True Ultra Hi-Speed Internet

[bkkmick]

Follow up: I have just visited a True shop. The surly 'girl' that I spoke with said that there is a 'general network problem' in my area (On Nut) and that it will take 24 hours to fix.

He/She wasn't interested in the fact that my neighbor doesn't have any problems with his True connection.

[infernalman7]

I can say that I'm "currently" satisfied with my True cable service.

Even though the first few months of it being available it was so erratic. Could be extremely good and terrible by the hours. Now I think the DOCSIS networks around my house is fairly stable now with no new people moving in the neighborhood or converting in/out of the network itself. And this I think is part of the reason why the connection has been so unstable when they first introduced it.

I'm subscribed to 50Mbps and I usually get around 25-30Mbps during the day and 45Mbps at night. Not so bad. Pings are still not that satisfactory (I still get killed constantly on US-based online games) but streaming isn't an issue.

[Pib]

I can say that I'm "currently" satisfied with my True cable service.

Even though the first few months of it being available it was so erratic. Could be extremely good and terrible by the hours. Now I think the DOCSIS networks around my house is fairly stable now with no new people moving in the neighborhood or converting in/out of the network itself. And this I think is part of the reason why the connection has been so unstable when they first introduced it.

I'm subscribed to 50Mbps and I usually get around 25-30Mbps during the day and 45Mbps at night. Not so bad. Pings are still not that satisfactory (I still get killed constantly on US-based online games) but streaming isn't an issue.

If your are using a "wireless" router you may be losing quite a few Mbps there. If using a wireless router try the speed test with computer hooked directly to a router port.

[Alfahane]

I just got True Ultra 100/10 mbit at the outskirt of Patong.

Torrent running at about 6MB/s (yes, megabytes. not megabits). For me, being in Thailand, this is incredible.

Speedtest (Sweden!) _while_ running the torrent at about 6MB/s

Again, this is _while_ running torrent.

While writing this uTorrent peaks at 8,5MB/s.

I just hope that this is not the first day glory that some of you have been describing.

For those of you that experience low torrent speed. Try doubling maximum connections per torrent and total. I set this to 800. It did help dramatically.

[TallGuyJohninBKK]

Dunno about your other results, but the Speedtest.net result you've posted below is clearly false... and a function of the way True configures its system.

It's physically impossible to get anything close to a 21 ms ping time to a server location 5400 miles away.

That's measuring the contact time and speed from your location to True's server in Thailand... not the end point.

I just got True Ultra 100/10 mbit at the outskirt of Patong.

Torrent running at about 6MB/s (yes, megabytes. not megabits). For me, being in Thailand, this is incredible.

Speedtest (Sweden!) _while_ running the torrent at about 6MB/s

Again, this is _while_ running torrent.

While writing this uTorrent peaks at 8,5MB/s.

I just hope that this is not the first day glory that some of you have been describing.

For those of you that experience low torrent speed. Try doubling maximum connections per torrent and total. I set this to 800. It did help dramatically.

[Alfahane]

Dunno about your other results, but the Speedtest.net result you've posted below is clearly false... and a function of the way True configures its system.

It's physically impossible to get anything close to a 21 ms ping time to a server location 5400 miles away.

That's measuring the contact time and speed from your location to True's server in Thailand... not the end point.

Yes, the ping is incorrect.

5400 miles = 8690km --> 8690 / 300 000 = 29 ms

A minimum of 29ms if in a straight line and optic only. (Electric signal is about 2/3 of the speed of light)

I'd guess it's more like 200-300 ms.

But the torrent speed is correct. I've been downloading video etc since my last post and it's really stable. Maybe because True Docsis is new here in Patong and everybody seem to be using ADSL. Also, me and my spouse have been very closely involved in planning and overseeing the installation trying to make sure the technicians did good job.

Parallell to torrent I've been using remote desktop (Sweden), streaming video (Sweden) and doing other internet based things. Remote desktop does lag when I scroll in a window, but nothing bad.

[Alfahane]

Here is the test again with torrent turned OFF.

Phuket <-> Sweden

Phuket <-> San Fransisco

Phuket <-> Sidney

Phuket <-> London

Phuket <-> Surat Thani

[TallGuyJohninBKK]

Once again, those results above clearly tell you you're getting false readings...

You're getting twice as fast ping times to the various international locations, low 20s ms, as you are to a domestic Thailand location, 48 ms. to Surathani... By the way... you're generally right about the times, as the typical ping from Thailand to the U.S. is 250 to 300 ms for a distance of 7000 or so miles, roughly speaking.

Usually, when the ping times are wrong like that, the speed results are wrong too, in that they're actually measuring the speed between your location and True's domestic server -- not the international location shown on the speed test result.

This is a common, long-standing problem with True and SpeedTest.net. You'll find scores and scores of TV posts on the same subject.

You may have other ways of measuring your speed results apart from Speedtest.net... But putting True and Speedtest.net together is simply a recipe for fake results -- unless for some reason you're also using True's proxy server settings, which then remove that issue and allow Speedtest.net to show real-world results.

BTW, congratulations on getting the 100/10 True cable plan, and for posting about it here. There are a few of us with the 10 and 20 Mbps plans... but not many or any posting with the 50 or 100 Mbps plans... So if we can nudge you into generating meaningful, accurate results, then it would be an interesting education for all of us.

One of the real unknowns for the user community here is how much additional speed/performance do True users get for international connections via their cable internet service on the lower speed plans (10-20 Mbps) vs. their higher speed plans (50 or 100 Mbps). We really haven't had enough of the higher speed subscribers posting here to gain any real insight.

We have done some head to head 10 Mbps vs. 20 Mbps plans tests before -- connecting to the same sites at the same times using the two different plans -- and found there was little sustained advantage for the higher speed/more expensive plan over the slower one for those international connections. Faster inside Thailand and maybe regionally, yes. But no evidence of any meaningful advantage internationally.

Now, if we were comparing 10 and 20 Mbps vs. 100 Mbps True Online services on actual results in identical tests, I'd love to know those results.

[Alfahane]

Yes, the ping was of course wrong for the reasons you mention. I think though that the DL/UL speeds are somewhat representative.

I still have 3BB 12mbit and speeds between me and the servers in Sweden I mostly work with is multiplied. This applies to everything I've tested.

I understand that comparing 10/20/50/100 Docsis would be more interesting. =)

Please do nudge me in the right direction to make proper tests. I'd be happy to share.

[TallGuyJohninBKK]

Please do nudge me in the right direction to make proper tests. I'd be happy to share.

A good start is to read this page below earlier in this same thread, that deals with speed test results and True's proxy settings:

Pay particular attention to TV member Pib's post #355 on that page, and its instructions, and then the other posts that follow it...

And you'll be on your way to getting real, accurate results from Speedtest.net

As I mentioned before, True and Speedtest.net are an often discussed subject around here...

[Pib]

Yes, setup your browser to use one of the True proxy servers and this should prevent Speedtest.net from giving you bogus download speed and ping results. . Something about the proxy setup prevents "in-Thailand, cached server" from skewing the speedtest results....prevents an L.A. speedtest server from giving results like it's just across the soi. Now whether you use the proxy regularly is up to you as it has its good and bad speed days (mostly good days), but it will eliminate the bogus speedtest results.

Also, if using a wiress router/modem with your 100Mb cable plan you are probably going to have to not use wireless to achieve the full 100Mb as most wireless routers, even N routers, have a hard time getting anywhere close to a 100Mb actual throughput in wireless mode.

[Supernova]

Proxy or no proxy, speedtest.net results are often exaggerated. Use the following services to measure speed instead.

http://www.dslreports.com/stest

http://speedtest.microsoftonline.com/

[TallGuyJohninBKK]

Yes.. SN, for whatever reason, neither of those speed test sites fall victim to True's caching fakeout...

I've always liked and used DSL Reports, in addition to the MS site, which provides much more detailed and useful info than just speed results...

But SpeedTest.net is such a common and oft used site, it's sometimes seemed easier with other folks to try to get them to better use the site they already want to use.

I've never used the True proxy or tried Pib's approach with Speedtest.net. But he's reported that doing so resolved the accuracy problem with the site.

[Supernova]

I've never used the True proxy or tried Pib's approach with Speedtest.net. But he's reported that doing so resolved the accuracy problem with the site.

See this post and the one above it.

[TallGuyJohninBKK]

Maybe I'm not following your point, SN... The posts you linked to seem to reconfirm what we were saying above about the use of True's proxy with the Speedtest.net website???

[shariq607]

you guys wont believe this but i just found out how deceiving true is

My friend who has 9 mb true adsl connection has been getting very slow downloads and slow browsing even from thai websites,

but of-course true would never believe him since true technicians have been coming his house and doing various tests from speedtest.net bangkok servers and he is getting full speed on the test sites

is there any other website which has thai servers and will not be fooled by true's deceiving non sense ?

[TallGuyJohninBKK]

Here are some other Thai sites, but I don't know if they have any connection with True and its accuracy issues or not.

http://speedtest1.totbb.net/

http://www.speedtest.or.th/

[Pib]

you guys wont believe this but i just found out how deceiving true is

My friend who has 9 mb true adsl connection has been getting very slow downloads and slow browsing even from thai websites,

but of-course true would never believe him since true technicians have been coming his house and doing various tests from speedtest.net bangkok servers and he is getting full speed on the test sites

is there any other website which has thai servers and will not be fooled by true's deceiving non sense ?

If the Speedtest.net in Bangkok server is giving a 9Mb result for a 9Mb plan then your friend is getting 9Mb speed, regardless of whether a cached or non-cached fill is being downloaded from the Bangkok server. Thai ISPs only advertise and supposedly guarantee "in-Thailand" speed to a nearby server. Also, I expect the techs probably used their own laptops connected directly/hardwired to the modem/router to eliminate any throughput problems caused by wireless issues, setup issues with the customer's computer, etc. Additionally, wireless issues (problems) can cause a lot of download speed problems, although you normally won't see many problems on a 10Mb or below plan...but once going above 10Mb "actual throughput" of wireless routers can be reduced due to distance, walls, floors, poor design, etc. While wireless routers may advertise say 100Mb speed this is probably the basic/max connection/sync speed and not the actual/maximum throughput speed.

[Alfahane]

Update on True Ultra 100/10, Patong:

Speed Test #99922900 by dslreports.com

Run: 2011-12-10 02:07:03 EST

Download: 1945 (Kbps)

Upload: 1038 (Kbps)

In kilobytes per second: 237.5 down 126.7 up

Tested by server: 54 java

User: 2 @ dslreports.com

User's DNS: asianet.co.th

No downtime since installation 2011-12-01.

Edited December 10, 2011 by Alfahane

[Pib]

Update on True Ultra 100/10, Patong:

Speed Test #99922900 by dslreports.com

Run: 2011-12-10 02:07:03 EST

Download: 1945 (Kbps)

Upload: 1038 (Kbps)

In kilobytes per second: 237.5 down 126.7 up

Tested by server: 54 java

User: 2 @ dslreports.com

User's DNS: asianet.co.th

No downtime since installation 2011-12-01.

Only 1.945Mb download speed on a 100Mb download plan and 1.038Mbupoload on the 10Mb upload side. I would be calling True technical support.

[TallGuyJohninBKK]

Additionally, wireless issues (problems) can cause a lot of download speed problems, although you normally won't see many problems on a 10Mb or below plan...but once going above 10Mb "actual throughput" of wireless routers can be reduced due to distance, walls, floors, poor design, etc. While wireless routers may advertise say 100Mb speed this is probably the basic/max connection/sync speed and not the actual/maximum throughput speed.

Re Pib's comments above, lately I've been using a piece of freeware software called InSSIDer, a wifi network scanning package that allows you to monitor your own and other wifi broadcast details around you. You install the software on whatever computer you're using with a wifi receiver, and then the software monitors the signal from your router to your wifi connected PC in real time.

The details about the software, which has both 32 and 64 bit versions, and user reviews from those who have used it are located here at CNET:

http://download.cnet.com/inSSIDer/3000-2085_4-10848357.html?tag=contentMain;contentBody;1d

Where I've found it particularly useful is:

--being able to see which 2.4 Ghz frequency bands (or 5 Ghz if you're using them) other competing wifi routers around me are using, and adjust my own frequencies in my routers to avoid local conflicts as much as possible.

--being able to compare the signal strength of different G and N routers that I have and use, and also see how locational changes in my home affect wifi signal strength in real time.

[johnnyj]

Quick note. Regardless of what they say, -any- Internet connection can run wireless, however you may need to set up the wifi aspect (or have someone who knows how) yourself.

There's plenty of info on google but the short answer is

Cable modem -> wireless switch/hub which will provide wifi and wired connections

[shariq607]

Wow good to see someone Actually posting results with true 100/10 mb Plan

getting only 2 mb from a 100mb plan seems a rip-off considering the fact that my true 6 mb adsl line does better or same as that

really Glad i got a fibre optic line then a true cable plan but i had high hopes for the 100 mb true cable plan

Ohh and pib do you really think true will care about international speed test results ? if so i have been trying since past few days and i get same answer 'we cant guarantee international speed"

just for a comparison purposes of what i am getting from my fibre optic plan

[Pib]

Update on True Ultra 100/10, Patong:

Speed Test #99922900 by dslreports.com

Run: 2011-12-10 02:07:03 EST

Download: 1945 (Kbps)

Upload: 1038 (Kbps)

In kilobytes per second: 237.5 down 126.7 up

Tested by server: 54 java

User: 2 @ dslreports.com

User's DNS: asianet.co.th

No downtime since installation 2011-12-01.

Only 1.945Mb download speed on a 100Mb download plan and 1.038Mbupoload on the 10Mb upload side. I would be calling True technical support.

On second thought and now fully realizing the test is being run on DSLReports.com (was was thinking Speedtest.net for some reason) I wouldn't call tech support as your speed using the dslreport.com speedtester is pretty much what a person would get whether they are on the 7, 10, 20, 30, 50, or 100Mb True cable plan. I just ran the dslreports.com test to L.A. on my True cable 20Mb/2Mb plan and got pretty much the same values as you did and I knew I would get those results before they even finished running as I've run them before. It's just an international bandwidth limitation/chokepoint and DSLReports thing....probably other factors involved but I'm not smart enough to figure them out.

As for no downtime, I've only had one 8 hour down period since I had the 20Mb plan installed in mid June...both the cable internet and TV were non-operational for that period due to a wide area outage....the second the cable TV came back on the internet came back on also. And during the recent flooding when my moobaan had a 1.5 meters of water in the streets, the cable internet and TV continued to hum along....now my neighbors on TOT ADSL weren't that luckly...they had very slow to no internet (or phone line) due to the ground level junction boxes getting flooded. Heck one of my TOT phone lines is basically still not working (my streets are dry now) due to high noise while my other TOT line is working fine....I expect the faulty line hooks up within the junction box near the bottom (the part that got flooded) while the other line hooks up near the top (the unflooded part). True cable TV/internet does not use ground level junction boxes (at least in my moobaan)...everything is high on the pole.

Edited December 10, 2011 by Pib

[TallGuyJohninBKK]

Update on True Ultra 100/10, Patong:

Speed Test #99922900 by dslreports.com

Run: 2011-12-10 02:07:03 EST

Download: 1945 (Kbps)

Upload: 1038 (Kbps)

In kilobytes per second: 237.5 down 126.7 up

Tested by server: 54 java

User: 2 @ dslreports.com

User's DNS: asianet.co.th

No downtime since installation 2011-12-01.

Only 1.945Mb download speed on a 100Mb download plan and 1.038Mbupoload on the 10Mb upload side. I would be calling True technical support.

What's to call about, Pib???

Seriously, I agree, I wouldn't be happy if I was PAYING the much higher monthly True Online bill for their 100 Mbps plan and only getting international speeds that are roughly the same as typical 10 or 20 Mbps plan customers who are paying 699 to 1100 baht per month.

But, as has been noted many times, True (and the other ISPs as well) never make any guarantees with their plans regarding international download/upload speeds... at least not on any of their retail consumer plans.

I know they have the technical capability to do better. I remember reading a True promo about their Knowledge Park set-up in the Digital Gateway center opposite Siam Paragon where they were claiming 30 Mbps or so international speeds, and 100 Mbps or higher local speeds, at their connection there. Of course, that location and setup, being True's own, is a lot different than a retail customer's home somewhere out on their network.

When we did our head-to-head True cable internet speed tests before pitting your 20 Mbps plan vs. my 10 Mbps plan, there really wasn't much of a speed advantage for the 20 Mbps plan for international connections, particularly to the U.S. and Europe, as I recall. Now, we're seeing more of that same kind of evidence above with Alfahane's post re his 100 Mbps plan.

Of course, it really depends on what kind of internet use a person is doing. For some folks, perhaps a speedy 100 Mbps for domestic traffic inside Thailand might have some advantage and purpose. But for most farangs, I suspect, we're more interested and focused on a variety of international connections and the speeds for them.

[shariq607]

Update on True Ultra 100/10, Patong:

Speed Test #99922900 by dslreports.com

Run: 2011-12-10 02:07:03 EST

Download: 1945 (Kbps)

Upload: 1038 (Kbps)

In kilobytes per second: 237.5 down 126.7 up

Tested by server: 54 java

User: 2 @ dslreports.com

User's DNS: asianet.co.th

No downtime since installation 2011-12-01.

Only 1.945Mb download speed on a 100Mb download plan and 1.038Mbupoload on the 10Mb upload side. I would be calling True technical support.

On second thought and now fully realizing the test is being run on DSLReports.com (was was thinking Speedtest.net for some reason) I wouldn't call tech support as your speed using the dslreport.com speedtester is pretty much what a person would get whether they are on the 7, 10, 20, 30, 50, or 100Mb True cable plan. I just ran the dslreports.com test to L.A. on my True cable 20Mb/2Mb plan and got pretty much the same values as you did and I knew I would get those results before they even finished running as I've run them before. It's just an international bandwidth limitation/chokepoint and DSLReports thing....probably other factors involved but I'm not smart enough to figure them out.

As for no downtime, I've only had one 8 hour down period since I had the 20Mb plan installed in mid June...both the cable internet and TV were non-operational for that period due to a wide area outage....the second the cable TV came back on the internet came back on also. And during the recent flooding when my moobaan had a 1.5 meters of water in the streets, the cable internet and TV continued to hum along....now my neighbors on TOT ADSL weren't that luckly...they had very slow to no internet (or phone line) due to the ground level junction boxes getting flooded. Heck one of my TOT phone lines is basically still not working (my streets are dry now) due to high noise while my other TOT line is working fine....I expect the faulty line hooks up within the junction box near the bottom (the part that got flooded) while the other line hooks up near the top (the unflooded part). True cable TV/internet does not use ground level junction boxes (at least in my moobaan)...everything is high on the pole.

i agree,

There is definitely something up with Dslreports

i get 23-25 mb in speedtest.net LA server

and in Dslreports 11-13 mb

Edited December 10, 2011 by shariq607

[shariq607]

[TallGuyJohninBKK]

BTW, for DSL Reports speed test, they recommend using their Java test, instead of their Flash test, for higher speed connections like 100 Mbps plans... Here's specifically what they say:

</h3>

<h3>Flash Speed Test

What is the maximum speed this test will correctly report?

<h3 style="border-top: 0em none; display: inline;"></h3>

The maximum speed the flash test will correctly report varies according to your PC performance. Flash is not as efficient as Java for very high speed connections, so if your connection speed is going to be above 20mbit, then you should also try the Java-based speed tester to verify that you are fully exercising your connection during the test. We have had reports from users on 100mbit connections and moderately powerful PCs that they cannot get more than 25 mbit reported by flash (any flash-based speed test, not just ours), but they can get 80 or 90 mbit reported by the Java test.

[Pib]

On the DSLReports and Microsoft Online Services speedtesters I've always got low results, with the lowest being on MS Online Services. Regardless of the international bandwidth limitation/chokepoint applicable to all Thai ISPs, it seems different speedtesters to the same the same city can give widely varying results when run within minutes of each other. Is this issue caused by DNS routing, speedtester server capability/capacity, speedtester programming, local server caching, etc...etc....etc...I just don't know. Take all speedtester results with a big grain of salt in trying to figure out which one is giving accurate results.

[Pib]

BTW, for DSL Reports speed test, they recommend using their Java test, instead of their Flash test, for higher speed connections like 100 Mbps plans... Here's specifically what they say:

</h3>

<h3>Flash Speed Test

What is the maximum speed this test will correctly report?

<h3 style="border-top: 0em none; display: inline;"></h3>

The maximum speed the flash test will correctly report varies according to your PC performance. Flash is not as efficient as Java for very high speed connections, so if your connection speed is going to be above 20mbit, then you should also try the Java-based speed tester to verify that you are fully exercising your connection during the test. We have had reports from users on 100mbit connections and moderately powerful PCs that they cannot get more than 25 mbit reported by flash (any flash-based speed test, not just ours), but they can get 80 or 90 mbit reported by the Java test.

I'm sure this is true but for me on my 20Mb plan running the Flash or Java based speedtest give me basically the same results...within a few percent of being the same. And I wouldn't be surprised if a person on a 30Mb to 100Mb plan got basically the same results (within a few percent) "due to True's international gateway bandwidth limitation/chokepoint" which basically slaps down a person's 6Mb (or lower) to 100Mb in-Thailand connection (cable or ADSL) to a much lower international connection. It seems all Thai ISP international gateways due a certain amount of "choking" on their international bandwidth....probably a money thing as bandwidth costs money.

Edited December 10, 2011 by Pib