dslreports logo
 
    All Forums Hot Topics Gallery
spc
Search similar:


uniqs
75753

maartena
Elmo
Premium Member
join:2002-05-10
Orange, CA

maartena to houkouonchi

Premium Member

to houkouonchi

Re: 45 MB internet service

said by houkouonchi:

I think you can ask any FPS gamer and they will agree that 20ms is significant.

No FPS gamer with a brain would have signed up for U-verse in the first place, and stuck with cable or traditional DSL. The latency issue is one of the easiest things to find with a few Google searches.

houkouonchi
join:2002-07-22
Ontario, CA

houkouonchi

Member

said by maartena:

said by houkouonchi:

I think you can ask any FPS gamer and they will agree that 20ms is significant.

No FPS gamer with a brain would have signed up for U-verse in the first place, and stuck with cable or traditional DSL. The latency issue is one of the easiest things to find with a few Google searches.

Hope that wasn't aimed at me but agreed. Honestly I never used uverse for gaming when I still lived with my dad. I actually also had DSL and the fact that they made me cancel my DSL and lose static IPs of 6+ years is one of the reasons it took me so long to signup for uverse.

I wanted uverse simply for the extra speed.

Once I got uverse I re-added DSL and used that for gaming. I still did test latency just to see how it was and was stuck with uverse only for a few weeks until I could get DSL installed again after getting uverse installed. I also had cable but unfortunately cable can also really suck for gaming at times as well. I actually did a post after i finally did get uverse with snapshots of the lagometer to the quakelive LA server between uverse, charter, and dslextreme dsl and there was a pretty significant difference:

»Re: Finally getting U-verse Will attempt to add DSL after its up

Also I was not the only one to report 60+ pings to local quakelive servers from uverse:

»Quake Live Players in Socal, High pings to LA based servers
WhyMe420
Premium Member
join:2009-04-06

WhyMe420 to houkouonchi

Premium Member

to houkouonchi
Well quake live is well known for bad netcode thus, you are correct, latency matters a lot more in that game than it does in most modern FPS games.

In those examples you stated, sure L.A. to seattle is 25ms on FiOS, but that would only be 45ms on U-verse, still well within acceptable latency for 99% of games including FPS games. LA to Chicago would be 71ms, again still within range. Anything below 100ms is generally acceptable.

houkouonchi
join:2002-07-22
Ontario, CA

houkouonchi

Member

said by WhyMe420:

In those examples you stated, sure L.A. to seattle is 25ms on FiOS, but that would only be 45ms on U-verse, still well within acceptable latency for 99% of games including FPS games. LA to Chicago would be 71ms, again still within range. Anything below 100ms is generally acceptable.

LOL. No even with good netcode just under 100ms is not acceptable. With a good player just a 20-30 ping difference can be huge as even if you shoot at the same time you enter into a race condition and it still comes out to whoever's packets make it first wins. If you are just a casual gamer than I can see that type of statement being made. If your a 15 year quake 3 veteran then no...
WhyMe420
Premium Member
join:2009-04-06

WhyMe420

Premium Member

As I said. Games like Quake, especially the old ones, yes with their bad netcode ping matters. But any modern FPS with good netcode does just fine with anything under 100ms. It's very simple, as I said, human reaction speed is 100ms, therefore with any ping lower than that, anything that you "feel" is pure placebo. I understand that people like to have higher e-penises with lower ping, but having a lower ping doesn't automatically make you a great player, nor does having a reasonable (under 100ms ping) make you a terrible player.

It's like saying only those with 120Hz monitors are hardcore gamers, and they will always play better than those on 60Hz monitors.

Yet again, I'll state it, in your example, your Chicago ping is not ideal for all FPS games, even on FiOS. In a game like Quake, with bad netcode, you will want to be on a closer/lower ping server. If a 20ms difference is going to make-or-break your connection, then again as I said, you need to reconsider which server you are playing on.

houkouonchi
join:2002-07-22
Ontario, CA

houkouonchi

Member

We will just have to agree to disagree on this issue. Even with perfect server-side lag compensation and good netcode there 20ms is noticable with two good players. 100 ping vs 20 ping is hugely noticeable especially when it makes the difference between getting the first shot in and messes with recoil in games like CS:S or q3ut. Also guess what, no type of netcode is going to really properly compensate for projectile weapons especially at 100ms. When I say projectile I mean ones that arent instant hit weapons IE on quake:

grenade launcher
rocket launcher
plasma gun
bfg

Metatron2008
You're it
Premium Member
join:2008-09-02
united state

Metatron2008 to WhyMe420

Premium Member

to WhyMe420
said by WhyMe420:

As I said. Games like Quake, especially the old ones, yes with their bad netcode ping matters. But any modern FPS with good netcode does just fine with anything under 100ms. It's very simple, as I said, human reaction speed is 100ms, therefore with any ping lower than that, anything that you "feel" is pure placebo. I understand that people like to have higher e-penises with lower ping, but having a lower ping doesn't automatically make you a great player, nor does having a reasonable (under 100ms ping) make you a terrible player.

It's like saying only those with 120Hz monitors are hardcore gamers, and they will always play better than those on 60Hz monitors.

Yet again, I'll state it, in your example, your Chicago ping is not ideal for all FPS games, even on FiOS. In a game like Quake, with bad netcode, you will want to be on a closer/lower ping server. If a 20ms difference is going to make-or-break your connection, then again as I said, you need to reconsider which server you are playing on.

20 Ms makes a HUGE difference in ut3 and tf2 with lots of rockets going by. Stop talking about epenises and talk about something you understand..
WhyMe420
Premium Member
join:2009-04-06

WhyMe420 to fltelman

Premium Member

to fltelman
Troll responses aside... (Not you houkouonchi) Of course 100ms is not ideal, it's the bare minimum for FPS gaming, but my point is that anything below will still be faster than human reaction time. Of course the person with 30ms ping will "fire sooner" than the person with a 100ms ping, but the same can be said about the person with the 3ms ping compared to the 30ms ping player. That's where netcode comes in.

30ms is still 10 times slower than 3ms, but that doesn't automatically mean that the player will be 10 times worse, 10 times slower to respond, or miss 10 times more shots.

Nuckfuts
Premium Member
join:2003-10-18
Joliet, IL

Nuckfuts to maartena

Premium Member

to maartena
I use U-Verse for online gaming on LIVE. Never had a problem, hosted rooms with people saying the server ran great, and only saw my first-hop pings go up 5-7ms from traditional xDSL to VDSL. I even played Quake3 arena a bit on my PC and it ran pretty good. Not everyone has high-latency VDSL here. I have a friend in Texas who switched from U-Verse to Comcast and he has latency issues all the time. I thought this thread was about 45MB profile=)
StLCardsFan
join:2011-06-06
Lafayette, LA

1 edit

StLCardsFan to WhyMe420

Member

to WhyMe420
said by WhyMe420:

Troll responses aside... (Not you houkouonchi) Of course 100ms is not ideal, it's the bare minimum for FPS gaming, but my point is that anything below will still be faster than human reaction time. Of course the person with 30ms ping will "fire sooner" than the person with a 100ms ping, but the same can be said about the person with the 3ms ping compared to the 30ms ping player. That's where netcode comes in.

30ms is still 10 times slower than 3ms, but that doesn't automatically mean that the player will be 10 times worse, 10 times slower to respond, or miss 10 times more shots.

said by WhyMe420:

Troll responses aside... (Not you houkouonchi) Of course 100ms is not ideal, it's the bare minimum for FPS gaming, but my point is that anything below will still be faster than human reaction time. Of course the person with 30ms ping will "fire sooner" than the person with a 100ms ping, but the same can be said about the person with the 3ms ping compared to the 30ms ping player. That's where netcode comes in.

30ms is still 10 times slower than 3ms, but that doesn't automatically mean that the player will be 10 times worse, 10 times slower to respond, or miss 10 times more shots.

This is not correct. Netcode is some generic term for I really don't understand.

If your ping is 100ms ..that also means 100ms plus overhead plus the ping of the other player you are playing. No "netcode" corrects that. The only thing that can possibly happen is the person with the "lowest" latency gets screwed with buffers.

In the world of online gaming ...with this "netcode" ... the game is only as good as the person with the shittiest connection.

So what really happens ...if you are playing with an established 100ms ping ... this "netcode" you are referring to will try and establish a connected server with people of like latency ... so ... if you have 100ms ..and everyone else has 100ms or greater ...what you really can say is you add your latency and theirs plus overhead and buffer = 200ms + overhead .... in comparison with servers where similar people are connecting with say fios and have >10ms ... then the total package = 10ms + 10ms + overhead = a much lower latency game.

I haven't even stated the importance of upload bandwidth or the power of your connected NIC and/or gateway. Most gateways will add additional latency when using upload / download simultaneously... something most people don't recognize as when you are initially searching for games only upload is being scrutinized by the game server.

K3
join:2008-05-06
Kankakee, IL

K3 to WhyMe420

Member

to WhyMe420
I think that you guys have gone way off topic with this, therefore you are the trolls. This is about 45Mb internet service from U-verse, not about your precious latency times for gameplay. Start a new topic before this one gets locked out for your offpost gibber gabber. So can we get back on base now? What's the news on this here faster internet? I want to trade this clunker speed for a ferrari
StLCardsFan
join:2011-06-06
Lafayette, LA

StLCardsFan

Member

said by K3:

I think that you guys have gone way off topic with this, therefore you are the trolls. This is about 45Mb internet service from U-verse, not about your precious latency times for gameplay. Start a new topic before this one gets locked out for your offpost gibber gabber. So can we get back on base now? What's the news on this here faster internet? I want to trade this clunker speed for a ferrari

this isn't off topic at all. someone asked about added latency of bonded pairs and VDSL2+ and someone added the point that uverse vdsl2+ is at least 20ms higher than other solutions for home internet. Then some of us gave examples of where higher latency = bad performance for some of us.

I mean why even mention the net police on this thread? Latency is probably the most important aspect of the home connection ...so it is relevant...even if you dont like it....

I for one would also like to know if bonded pairs adds even more latency to an already sluggish vdsl2+ connection.

trparky
Premium Member
join:2000-05-24
Cleveland, OH
·AT&T U-Verse

trparky

Premium Member

said by StLCardsFan:

I for one would also like to know if bonded pairs adds even more latency to an already sluggish vdsl2+ connection.

My guess would be... yes, unfortunately.

Think about it, you're splitting the data stream between two data transfer paths.

Each data transfer path would have to be as close to being perfectly in sync with each other to make sure that packet A with path A and packet B with path B is received and processed at the endpoint at exactly the same time to maintain a coherent TCP/IP connection. If packet A was received and processed even a few milliseconds off from packet B, it can break down the TCP/IP connection.

Unless, there's something in the hardware to make sure that a single TCP/IP connection isn't split between the two transfer paths.
doubleohwhat
join:2008-10-25
Birmingham, AL

doubleohwhat

Member

Unless, there's something in the hardware to make sure that a single TCP/IP connection isn't split between the two transfer paths.

I would imagine that would create more headaches actually.
WhyMe420
Premium Member
join:2009-04-06

1 edit

WhyMe420 to StLCardsFan

Premium Member

to StLCardsFan
said by StLCardsFan:

said by WhyMe420:

Troll responses aside... (Not you houkouonchi) Of course 100ms is not ideal, it's the bare minimum for FPS gaming, but my point is that anything below will still be faster than human reaction time. Of course the person with 30ms ping will "fire sooner" than the person with a 100ms ping, but the same can be said about the person with the 3ms ping compared to the 30ms ping player. That's where netcode comes in.

30ms is still 10 times slower than 3ms, but that doesn't automatically mean that the player will be 10 times worse, 10 times slower to respond, or miss 10 times more shots.

said by WhyMe420:

Troll responses aside... (Not you houkouonchi) Of course 100ms is not ideal, it's the bare minimum for FPS gaming, but my point is that anything below will still be faster than human reaction time. Of course the person with 30ms ping will "fire sooner" than the person with a 100ms ping, but the same can be said about the person with the 3ms ping compared to the 30ms ping player. That's where netcode comes in.

30ms is still 10 times slower than 3ms, but that doesn't automatically mean that the player will be 10 times worse, 10 times slower to respond, or miss 10 times more shots.

This is not correct. Netcode is some generic term for I really don't understand.

If your ping is 100ms ..that also means 100ms plus overhead plus the ping of the other player you are playing. No "netcode" corrects that. The only thing that can possibly happen is the person with the "lowest" latency gets screwed with buffers.

In the world of online gaming ...with this "netcode" ... the game is only as good as the person with the shittiest connection.

So what really happens ...if you are playing with an established 100ms ping ... this "netcode" you are referring to will try and establish a connected server with people of like latency ... so ... if you have 100ms ..and everyone else has 100ms or greater ...what you really can say is you add your latency and theirs plus overhead and buffer = 200ms + overhead .... in comparison with servers where similar people are connecting with say fios and have >10ms ... then the total package = 10ms + 10ms + overhead = a much lower latency game.

This is not correct as 100ms is the ROUND TRIP time not a 1-way measure of latency. 100ms ping means 50ms 1-way. Therefore if you have 100ms and the other player has 100ms, guess what? The latency between you two is... 100ms! A packet of you shooting someone gets sent to the server in 50ms, then sent from the server to the other player in another 50ms. It's quite simple actually.

In other words, if you both have a 100ms ping, then what you are seeing him do is what happened 100ms ago, and the same goes for him. Considering that human reaction time is 100ms, you will not notice the latency. However this is in PERFECT conditions. In reality, if you both have 100ms ping, it will fluctuate, and you can in fact notice minor issues. As I said before, 100ms is far from ideal, it is the bare minimum required for FPS gaming.

That being said, I wonder if someone with pair bonding could post their first hop latency?
StLCardsFan
join:2011-06-06
Lafayette, LA

2 edits

StLCardsFan

Member

thats not right either.. maybe I didnt explain myself well enough. Any online FPS game has a central server...whether it be a dedicated server or player hosted from their home system.

When you connect to a game ... there is one system in the mix that dictates all players... you connect to one computer or server just like everyone else connects to that same computer or server. There is no direct connections from you to other players. This is why player based connections with 5mbit or greater always get automatically picked to host. 3mbit does suffice but runs into bottlenecks and adds lag/jitter to the entire game...for everyone involved. Only the host machine makes a direct connection to everyone else.

You can look at a game server this way .. it is an additional hop or gateway from you to everyone else ..and them to you.

so when player a has 100ms ping from them to the server ..that doesn't mean you see the same ... that 100ms of theirs is to the host ..and then you add the host overhead ..and then your latency. Now the server will try and compensate for higher latency connections by adding a buffer to the lower latency players. So what you are really seeing is what the host computer/connection is seeing plus your latency and the rest of the players latency...

while its been said it takes more than 100ms for human reaction... that is based on 0ms - 100ms. if you are starting at 100ms ...latency is even more exaggerated.

Next time youre playing online ...say COD for instance ..ask in the lobby who has FiOS ... chances are you will rarely if ever find them in your lobbies.... unless of course you have a blazingly fast super-low latency connection. This is I guess what is meant when people throw around the term "netcode."

Now of course if you are manually connecting to say a Dice or EA server ... none of the above makes any difference ...people with shit connections will be laggy ..and those who have good connections will not be laggy.

Lastly ..i am interested to see pair bonding latency too.
WhyMe420
Premium Member
join:2009-04-06

WhyMe420

Premium Member

Again, you're pretty much saying what I'm saying, which is anything less than 100ms is acceptable. That is not to say that 100ms ping is necessarily playable. It depends on the game/netcode.

Anyway, we can argue about 100ms all we want, but U-verse, even with it's higher latency, doesn't approach anywhere near 100ms in most cases. If a server were to have a 100ms ping on U-verse, then it would still be 80ms on even FiOS. There's not a big difference in 80ms and 100ms. The 20ms difference is 1/50th of a second, still 5 times faster than the blink of an eye.

maartena
Elmo
Premium Member
join:2002-05-10
Orange, CA

maartena to fltelman

Premium Member

to fltelman
I'm still annoyed on how AT&T is wanting $$$ from people just to upgrade. Truck roll.... fine, but if you want me to upgrade, you come to my house and get it done without charging me an arm and a leg.

I don't want to be nickled and dimed. I don't have to pay if I upgrade a cable connection to a new speed, if a new modem is needed (e.g. a DOCSIS upgrade), I can just take my old modem to the local TWC office, and exchange it for a new one. And if you already have a good modem, they just up the speed, and off you go.

It's really that I like the U-Verse voice so much (we have 2 lines), but I have been thinking about going back to TWC for cable internet. Either way, I won't be paying for a truck roll.
WhyMe420
Premium Member
join:2009-04-06

WhyMe420

Premium Member

Well, at least around here, as long as U-verse is uncapped, I'll gladly pay for a truck roll if it means that I can avoid Charter and their caps! Plus, U-verse is a lot more reliable for me than Charter is for my friends. Consistent speeds and connectivity, whereas friends on Charter constantly report monthly downtime and slower speeds than they are paying for. I have an SB6120 modem lying around for whenever U-verse gets the cap up and going, but whenever I get Charter (depending on what AT&T does,) I will be there during the install, to make sure everything's as close to perfect as possible. Though, I still don't think it will be as reliable as U-verse. Not looking forward to their VoIP either, as I'm sure it's not as reliable as POTS, but as soon as AT&T implements their caps, I'm done with them.

shortyd999
join:2008-10-21
Birmingham, AL

shortyd999 to fltelman

Member

to fltelman
Wonder if they were to offer higher profiles will they also increase the bit rate for the HD streams for the TV service? If I'm not mistaking each stream is around 5-6mbits. An increase to 10 or maybe 15 would help improve the PQ quality of Uverse TV. Or will they use the extra bandwidth strictly for Internet.
StLCardsFan
join:2011-06-06
Lafayette, LA

StLCardsFan

Member

this would be a good idea... or instead of that offer the entire profile in net speed increases.

Nuckfuts
Premium Member
join:2003-10-18
Joliet, IL

Nuckfuts

Premium Member

I could see it for higher internet speeds for competition or AT&T would just add more streams but leave the bitrate the same. I do not think they really care about their HD PQ.

oneoone
join:2010-01-20
North Hollywood, CA

oneoone to shortyd999

Member

to shortyd999
I pray they do!

Michail
Premium Member
join:2000-08-02
Boynton Beach, FL

Michail to Fakarooz

Premium Member

to Fakarooz
said by Fakarooz :

At this point, the pricing is not determinined yet, but the 3/1 is going to disappear and the 18/1.5 customers are going to be upgraded to 24/3 if you are currently on the 32M profile while keeping the price he same. Those currently on 24/3 can get a free upgrade to the higher 36/3 tier, but will need to pay for a truck roll to install an iNID. Thus the 36/3 will be at the same price range as the current 24/3 and the new 48/6 will be between $80-$100. VDSL2 does allow for up to 12MHz in spectrum to be used while currently it maxes at around 8.5MHz. AT&T is looking into utilizing this additional spectrum in areas that have good line quality. In addition AT&T is also looking at vectoring but may not be implemented for some time. Also AT&T is considering to spinoff or sellout it's non-Uverse territory. This means areas that do not have FTTP FTTN OR ADSL2+ in order to gain cash for the recent failed buyout of T-mobile and purchase of additional spectrum for its mobility operations.

I would be willing to pay an upgrade for 36/6 but I'm not so sure 36/3 is much of a value compared to 24/3.

fakarooz
@sbcglobal.net

fakarooz to fltelman

Anon

to fltelman
Ok I've received word from the company that we are not going to announce the new speed tiers next thursday (jan. 26). Seems as if we were all ready for the announcement of the roll out, but apparently our bean counters say its not worth it right now to deploy the 60Meg profile. Basically at&t is not ready to invest in pair bonding for those on the current 25/32meg profiles. The iNID are pretty expensive and labor costs to change out customer equipment and install the iNID outweighs the profit potential of Uverse.

So it seems at&t is going to wait until next quarter to announce new speed tiers. What we are currently looking at right now is simply an upgrade to the 45Meg profile to anyone that has line statistics to support that tier. If you don't you're stuck on the lower 32meg / 25 meg profiles.

With the new 45Meg profile, we will see one new hsi speed tier at 36mbps down / 6mbps up. The 3/1 hsi speed tier will disappear for the FTTP/FTTN customers. iNIDs will still only be used on customers to bring their profiles up to the 25megs. Pricing is still undetermined for the 36/6 tier, but expect it to be $10 more than the current 24/3 speed tier.

Thank you all for being at&t customers! Hopefully 2021 will be filled with many surprises to all of our at&t customers!
doubleohwhat
join:2008-10-25
Birmingham, AL

doubleohwhat

Member

said by fakarooz :

Hopefully 2021 will be filled with many surprises to all of our at&t customers!

Is that the year they plan on going with fiber throughout?

Jokes aside... what you're saying is that those of us with lines capable of handling the 45meg profile will get automatically upgraded to that profile. Is that correct or will we have to call in and request the profile upgrade?

Metatron2008
You're it
Premium Member
join:2008-09-02
united state

Metatron2008

Premium Member

2021 is probably when to expect to pay for at&ts bailout...
StLCardsFan
join:2011-06-06
Lafayette, LA

2 edits

StLCardsFan to fakarooz

Member

to fakarooz
said by fakarooz :

Ok I've received word from the company that we are not going to announce the new speed tiers next thursday (jan. 26).

So it seems at&t is going to wait until next quarter to announce new speed tiers.

LULZ. Wonder how many more customers ATT is gonna bleed. I mean really ... in the St. Louis market ... charter 100mbit internet can be had for about 80 bucks ..60mbit for 60 bucks ... compared to 64.99 for 24mbit uverse.

Do the math.

I would say ATT decided that charging 80 bucks and beyond for internet not even half the speed of competitors is not a good idea... meanwhile Charter is slowly winning back its once huge customer base with better customer service, much better PQ... and well ... seem to have gotten their ducks in a row.... thanks to ATT.

So what ATT should do ..or did ...or who knows ... is add higher tiers and lowered the pricing of current tiers.

I think as time progresses and more things shift to net throughput ..consumers are becoming more educated...and ATT will continue to fall behind.

AnonMan
@sct.com

AnonMan

Anon

I wonder if AT&T knows that the POS router they provide the wifi cant support beyond the 24/3 plan because 54G with overhead maxes out at about 25Mbit lmao.

So they will just make more issues, people will upgrade but be on wifi and complain they wont get the speeds because 54G cant handle beyond 25Mbit... Is funny, the router itself is somewhat decent, just firmware locked down and crappy Wifi... Oh well.. Good thing I use my own wifi, just sucks that I have to when I paid $100 for one. (HSI only user)

djrobx
Premium Member
join:2000-05-31
Reno, NV

djrobx to fakarooz

Premium Member

to fakarooz

apparently our bean counters say its not worth it right now to deploy the 60Meg profile

OK. What you had posted previously sounded good, but if the bean counters don't like it, it looks like I need to switch.

I checked my options with Time Warner. I can either get 30/5 for $59 (12 month promo) or 50/5 for $99, with an included Wireless "N" router. No install fees. With the increased equipment fees and new addition of taxes to my bill, my U-verse internet bill has gone up $5 per month. I'm now paying $70/month. There's a lot of potential for shenanigans with Time Warner, but switching to wherever the deal is best is pretty painless when you're not bundled.

I don't think I'm going to qualify for the 45mbps tier without pair bonding. My max rate is 53mbps. I recall that 42mbps was about the lower limit for 32mbps. Extrapolating, I need about 60mbps to qualify for 45.