Geekzone: technology news, blogs, forums
Guest
Welcome Guest.
You haven't logged in yet. If you don't have an account you can register now.

View this topic in a long page with up to 500 replies per page Create new topic
1 | 2 | 3 | 4
3932 posts

Uber Geek
+1 received by user: 175

Trusted
Subscriber

  Reply # 738627 30-Dec-2012 17:39 Send private message

sbiddle: If you're in a cabinetised area around 80% of MPF's have a loop distance of 1km or less, with ~800m being the maximum reach of VDSL2+ if you want acceptable performance. Greater than this and you'll get better downstream with ADSL2+, but will still get better upstream than ADSL2+.

The TrueNet stats corroborate that; it's been a little while since I've looked at them but I think I saw places at ~1.2 km that still had several Mb/s upstream. I'd love to give it a try here but my ISP doesn't offer it, and that's the other problem; limited ISP support for VDSL.

420 posts

Ultimate Geek
+1 received by user: 84


  Reply # 738648 30-Dec-2012 18:03 Send private message

It'd be nice if Chorus made VDSL more affordable in areas where there is no UFB service yet.

19764 posts

Uber Geek
+1 received by user: 1515

Moderator
Trusted
Biddle Corp
Subscriber

  Reply # 738662 30-Dec-2012 18:47 Send private message

Behodar:
sbiddle: If you're in a cabinetised area around 80% of MPF's have a loop distance of 1km or less, with ~800m being the maximum reach of VDSL2+ if you want acceptable performance. Greater than this and you'll get better downstream with ADSL2+, but will still get better upstream than ADSL2+.

The TrueNet stats corroborate that; it's been a little while since I've looked at them but I think I saw places at ~1.2 km that still had several Mb/s upstream. I'd love to give it a try here but my ISP doesn't offer it, and that's the other problem; limited ISP support for VDSL.


Last I heard there were 20 ISP's offering VDSL2.



195 posts

Master Geek
+1 received by user: 5


  Reply # 738677 30-Dec-2012 19:13 Send private message

 
The TrueNet stats corroborate that; it's been a little while since I've looked at them but I think I saw places at ~1.2 km that still had several Mb/s upstream. I'd love to give it a try here but my ISP doesn't offer it, and that's the other problem; limited ISP support for VDSL.


I have the fastest download speed on the Trunet stats.  What are your current speeds or do you know your downstream attenuation or approximate distance from the cabinet?  Snap offer VDSL from $95 a month so getting is no problem.

813 posts

Ultimate Geek
+1 received by user: 33

Trusted
Subscriber

  Reply # 738703 30-Dec-2012 20:50 Send private message

sbiddle:
Behodar:
sbiddle: If you're in a cabinetised area around 80% of MPF's have a loop distance of 1km or less, with ~800m being the maximum reach of VDSL2+ if you want acceptable performance. Greater than this and you'll get better downstream with ADSL2+, but will still get better upstream than ADSL2+.

The TrueNet stats corroborate that; it's been a little while since I've looked at them but I think I saw places at ~1.2 km that still had several Mb/s upstream. I'd love to give it a try here but my ISP doesn't offer it, and that's the other problem; limited ISP support for VDSL.


Last I heard there were 20 ISP's offering VDSL2.


I guess the challenge is that if it isn't Telecom, Vodafone or TelstraClear that excludes a large part of NZ and forces people to have to change ISP - which some folk just won't want to do, or will be tied into a contract preventing them from doing so easily...




________
AK

269 posts

Ultimate Geek
+1 received by user: 9

Trusted
Subscriber

  Reply # 738711 30-Dec-2012 21:10 Send private message

I posted some TrueNet stats on VDSL vs ADSL at the beginning of the month, while the speeds above are not being reached with our volunteers, the speeds reached are dramatic when compared to ADSL (note the scales below had to be 2:1). Our volunteers do not get the elusive 40+Mbps, but that is quite likely due to in-house wiring issues - filters etc. My own experience is visible in the chart as probe 555, 900m from the exchange 27/10Mbps, but I have VoIP and Cat6 throughout including to the outside line - so the best wiring possible. It is very nice :-)



A comparison with Telecom ADSL probes is also included, I used Telecom because their distribution most closely matches the overall distribution of probes, providing an approximation of the proportion of connections able to get great speeds on VDSL - I expect volunteers are within reach of VDSL at say 1.6km, rather than 1km our volunteers seem restricted to.



VDSL upload and download speeds sorted by distance to the DSLAM

I also published a performance x distance comparison in June that is helpful as it includes all probes:




www.truenet.co.nz

We are seeking more volunteers here :-)



195 posts

Master Geek
+1 received by user: 5


  Reply # 738735 30-Dec-2012 23:12 Send private message

Thanks John for the info its good to have stats but to be clear my line can hit 40mb/s.




195 posts

Master Geek
+1 received by user: 5


  Reply # 738737 30-Dec-2012 23:25 Send private message

Also one more image of a speedtest. This line is aprox 900m from the exchange in Te Awamutu with a line attenuation of 12 and on DLM 1. So its not all about distance from the exchange there can be a lot of interference that can cause lower speeds.


3015 posts

Uber Geek
+1 received by user: 196

Trusted
Subscriber

  Reply # 738747 30-Dec-2012 23:55 Send private message

THe pricing my ISP has quoted for a full VDSL install are crazy, like $400. I've been having success just getting it with connection only install and then logging it as a fault. Usually its not jumpered to the correct property etc. so pretty much is a fault.

In regards to the master splitter, in all cases no POTS have been delivered and only one doesn't have a direct patch to the ETP. Will be checking out a place tomorrow which doesn't have a master filter and multiple jackpoints.





1910 posts

Uber Geek
+1 received by user: 453

Trusted
Spark NZ

  Reply # 738748 31-Dec-2012 00:04 Send private message

Many speedtests make stuff up to give the highest estimation of throughput that is credible. Read the methods they use - they open up many threads, they discard some results (high and low) and then extrapolate to what the best possible speed could be.

Cheers - N



195 posts

Master Geek
+1 received by user: 5


  Reply # 738749 31-Dec-2012 00:16 Send private message

Talkiet: Many speedtests make stuff up to give the highest estimation of throughput that is credible. Read the methods they use - they open up many threads, they discard some results (high and low) and then extrapolate to what the best possible speed could be.

Cheers - N


Still seen my connection hit 5MB/s in real world downloads and we are talking possible line maximums in best case scenarios not what every server in the world is going to give you.



195 posts

Master Geek
+1 received by user: 5


  Reply # 738750 31-Dec-2012 00:21 Send private message

Zeon: THe pricing my ISP has quoted for a full VDSL install are crazy, like $400. I've been having success just getting it with connection only install and then logging it as a fault. Usually its not jumpered to the correct property etc. so pretty much is a fault.

In regards to the master splitter, in all cases no POTS have been delivered and only one doesn't have a direct patch to the ETP. Will be checking out a place tomorrow which doesn't have a master filter and multiple jackpoints.


Interesting have you just suggested adding a direct connection to the ETP when the tech comes out?  This is actually what happened at my house but was actually a fault in the house wiring when I had ADSL2.  New Cat5 direct from the ETP to a new Jackpoint and didn't pay a cent for VDSL install.  If there is only one Jackpoint connected or find the master one then disconnect the others and there should be no need for a splitter or wiring charge.

269 posts

Ultimate Geek
+1 received by user: 9

Trusted
Subscriber

  Reply # 738772 31-Dec-2012 08:30 Send private message

Speedtests are often very different from the testing TrueNet completes as a standard test.  Although the speedtest.net test is a download from a site of a random picture file, much like our own standard test.

TrueNet measures the speed of download by recording the best quartile bits/time of a file being downloaded from a very common website server through the ISP to the probe.  

We became curious about the differences recently for very high speed connections and tested a comparison using the same methodology.  This is presented below for a group of probes with common characteristics - 100Mb/s (unfortunately we did not include fellaintga in these tests);



AK Speedtest is http://speedtest.citylink.co.nz/speedtest/random2000x2000.jpg 
CH Speedtest is http://speed.snap.net.nz/speedtest/random2000x2000.jpg
TP_1MB_file_Dallas is our international test
TP_National is our standard download test we report in charts above

The results show the difference between great connections to a server and not-so-good connections. Download speed performance is all about speed from a wide variety of sources so it is difficult to define the perfect test, but we believe our use of a popular server represents the best option for reporting comparisons.

The results above show that this ISP has a great connection to the Christchurch speedtest server, but not so good to the Auckland one.  The TrueNet standard test at typically ~30Mb/s is very slow for these connections.




www.truenet.co.nz

We are seeking more volunteers here :-)

1910 posts

Uber Geek
+1 received by user: 453

Trusted
Spark NZ

  Reply # 738791 31-Dec-2012 09:34 Send private message

JohnButt: Speedtests are often very different from the testing TrueNet completes as a standard test.  Although the speedtest.net test is a download from a site of a random picture file, much like our own standard test.

TrueNet measures the speed of download by recording the best quartile bits/time of a file being downloaded from a very common website server through the ISP to the probe.  
[snip]


Thanks for the insight John, very useful to help with interpreting the Truenet data... 

However, the speedtest.net test is hardly just a download of a random picture file... From their own help pages...

Download Speed
  1. Your computer downloads small binary files from the web server to the client, and we measure that download to estimate the connection speed.
  2. Based off this result, we choose how much data to download for the real test. Our goal is to pick the right amount of data that you can download in 10 seconds, ensuring we get enough for an accurate result but not take too long.
  3. We prevent caches from throwing off results by appending random strings to each download.
  4. Once we start downloading, we use up to four HTTP threads to saturate your connection and get an accurate measurement.
  5. Throughput samples are received at up to 30 times per second.
  6. These samples are then aggregated into 20 slices (each being 5% of the samples).
  7. The fastest 10% and slowest 30% of the slices are then discarded. We'll explain that more below.
  8. The remaining slices are averaged together to determine the final result.
Since we are measuring data transported over HTTP (via Flash), there are the following factors that can affect speed: potential protocol overhead, buffering due to the many layers between our application and the raw data transfer, and throughput bursting due primarily to CPU usage. These factors lead us to drop the top 10% and bottom 10% of our slices as outliers.

Additionally, we keep the default test length short for user experience. Because the test is shorter, the ramp-up period can take a significant part of the beginning of the test, leading us to drop another 20% of the bottom result slices.
(My bolding)

From this basic description you can see that they discard the lowest 30% of speed metrics (and the highest 10%) before calculating the speed to present to you. It's FAR form a simple x/y calculation.

They do something similar for uploads as well...

While I don't disagree with the concept, (so long as people keep in mind it is an ARTIFICIAL ESTIMATE of PEAK SPEED only, I think we've all see the literally impossible upload speed estimates that speedtest.net can give on ADSL lines. It frequently shows well over 1mbi (up to 2mbit even) for basic ADSL2+ upload speeds. These numbers are a result of their algorithms breaking somehow.

As for the comment on Truenet speed calculation, are you saying that the results you present represent ONLY the best 25% of performance across the file(s) downloaded during a test? Again, no issue if that's the case, but means it's not possible to even try and compare your results to any other testing someone might be doing.

Cheers - N

edit: I've just noticed the speedtest page has a mistake... It says top 10% and bottom 30% in one place, and then says top 10% and bottom 10% a few line later... Not sure which is right.

(ref: https://support.speedtest.net/entries/20862782-how-does-the-test-itself-work-how-is-the-result-calculated )

3932 posts

Uber Geek
+1 received by user: 175

Trusted
Subscriber

  Reply # 738798 31-Dec-2012 10:02 Send private message

fellaintga:
 
The TrueNet stats corroborate that; it's been a little while since I've looked at them but I think I saw places at ~1.2 km that still had several Mb/s upstream. I'd love to give it a try here but my ISP doesn't offer it, and that's the other problem; limited ISP support for VDSL.


I have the fastest download speed on the Trunet stats.  What are your current speeds or do you know your downstream attenuation or approximate distance from the cabinet?  Snap offer VDSL from $95 a month so getting is no problem.

Currently 14391/1176 on ADSL2+ with 26.5 dB downstream, with Fyx (Maxnet). I'm not sure exactly how the cable runs from the local cabinet but I'm guessing 1100-1200 m. The longest one on John's graph is 1000 m but even then it has a decent upload speed (which is the main thing that I'm interested in).

I'm unsure about Snap as their website is almost completely devoid of information.

1 | 2 | 3 | 4
View this topic in a long page with up to 500 replies per page Create new topic

Twitter »
Follow us to receive Twitter updates when new discussions are posted in our forums:



Follow us to receive Twitter updates when news items and blogs are posted in our frontpage:



Follow us to receive Twitter updates when tech item prices are listed in our price comparison site:





Trending now »

Hot discussions in our forums right now:

22nd Only: PB Tech BROTHER HL1110 Mono laser Printer $15 shipped(after $30 cashback)
Created by loceff13, last reply by old3eyes on 22-Oct-2014 18:40 (18 replies)
Pages... 2


Who Audits IRD?
Created by gundar, last reply by charsleysa on 22-Oct-2014 15:52 (18 replies)
Pages... 2


Spark Socialiser
Created by freitasm, last reply by freitasm on 22-Oct-2014 18:39 (34 replies)
Pages... 2 3


American legal jurisdiction in New Zealand
Created by ajobbins, last reply by gzt on 21-Oct-2014 14:58 (30 replies)
Pages... 2


HERE Maps beta available to all Android 4.4 devices and up
Created by freitasm, last reply by hamish225 on 22-Oct-2014 17:54 (14 replies)

Another Trade Me competitor: SellShed
Created by freitasm, last reply by SellShed on 22-Oct-2014 11:54 (42 replies)
Pages... 2 3


Abnormal upstream data usage - Vodafone Cable Wellington
Created by otherside, last reply by otherside on 22-Oct-2014 17:11 (12 replies)

Snap have failed our company!
Created by dafman, last reply by toejam316 on 22-Oct-2014 13:03 (25 replies)
Pages... 2



Geekzone Live »

Try automatic live updates from Geekzone directly in your browser, without refreshing the page, with Geekzone Live now.

Are you subscribed to our RSS feed? You can download the latest headlines and summaries from our stories directly to your computer or smartphone by using a feed reader.

Alternatively, you can receive a daily email with Geekzone updates.