Questions about gaming lag, virgin & infinity...

[Sephiroth]

I think there's EU legislation comin in next year that forces telecoms companies to offer 12 month contracts.

[qasdfdsaq]

Originally Posted by Kushan

Really? The VDSL modems currently in use only have 100M ports? I'm surprised at that. I know the 80Mbit limit is fine on that but still. What happens if/when vectoring rolls out? New modems for everyone?

Pretty much. Though new and upgrading customers will probably be supplied with a next-gen modem+router combo that has multiple gigabit ports.

[Chrysalis]

Originally Posted by Kushan

Really? The VDSL modems currently in use only have 100M ports? I'm surprised at that. I know the 80Mbit limit is fine on that but still. What happens if/when vectoring rolls out? New modems for everyone?

probably not as vectoring isnt going to bring 100mbit+ speeds. Its more a speed enablier to allow more people to get existing speeds. If higher speeds do come it will probably be a small bump to 100mbit not beyond. Higher then that will be if they use a bigger profile or line bonding or when gfast comes.

Also the days are counting down for these modems, soon self install will be rolled out and then user's will get supplied modem/router combo devices.

[DaMac]

So can it be worth the hassle of changing, and is there some way of me finding out how good the service is on my street?

[Kushan]

Originally Posted by Chrysalis

probably not as vectoring isnt going to bring 100mbit+ speeds. Its more a speed enablier to allow more people to get existing speeds. If higher speeds do come it will probably be a small bump to 100mbit not beyond. Higher then that will be if they use a bigger profile or line bonding or when gfast comes.

Also the days are counting down for these modems, soon self install will be rolled out and then user's will get supplied modem/router combo devices.

I don't know a lot about Vectoring, but I kept seeing 120Mbit floated around as a figure. Sure, as you say it's more to help people who struggle to get the 76/80Mbit speeds actually attain them but surely just equipping a gigabit port would help eek out a few extra mbits for those who can get it? I know when Virgin upped me to 100Meg (actually 105), I struggled to get more than about 94Mbit using the D-Link with its 100Mbit ports.

[Chrysalis]

Originally Posted by Kushan

I don't know a lot about Vectoring, but I kept seeing 120Mbit floated around as a figure. Sure, as you say it's more to help people who struggle to get the 76/80Mbit speeds actually attain them but surely just equipping a gigabit port would help eek out a few extra mbits for those who can get it? I know when Virgin upped me to 100Meg (actually 105), I struggled to get more than about 94Mbit using the D-Link with its 100Mbit ports.

120mbit I think is unlikely to be rolled out from vectoring alone.

To me as well 94mbit,100,120mbit no realistic difference between all 3, once you get to a certian speed its dimishing returns. BT dont need to rollout 120 on FTTC as they already the market leaders with their FTTP/FTPPoD products.

[qasdfdsaq]

Originally Posted by Chrysalis

probably not as vectoring isnt going to bring 100mbit+ speeds. Its more a speed enablier to allow more people to get existing speeds. If higher speeds do come it will probably be a small bump to 100mbit not beyond. Higher then that will be if they use a bigger profile or line bonding or when gfast comes.

I know a few people who already have lines capable of 140 when they were installed but have since dropped to 95 because of crosstalk - exactly the sort of situation where vectoring is supposed to restore the full 140. My line with is "average" length for a city does over 110Mbps without crosstalk and 80 with. Again, likely to increase to 150-200 with 30a.

Most people on the 76/80Mbps tariff already get existing speeds - 70% get 70Mbps or more - they don't need any speed enablers. It's the ones on 38/40Mbps and below it's expected to benefit the most.

Additionally Openreach are rumoured to be implementing profile 30a at the same time as vectoring which will double attainable speeds on short lines - up to e.g. 280Mbps without crosstalk in the real world.

Also the days are counting down for these modems, soon self install will be rolled out and then user's will get supplied modem/router combo devices.

That we can agree on.

---------- Post added at 07:16 ---------- Previous post was at 07:14 ----------

Originally Posted by Kushan

I don't know a lot about Vectoring, but I kept seeing 120Mbit floated around as a figure. Sure, as you say it's more to help people who struggle to get the 76/80Mbit speeds actually attain them but surely just equipping a gigabit port would help eek out a few extra mbits for those who can get it? I know when Virgin upped me to 100Meg (actually 105), I struggled to get more than about 94Mbit using the D-Link with its 100Mbit ports.

As mentioned, new modem/routers will become widespread before the new speeds.

And yes 120 is what I'd expect to be the headline figure for the upgraded current service, with 160/200 being implemented as a new tier. Most people on 76/80 already get 70+ as I've mentioned several times.

[Sephiroth]

Originally Posted by qasdfdsaq

.......

Most people on the 76/80Mbps tariff already get existing speeds - 70% get 70Mbps or more - they don't need any speed enablers. It's the ones on 38/40Mbps and below it's expected to benefit the most.

.....

Here's where my circuit differs. When I was on the 38/40 service, that is what I got. 38 meg.

As soon as the went to 17a, my download speed has never risen above 55 meg, which I put down to something in the 350 m to the cabinet. Aluminium, perhaps. I don't know if vectoring will help that situation.

[Chrysalis]

Originally Posted by qasdfdsaq

I know a few people who already have lines capable of 140 when they were installed but have since dropped to 95 because of crosstalk - exactly the sort of situation where vectoring is supposed to restore the full 140. My line with is "average" length for a city does over 110Mbps without crosstalk and 80 with. Again, likely to increase to 150-200 with 30a.

Most people on the 76/80Mbps tariff already get existing speeds - 70% get 70Mbps or more - they don't need any speed enablers. It's the ones on 38/40Mbps and below it's expected to benefit the most.

Additionally Openreach are rumoured to be implementing profile 30a at the same time as vectoring which will double attainable speeds on short lines - up to e.g. 280Mbps without crosstalk in the real world.


That we can agree on.

---------- Post added at 07:16 ---------- Previous post was at 07:14 ----------


As mentioned, new modem/routers will become widespread before the new speeds.

And yes 120 is what I'd expect to be the headline figure for the upgraded current service, with 160/200 being implemented as a new tier. Most people on 76/80 already get 70+ as I've mentioned several times.

I cant reveal too much sadly.

But you have to consider since there is a 40/10 and 40/2 product people who are unable to sync higher than 40 are likely to be on the lower product which in turn stuff's the stats somewhat. If my line only synced at 30mbit eg. why would I pay for the 80mbit package?

Whilst 140 is possible, its not a very high %, certianly not 10%. BT have learnt a little about managing expectations which is one reason we dont have vdsl max (fully rate adaptive no cap) by comparison to your city area example, I started off on 110mbit and am now down to a 71mbit sync, attainable is currently 70 so if I resync again my sync will go down. If someone syncs at 80 on a line that can do 100mbit and as such has a nice high snrm so little chance of been interleaved, they are much less likely to complain than someone with a line that can do the same 100mbit but the max speed is 140mbit on the product, the line is running at just 6db snrm, has too many errors, gets interleaved and the whole situation explodes. I think anything over 100mbit is very unlikely when vectoring is enabled unless its combined with bonding. Personally I think BT wont touch profile 30, gfast will come after vectoring. Also the way the dslam vendors are marketing vectoring to companies like BT is not to roll out lines pushing vdsl2 to the absolute limits but rather allowing rollout of speeds of 100mbit at 400m, isp's using whats capable at 400m as the benchmark.

I respect your opinion and I hope you respect mine, but given the farce adsl2+ is/was with marketed speeds I dont think BT want a repeat and as such will not market speeds based on a 0m attainable distance.

---------- Post added at 17:36 ---------- Previous post was at 17:30 ----------

Originally Posted by Sephiroth

Here's where my circuit differs. When I was on the 38/40 service, that is what I got. 38 meg.

As soon as the went to 17a, my download speed has never risen above 55 meg, which I put down to something in the 350 m to the cabinet. Aluminium, perhaps. I don't know if vectoring will help that situation.

When you get your unlocked modem seph you can check the quiet line noise level, thats an indicator of crosstalk levels on the line.

[qasdfdsaq]

Originally Posted by Sephiroth

Here's where my circuit differs. When I was on the 38/40 service, that is what I got. 38 meg.

As soon as the went to 17a, my download speed has never risen above 55 meg, which I put down to something in the 350 m to the cabinet. Aluminium, perhaps. I don't know if vectoring will help that situation.

Everybody's circuit differs, obviously yours is below average for a 76/20 customer but many in your situation wouldn't have gone for the upgrade at all (hence why the average is higher).

Vectoring won't help aluminium cable be less crap, but if that is being compounded by crosstalk then the latter will be helped.

---------- Post added at 23:58 ---------- Previous post was at 23:36 ----------

Originally Posted by Chrysalis

But you have to consider since there is a 40/10 and 40/2 product people who are unable to sync higher than 40 are likely to be on the lower product which in turn stuff's the stats somewhat. If my line only synced at 30mbit eg. why would I pay for the 80mbit package?

Well yes, that's certainly true. If everybody was on 76/20 then the averages would be lower, people who can't get more than 40/10 won't be contributing to the figures because they would be best off on the lower service. How many people would move from the lower band to the upper band as a result of vectoring improvements though, that's anybody's guess.

Whilst 140 is possible, its not a very high %, certianly not 10%. BT have learnt a little about managing expectations which is one reason we dont have vdsl max (fully rate adaptive no cap) by comparison to your city area example, I started off on 110mbit and am now down to a 71mbit sync, attainable is currently 70 so if I resync again my sync will go down.

Indeed, it's not often you see lines go that high but point is simple, it exists in the real world and it does happen. Not a very high % got 24Mbps on 24/2 either, but some still did. But yours is the exact situation that vectoring will drastically improve, bringing your line back up to 100+ without profile 30a which will further improve speeds drastically.

If someone syncs at 80 on a line that can do 100mbit and as such has a nice high snrm so little chance of been interleaved, they are much less likely to complain than someone with a line that can do the same 100mbit but the max speed is 140mbit on the product, the line is running at just 6db snrm, has too many errors, gets interleaved and the whole situation explodes. I think anything over 100mbit is very unlikely when vectoring is enabled unless its combined with bonding.

If 70% of people currently get 70+ and we assume they get similar amounts of crosstalk to you, then that means 70% of people will get over 100Mbps once vectoring is enabled. That certainly doesn't concur with the idea anything over 100 is very unlikely. You said yourself your own line can do at least 110 (assuming your 110 result was at zero crosstalk).

Personally I think BT wont touch profile 30, gfast will come after vectoring.

Perhaps, but it seems you are "widely" disagreed with:

http://www.ispreview.co.uk/index.php...s-bt-fttc.html

A further speed boost is also likely to surface in 2013 when BT is widely expected to increase its spectrum allocation to 30MHz (Profile 30a), which could result in headline speeds of above 100Mbps (BT hinted to ISPreview.co.uk last year that 120Mbps was a possibility)

Also the way the dslam vendors are marketing vectoring to companies like BT is not to roll out lines pushing vdsl2 to the absolute limits but rather allowing rollout of speeds of 100mbit at 400m, isp's using whats capable at 400m as the benchmark.

So when the majority of urban lines are <350m, and the national average line length is around 400-500m, and the "benchmark" is 100Mbit at 400m, then how do you come up with the conclusion 100Mbit is "very unlikely"?

Supposedly some others think vectoring + 17a provides 100Mbit out to 600m, not 400.

I respect your opinion and I hope you respect mine, but given the farce adsl2+ is/was with marketed speeds I dont think BT want a repeat and as such will not market speeds based on a 0m attainable distance

Marketing and attainability are always different, as you know. But I'd rather have an "Up to 24Mbps" service that does "Up to 24Mbps" than have what I am actually getting obfuscated by some arbitrary "realistic" marketing horsecrap when I can no longer tell if what I'm getting is really "Up to 16Mbps" or actually "Up to 24Mbps". Even then, marketing a 24Mbps service as 16Mbps is one thing, artificially capping everybody on a 24Mbps service to 16Mbps is another. In that vein, I'd rather have an "Up to 200Mbps" service knowing full well my line was likely to be able only do 120 instead of having an artificially capped "Up to 100Mbps" service knowing full well my line was actually able to do 120. Why penalize people who actually live 0m from the cab.

[Chrysalis]

I meant I think its unlikely BT will market higher than 100mbit just on vectoring alone.

If they did rollout 120 with profile 30 and vectoring or just vectoring/profile 30 alone I guess it makes you happy but I dont see the point, BT already win on headline speeds with FTTP and FTTPoD (which soon will be available in all FTTC areas), and most FTTC consumers are on 40/10 or 40/2.

Regarding my 110, sadly I didnt create any graphs, I unlocked the modem, the first night in the gui it showed 110, which dropped to 105 in the evening.
The next day it was at 90, which I then did make graphs, those graphs showed some QLN but it was low.

Now is some confusion, apparently when I got enabled a bunch of adsl lines got knocked offline, I suspect they forgot to enable the adsl power cutback when I was enabled (I was enabled before exchange was officially enabled), so its possible the 90 to 110 difference was due to full power on the adsl tones.

When the line dropped from 90 to 73 there is a clear QLN increase across pretty much all the vdsl frequency with most of it on the first downstream range. So my guestimate is crosstalk at least for the 90 to 73, but not so sure for the 110 down to 90. My install engineer did link both times to new installs tho so he thinks its crosstalk for both.

Noticed there is no STM today - has it been turned off?

[DaMac]

Hope so, can anyone confirm?

The day we can order fibre there's no STM, not complaining but it's a massive co-incidence. (not going to abuse it tho!)

oddly enough upload seems to still be going as speeds been cut in half.

[Martin_D]

Originally Posted by DaMac

Hope so, can anyone confirm?

Not been STM for about 3 weeks now