. It looks to me like the jitter is being introduced locally by your router as packets are buffered as they are exiting the WAN interface. That's not a Comcast/network problem. |
|
pepi3
join:2013-02-26
Hartford City, IN |
pepi3
Member
2013-Mar-4 8:02 am
I've been through 3 different cable modems and two different wireless routers. I also get the nasty Jitter signals when I do my test directly into the cable modem. Two ARRIS modems and now a Zoom 5341J with the exact same results on Jitter. The otherday for some reason my VoIP tests were all in the green. I called a few people on my phone and they said it sounded great. Later that evening back into the yellow and red and my phone was cutting out on my end again.
I tried setting up the QoS on my linksys wireless router and it doesn't seem to make any difference? |
|
| pepi3 |
pepi3
Member
2013-Mar-4 9:08 am
I was also told that maybe my upload has too much noise on it?
Upstream Bonded ChannelsChannel Lock Status US Channel Type Channel ID Symbol Rate Frequency Power
1 Locked ATDMA 7 5120 Ksym/sec 31600000 Hz 35.5 dBmV
2 Locked TDMA 1 2560 Ksym/sec 37500000 Hz 37.0 dBmV
3 Not Locked Unknown 0 0 Ksym/sec 0 Hz 0.0 dBmV
4 Not Locked Unknown 0 0 Ksym/sec 0 Hz 0.0 dBmV |
|
gar187erI DID this for a living
join:2006-06-24
Seattle, WA |
this is upstream signal to noise, and can only be read from the CMTS,not from the house. |
|
| |
to pepi3
Completely depends on your router, if you can get ahold of an IOS device than I may be able to send you some commands. I know an ASUS rt-n66u can do this. There maybe something in DD-wrt or tomato as well. |
|
