dslreports logo
site
 
    All Forums Hot Topics Gallery
spc

spacer




how-to block ads


Search Topic:
uniqs
3858
share rss forum feed


txrr

@rr.com

does twc rig the bandwidth on speedtest.net ?




why do i constantly keep getting 20+mbit results on speedtest when i only have 10mbit

yes i know there's a thing called turboboost, but the turboboost on speedtest.net seems to go on for a good 10-15 seconds and actually SPEEDS UP in the middle of the test (i hit around 17mbit most times but always end up with 20+mbit)

when i'm downloading a file from any regular site, my turbo boost never jumps to 30mbit like it does on speedtest, and the speeds only last for 3-5 seconds, and never do they actually SPEED UP during the turbo boost - it's a fast jump to my regular 10mbit speeds unlike on speedtest.net where the speeds stay around 20mbit for like 15 seconds


banditws6
Shrinking Time and Distance
Premium
join:2001-08-18
Frisco, TX
Reviews:
·Time Warner Cable

I've come to believe that speedtest.net is next to worthless at least on cable. I was seeing the same results as you all day long on my 10/1 connection, when in actuality I was only getting about 3-4 megabits of usable downstream bandwidth -- as verified by actual file downloads as well as tests from testmy.net and »myspeed.visualware.com. (Which are much better speed test services, IMO.) Incidentally I solved that speed problem by switching to a DOCSIS 3 modem so I could bond channels.

I don't think TWC is "rigging" speedtest.net but the methods used by that service to test your connection seem to only be good at telling you a theoretical maximum, Powerboost included.
--
"The counsel of fools is all the more dangerous the more of them there are." -Ólafr Höskuldsson



Smith6612
Premium,MVM
join:2008-02-01
North Tonawanda, NY
kudos:24
Reviews:
·Verizon Online DSL
·Frontier Communi..
reply to txrr

Those Flash speed tests seem to work on an average. If your speed is bursting for a long enough time but is then slowing down it will inaccurately measure the speed. Defeating Powerboost also requires saturating the connection for a long enough time. A good speed test will either download a large enough file to account for the PowerBoost, or it will run the connection until PowerBoost wears off, then start logging speed.



sivran
Opera ex-pat
Premium
join:2003-09-15
Irving, TX
kudos:1

1 recommendation

reply to txrr

Flash speed tests suck. Try testmy.net.


Mele20
Premium
join:2001-06-05
Hilo, HI
kudos:5
reply to txrr

Use Visualware's tests and use Java based tests...not Flash. Visualware has the best tests of all. TWC uses Visualware's tests on all its gateway servers and always has. TWC here in Hawaii leased MySpeed test for us RR users about 4-5 years ago (their second choice was to lease Justin's Java speed test). I was a main force in persuading OTWC to get MySpeed test. They had it for two years and then didn't renew their contract with Visualware but went with Speedtest.net. Why? (1) because Speedtest.net is FREE and (2) because the most important measurement which is the Quality of Service one OTWC had lots of problems giving us a strong, non variable speed. That severe shortcoming showed up glaringly on MySpeed tests. Speedtest.net cannot measure QOS. I was told that too many users were calling in about poor QOS so OTWC decided to move to Speedtest.net with its inflated values especially if you have PowerBoost.

NO speed tests out there can do an accurate test if you have PowerBoost except for CERTAIN (not all) Visualware MySpeed tests and SamKnows tests which are able to factor out Speedboost into a separate rating so you can see your speed with PowerBoost and your REAL SUSTAINED speed without PowerBoost.

Do this Quality test which is a Visualware test. There is a server in Los Angeles, Texas and Georgia. The link is to the Los Angeles one but pick the one closer to you. Read carefully and do the route test and capacity one also. Capacity test is what speedtest.net is doing and all other speed tests except MySpeed and old fashioned download of a very large file.

"This test is special in that it is an application speed test and displays real world speeds and connection quality as would be observed by most applications that you currently operate such as streaming or voice over IP and other single socket applications. Do NOT treat the results from this test as you would results from other broadband speed tests".

This test can filter out PowerBoost...at least it has done so for me when I had 10/1 and Powerboost. I have 15/1 now and there is no Powerboost on it. Your QOS (Quality of Service) should always be in the high 90's.

»www.ispgeeks.com/wild/modules.ph···lityTest

You can also do an old fashioned big file download from a known fast server as PowerBoost will exhaust itself and then the latter part of the test will show your true speed (of course you have sit and watch the downloading in your browser to see this). This is the one I have used for many years:

»ftp://ftp1.optonline.net/test64
--
When governments fear people, there is liberty. When the people fear the government, there is tyranny. Thomas Jefferson



Packeteers
Premium
join:2005-06-18
Forest Hills, NY
kudos:1
Reviews:
·Time Warner Cable
reply to sivran

IMHO testmy.net is the only true test out there.

make sure a test sample is double your speed rates

also make sure to pick the right test server as that
setting is hidden at the top center of the test page,
and often auto detects incorrectly. my NYC tests
will run out of Dallas TX instead of Washington DC,
thus giving me unfair results - so pay attention



Selenia
I love Debian
Premium
join:2006-09-22
Fort Smith, AR
kudos:2
reply to txrr

Nah. Speedtests usually choose shorter non-congested routes. Routes to many sites may be longer or congested. Some smaller sites struggle under peak load and will not max out your line.