ISP's with the 'best' (lowest) international latency

Forums › New Zealand Broadband › ISP's with the 'best' (lowest) international latency

1 | 2 | 3 | 4

34 posts

Geek

#1982997 25-Mar-2018 19:07

Linux:

Sorry but the 20 to 25ms difference would not make any difference

Linux

I mean everyone's different. That's like telling an audiophile there's no difference in two different pairs of headphones. 50ms to 75ms is a 50% jump. For people who are attentive to that kind of thing, it's definitely noticeable.

ctucks

34 posts

Geek

#1982998 25-Mar-2018 19:09

We don't have a master filter, no. We were meaning to get one installed a while back but never got around to it. It's definitely something I regret not doing now, but I suppose it doesn't matter anymore.

michaelmurfy

meow
13265 posts

Uber Geek

Moderator

ID Verified

Trusted

Lifetime subscriber

#1983012 25-Mar-2018 20:21

Please - can we honestly stop with these discussions?

50ms to 75ms is a 50% jump. For people who are attentive to that kind of thing, it's definitely noticeable.

That is incorrect. If we put you to a blind test you wouldn't be able to tell. The whole "ping" thing in games and thinking you're somehow better with a lower ping time is incorrect as there is no way your reaction time is this good. What you're seeing is a pure placebo just like how a monster HDMI cable could increase your image quality.

However - no ISP is right for you. You're not going to notice any difference with even changing. Orcon and 2degrees are potentially the best but again your bottleneck is having VDSL which isn't great for gaming.

Really - just enjoy your game. Don't look at the ping. You're going to have a slight disadvantage (and by slight - I mean it) playing a server based in the US but your ping suddenly dropping by lets say 20ms isn't going to magically make you a better gamer, it may however give you a placebo effect but that is it.

What I am trying to say is it is only noticeable because you've seen it go up but you wouldn't have noticed it otherwise.

Michael Murphy | https://murfy.nz
Referral Links: Quic Broadband (use R122101E7CV7Q for free setup)

Are you happy with what you get from Geekzone? Please consider supporting us by subscribing.
Opinions are my own and not the views of my employer.

ctucks

34 posts

Geek

#1983090 25-Mar-2018 23:06

Except I do notice a difference? I appreciate a lot of people that message because they don’t waltz in and immediately act like they own the place. I’m asking a question, and looking for constructive messages. It seems like you didn’t even read my initial post.

I feel like you’re the kind of guy that will tell people there isn’t a difference between 60hz and 120hz. People want to get the best available service they can, and I personally want to minimise my latency as much as possible.

Anyways, thanks to the people trying to help. I’m not trying to bring in any toxicity here, but it seems that every time I make a post, someone has to jump on their high horse and lecture people.

I think I’m gonna stick with my initial idea and get Orcon Fibre.

ripdog

548 posts

Ultimate Geek
Inactive user

#1983106 26-Mar-2018 02:46

There's nothing geekzoners like more than to speak down at people and tell them "no, you DON'T want that!". I see it time and time again, and it's never called out.

Seriously, let the guy optimize for latency, we're all geeks here, what's wrong with some nice data on which ISPs have the lowest latency on average?

michaelmurfy

meow
13265 posts

Uber Geek

Moderator

ID Verified

Trusted

Lifetime subscriber

#1983207 26-Mar-2018 09:59

ctucks: Except I do notice a difference? I appreciate a lot of people that message because they don’t waltz in and immediately act like they own the place. I’m asking a question, and looking for constructive messages. It seems like you didn’t even read my initial post.

Re-read my post. You only notice it because you see the ping. If you didn't see the ping you wouldn't notice any difference. It is a pure placebo you're falling into.

Personally, I think 2degrees are better latency and capacity wise but really do what you want here as it isn't going to make any difference.

Regarding your monitor side of things - you're right, but this is in no way related.

webwat

2036 posts

Uber Geek

Trusted

#1983513 26-Mar-2018 14:52

ctucks:

I wouldn't be surprised. When our old VDSL connection with Orcon was in the early days of it's profile switching, every other week, on a Saturday night, our profile would swap to a decently low latency one with a significantly faster speed. We'd experience 0 issues on our end (no disconnections, packet loss etc...) But then 2-3 days afterwards, it'd reset and we'd lose the speed and latency drop. Weird.

It probably responded to line conditions that may or may not have been house wiring.

ctucks:

We don't have a master filter, no. We were meaning to get one installed a while back but never got around to it. It's definitely something I regret not doing now, but I suppose it doesn't matter anymore.

Always best to have a VDSL splitter and dedicated outlet to be sure. If you are getting lags or anything on your gaming then a proper splitter would be the very first thing to fix, normally before the connection goes live.

ripdog:

There's nothing geekzoners like more than to speak down at people and tell them "no, you DON'T want that!". I see it time and time again, and it's never called out.

Seriously, let the guy optimize for latency, we're all geeks here, what's wrong with some nice data on which ISPs have the lowest latency on average?

Some people (especially geeks) cannot accept arguments that something thats false should really be true, just because people want to believe it without any evidence to prove otherwise. In terms of pings, so-called latency is less relevant than QoS for different types of traffic because restricting some types of traffic has potential to improve a "ping". Obviously the ISPs try to give realtime traffic such as gaming or voice higher priority, but its too complex to be boiled down to a "ping" time. Perceptions are interesting: you cannot distinguish individual movie frames at 15 fps, but some things might affect the subconcious without being detected. So no you can't detect a 25ms difference in latency unless you are superhuman, but you probably can detect queueing of video traffic that was coincided with a negligible difference in "ping" times. Keep up the good work with the placebos though, its the best way to keep you happy!

Time to find a new industry!

Sharesies

Trade NZ and US shares and funds with Sharesies (affiliate link).

unsignedint

69 posts

Master Geek

#1984222 27-Mar-2018 12:13

FWIW here are my latency results from a little while ago...

https://gist.github.com/unsignedint/6aa9b35d0202509f91b38a7957d6a3e6

Methodology is included, tested across 335 end-points.

surfisup1000

5288 posts

Uber Geek

#1984242 27-Mar-2018 12:33

Linux:Sorry but the 20 to 25ms difference would not make any difference

Linux

All other things being equal, it would make a difference .

Why do you think it wouldn't make a difference?

Linux

11434 posts

Uber Geek

Trusted

Lifetime subscriber

#1984249 27-Mar-2018 12:51

surfisup1000:

Linux:Sorry but the 20 to 25ms difference would not make any difference

Linux

All other things being equal, it would make a difference .

Why do you think it wouldn't make a difference?

No difference to the end user or game play

Linux

Talkiet

4793 posts

Uber Geek

Trusted

#1984262 27-Mar-2018 13:29

ripdog:

Seriously, let the guy optimize for latency, we're all geeks here, what's wrong with some nice data on which ISPs have the lowest latency on average?

Mainly the flawed assumption that the data won't change day to day, and that the signal/noise ratio or this sort of thing is so bad that basing any decisions on the data is essentially the same as rolling a dice. There are more important things to be concerned about, and there are also reasons why sometimes a network design decision MIGHT mean slightly higher latency in some cases, but this has benefits elsewhere.

For example, what was the impact on Spark customers of the fibre cut in Aussie a few days ago that affected a lot of people? Answer, Just about nothing - and it's not because we didn't have traffic and links that were affected... It's because our service design took account of that sort of (and magnitude of) failure and had mitigations designed in.

For someone to suggest than GGIS routing or international performance is bad is both A) an insult to the people that spend a lot of time making it work as well as possible and B) ignorant of that fact that GGIS has multiple services they offer to RSPs. Individual RSPs make investment choices that reflect their desire to protect against various failures and shocker, not all RSPs make the same choice.

Finally, applications are so good now that a few ms will barely have a perceptible affect (20 to 25ms is detectable in synthetic tests, but won't have an impact on any real world use; 150-170ms is again detectable, but will have a barely perceptible impact)

Cheers - N

Please note all comments are from my own brain and don't necessarily represent the position or opinions of my employer, previous employers, colleagues, friends or pets.

Talkiet

4793 posts

Uber Geek

Trusted

#1984264 27-Mar-2018 13:32

unsignedint:

FWIW here are my latency results from a little while ago...

https://gist.github.com/unsignedint/6aa9b35d0202509f91b38a7957d6a3e6

Methodology is included, tested across 335 end-points.

Good dataset, but I can confirm that it's no longer relevant to Spark customers at least due to changes made in the service since the testing was done. Changes have been made specifically to improve latency to some locations (within the last 6 months)

Cheers - N

Please note all comments are from my own brain and don't necessarily represent the position or opinions of my employer, previous employers, colleagues, friends or pets.

ripdog

548 posts

Ultimate Geek
Inactive user

#1984316 27-Mar-2018 14:58

Talkiet, it seems to me like every single one of the arguments you're making are just as applicable against the comparative speed tests which Truenet did. The difference between 5mbps and 6mbps is similarly little when loading webpages, but through the amazing transparency we gained from truenet's testing, NZ ISPs were inspired/shamed into significantly improving their average speeds and especially their consistency around the 9pm time.

So in other words, the best way forward for latency in NZ is not to cease testing because it's 'unfair' or 'not useful', it's to expand the test nationwide and automate it. This is in order to give ISPs the data they need to know if they're doing right by their customers, and customers the power to choose the ISP which tries the hardest.

unsignedint

69 posts

Master Geek

#1984345 27-Mar-2018 15:21

Of course latency matters. It is easily measurable. Ever heard of High Frequency Trading? That is the most extreme example of a latency sensitive application. For people working in that area, they are more than happy to pay for lower latency.

Moreover, some ISPs will generally offer lower latencies because they may have shorter routes. In general, a shorter route from A to B will come at a higher cost.

Talkiet

4793 posts

Uber Geek

Trusted

#1984357 27-Mar-2018 15:36

ripdog:

Talkiet, it seems to me like every single one of the arguments you're making are just as applicable against the comparative speed tests which Truenet did. The difference between 5mbps and 6mbps is similarly little when loading webpages, but through the amazing transparency we gained from truenet's testing, NZ ISPs were inspired/shamed into significantly improving their average speeds and especially their consistency around the 9pm time.

So in other words, the best way forward for latency in NZ is not to cease testing because it's 'unfair' or 'not useful', it's to expand the test nationwide and automate it. This is in order to give ISPs the data they need to know if they're doing right by their customers, and customers the power to choose the ISP which tries the hardest.

I'm not authorised to make specific comments on the Truenet results but I will say that in no way did their results shame or inspire us to make changes. In fact I gave more feedback to truenet to improve the accuracy of their tests than we garnered useful and accurate information from their tests. Up until the end there were results that did not accurately reflect real world performance on some ISPs. (Note that there is a decision due from MBIE or the comcom about an Independent Test Provider to run similar tests - I don't think there's any public announcement yet but I definitely look forward to continued external monitoring of the services in the marketplace.

As to your second point - if lower latency was the only or most significant measure of good routing and investment decisions, I'd agree with you. I don't agree with you. "Doing right by their customers" is NOT always giving lower latency.

Cheers - N

Please note all comments are from my own brain and don't necessarily represent the position or opinions of my employer, previous employers, colleagues, friends or pets.

1 | 2 | 3 | 4