TR Forums

Full well knowing that this topic will be filled by a bunch of posts that essentially will say "just don't do it" and full well knowing that my admonition to please refrain from such posts will go ignored (please don't) I am looking for tests from a reliable source or from those of you who have done it as I can't seem to find any on my own.

I have recently moved to an apartment in a small building and I have no way to get a wired connection from my computer to my routers location which is less than 20 feet away from my PC and 2 walls. I mostly play FPS (Battlefield 4) games so latency and lag are always an issue.

My current router is a D Link DIR-655, which is getting long in the tooth(4-5 years old I think). On my current network I have 3 devices hooked up through wireless and all are 802.11n, an IPad, Windows 8.1 phone and a Roku streaming device (which actually streams through the Ethernet cable and there are about 3 wireless networks that I can see on my phone. I need to get a wireless adapter to so I can get back to gaming. My thought was to buy an ASUS USB-AC53. I know my current router does not broadcast on the 5 Ghz band but if I have issues gaming on the DIR-655 I always can step up to the ASUS RT-AC56U router and set-up the computer on the 5 Ghz band as the only device.
I don't think the RT-AC66U or the RT-AC68U would do that much for me. FPS games don't use larges amounts of bandwidth and the dual core router on the RT-AC56U would help with any latency issues.
Any help would be appreciated.

I don't recommend USB adapters. USB devices are limited to supporting a single machine, they use drivers which can limit use on other platforms, and can create kernel panics.

I recommend wireless bridges. If you desire you can uplink the wireless bridge into a switch and support many other wired devices regardless of the platform they use. A wireless bridge does not require drivers and thus can't lead to kernel panics.

I do wireless network gaming and see steady and low pings.

I use RT-N66U Multifunction Router & a EA-N66 AP/Bridge on the 5GHz band.

Pville_Piper wrote:

and I have no way to get a wired connection from my computer to my routers location which is less than 20 feet away from my PC and 2 walls.

http://www.amazon.com/Best-Sellers-Elec ... 6577538011

Seems worth it to me.

I do not quite understand the question... Are FPS games still viable over WiFi? Yes, they are - I use WiFi for all home devices (well, except for IP cams and DVR for them because of PoE) for 5+ years, never had issues with games - you can still get excellent ping in BF4, CS:GO, Titanfall and other stuff. I don't use 5GHz (except few devices which are very close to router) due to severe signal attenuation, 2.4GHz band works just fine and doesn't negatively affect my download/upload speeds (which are around 60/25) or in-game latency even with multiple users streaming legal video content. So WiFi works perfectly well for all tasks - you just gotta keep aware of few things, like NEVER using USB WiFi adapters (they don't last very long and some tend to overheat during prolonged intensive usage) and using PCI/PCIe cards instead (I like the ones which use Atheros chipset even though there aren't many of them compared to Ralink/Realtek ones) and preferably using routers with external antennas with a good cooling (some Asus models do get very warm, which does shorten their lifespan), powerful chipset and firmware with less relevant drawbacks (all WiFi routers like Asus RT-AC series, Netgear R7000 and most likely Linksys WRT1900AC have at least some firmware bugs/deficiencies, not necessarily solved by custom firmwares like DD-WRT, you just gotta read what these are and whether they will affect your particular environment).

If you wanna read more - SmallNetBuilder forums have a good section related to Asus routers, plus there are official Netgear/Linksys forums, as well as "unofficial" ones for custom firmwares for these devices.

To avoid all the grief wireless will cause you when gaming, use PowerLine networking.

You can game on wireless, but unless you live in a field with only one wireless network, there will always be contention and jitter.

All good comments so far and y'all have given me a lot to look into.

@ JohnC - the issuer I see is that a lot of people say it doesn't work well due to increased lag, interference, ect... What I am trying to find out is if there is any built in lag to a WiFi system and if there are any other issues that I am not aware of. The big thing that I can see the amount of WiFi in the area and the fact that I don't have a lot of networks in my building. Although I am in an apartment, the building only has 4 apartments in it and it is fairly far from other buildings. As far a throughput I don't need much, the computer only talks to the internet which is 10 mb so it really doesn't matter if I get a middle of the road router as I can only stream what the internet gives me. I figured I would get a 5 GHz unit because it is unlikely that anyone is using it. As for distance I'm only talking 20 feet so I just can't see where a really high end unit will do much good. I thought the dual core router on the RT-AC56U would help with the latency issues. My router is suspect on the wireless side so anything I do wireless will likely involve getting a new router.

@ Chrispy - I am still looking into the Powerline networking. I need to understand the pros and cons better but that idea is promising... Do you use one?

@ Ryu Conner- I really don't see the need for an access point but price wise they are about the same as a USB module. I haven't had the chance to look up the issue you described nor have I seen any indication of this problem in other forums. Would a dedicated card be a better solution than the USB or would the access point still be a better option? The biggest hit I've seen on the USB is the faster ones tend to run a little hot which will likely shorten their life and add to reliability issues.

Pville_Piper wrote:

@ Chrispy - I am still looking into the Powerline networking. I need to understand the pros and cons better but that idea is promising... Do you use one?

Nope, I have CAT6 in the walls at home.

However I've installed Poweline homeplugs several times both as a consultant and a couple of times as a friend helping a fellow gamer.

Yes, it solves the lag spikes, bandwidth and jittery ping that frequently affect wireless connections.

• If your house wiring is a complete mess it's still faster than 2.4GHz wireless and still gives you more consistent pings, though there can still be some jitter.
• If your house wiring is decent (ie, last 50 years or so) you'll probably find it indistinguishable to a direct ethernet connection.

Pville_Piper wrote:

@ Ryu Conner- I really don't see the need for an access point but price wise they are about the same as a USB module. I haven't had the chance to look up the issue you described nor have I seen any indication of this problem in other forums. Would a dedicated card be a better solution than the USB or would the access point still be a better option? The biggest hit I've seen on the USB is the faster ones tend to run a little hot which will likely shorten their life and add to reliability issues.

It's technically a multifunction device an AP and a Bridge, you'd use it in Bridge mode.

A dedicated PCIe Wi-Fi adapter is an acceptable choice in comparison to a bridge. The PCIe Wi-Fi adapter comes with some of the same pitfalls as a USB adapter (potential limited platform support, potential stability woes, and singular use).

Bridges simply make things - IMHO - more straight forward, less problematic, and more adaptable to your needs over time.

I'm hijacking this thread only because I think a powerline adapter is the best solution here.

Question: Is it worth-while to run the "receiving" powerline adapter off a surge protector to protect against power surges? (I could care less if the "transmitting" end fries my $60 router, better that than my $1000 gaming tower) Does this have any negative effects? Am I being paranoid?

Pville_Piper wrote:

@ JohnC - the issuer I see is that a lot of people say it doesn't work well due to increased lag, interference, ect... What I am trying to find out is if there is any built in lag to a WiFi system and if there are any other issues that I am not aware of.

There always is an additional lag, but it's extremely insignificant - I still get sub-50ms ping to all local gaming servers with no noticeable lag spikes, packet loss or any other drawbacks (unless local ISP is doing some "maintenance" work or routes me through overloaded node or whatever, which unfortunately happens very often with EVERY local ISP).

Pville_Piper wrote:

The big thing that I can see the amount of WiFi in the area and the fact that I don't have a lot of networks in my building. Although I am in an apartment, the building only has 4 apartments in it and it is fairly far from other buildings. As far a throughput I don't need much, the computer only talks to the internet which is 10 mb so it really doesn't matter if I get a middle of the road router as I can only stream what the internet gives me. I figured I would get a 5 GHz unit because it is unlikely that anyone is using it. As for distance I'm only talking 20 feet so I just can't see where a really high end unit will do much good. I thought the dual core router on the RT-AC56U would help with the latency issues. My router is suspect on the wireless side so anything I do wireless will likely involve getting a new router.

If you have close distance then 5GHz should provide you with an acceptable signal strength, if not - you can always fall back to 2.4GHz. And it's not about "dual cores" in routers, it's mostly about how mature and stable their firmware is I currently use Netgear's R7000 model - it's still relatively "new" and official firmware has few issues (mostly related to the usage of external storage connected through USB ports and some very buggy functions like QoS which currently should not be used especially with Apple devices) but for the most part (for tasks like gaming or downloading programs or watching online streams on PC or portable devices) it's perfectly usable and should get even better over time Other router manufacturers are unfortunately not really better in this regard (I've used many Asus wireless routers as well) but the convenience of fully wireless access for all devices still beats any minor firmware drawbacks, at least in my current environment.

I would use powerline before wifi for FPS gaming. That or run a long ethernet cable. Sorry if that's not what you want to hear but I would not ever advise wifi for FPS.

I game just fine on wifi. If you want to go the wifi route, make sure you get a good router, like the asus r-n66u. If you're in an area with lots of interference, you need a dual band router. You can see your wifi interference with inssider.
A wired connected will always have lower ping times and a more stable connection.

Whatever anyone else in the thread says about gaming on wireless being subjectively okay, here are the graphs from Valve who actually measured it and published the results.

Don't do competetive gaming on wireless when you have better alternatives; It should be your last resort.

That's a pretty crappy "good" network from Valve. My wireless network doesn't even have close to an average 10ms transmit time.

That being said I don't think anyone here is playing for cash in their games.

Chrispy_ wrote:

Whatever anyone else in the thread says about gaming on wireless being subjectively okay, here are the graphs from Valve who actually measured it and published the results.

Don't do competetive gaming on wireless when you have better alternatives; It should be your last resort.

/thread

Chrispy_ wrote:

Whatever anyone else in the thread says about gaming on wireless being subjectively okay, here are the graphs from Valve who actually measured it and published the results.

Don't do competetive gaming on wireless when you have better alternatives; It should be your last resort.

I'd rather believe my own objective experience instead of random lines on some internets .jpg, thank you very much.

Chrispy_ wrote:

Whatever anyone else in the thread says about gaming on wireless being subjectively okay, here are the graphs from Valve who actually measured it and published the results.

Don't do competetive gaming on wireless when you have better alternatives; It should be your last resort.

Ah, well there we have it... Valve has a graph showing a "good" wireless network and a "bad" wireless network; and the blue line is obviously more squiggly for the "bad" network... so clearly... umm... something?

Plenty of people game on wireless networks without experiencing any problems; in fact, in many cases you would be very hard pressed to even notice a difference between playing on wireless and wired. Just like plenty of people still manage to play over a DSL connection.

All of this stuff about "never" playing over WiFi is bunk. Try it. Chances are it'll work perfectly well. If it doesn't, then you start looking for alternatives. Don't pass up the most obvious solution just because of a graph without any context.

JohnC wrote:

Chrispy_ wrote:

Whatever anyone else in the thread says about gaming on wireless being subjectively okay, here are the graphs from Valve who actually measured it and published the results.

Don't do competetive gaming on wireless when you have better alternatives; It should be your last resort.

I'd rather believe my own objective experience instead of random lines on some internets .jpg, thank you very much.

Especially when the only information we have on the graph is that it represents a "weak signal"... Really? A good signal is better than a weak signal? You don't say...

cphite wrote:

Don't pass up the most obvious solution just because of a graph without any context.

The only other solutions offered so far are

• an $86 wireless bridge
• some floor matting to cover a CAT6 cable
• don't use a USB adapter
• A dedicated PCIe wireless card acting as a brigde ($20-50$)

A pair of powerline adapters cost $29.99 making them close to the cheapest option, and practically guarateeing better latency than wireless for an OP who is primarily concerned with latency.
How is that /NOT/ the obvious solution?

With WiFi we know he has other devices fighting for bandwidth on his own network, we know he has at least three competing networks in range, we know the signal has two walls to get through. None of these bode well for flawless WiFi performance.

cphite wrote:

JohnC wrote:

I'd rather believe my own objective experience instead of random lines on some internets .jpg, thank you very much.

Especially when the only information we have on the graph is that it represents a "weak signal"... Really? A good signal is better than a weak signal? You don't say...

It's not quite just some jpg from the internets if Valve is publishing the data now is it? If you have objective results as you claim JohnC, go ahead and post them here to bolster whichever argument you may have.

the valve link refers to game streaming, not gaming over wifi.

i personally game over wifi almost exclusively, and i've had no issues at all. i tried powerline adapters, and unless something changed drastically over the last year or so, i wouldn't touch that with a ten foot pole. terrible performance, even in a house built around 2010.

USB wireless adapters are perfectly fine. i use an asus USB-N66 with a RT-N66U over the 5 GHz band. terrific performance (5x as fast as the powerline adapter that i tried previously), no measureable jitter, and never disconnects. this is even with numerous other devices connected to the 2.4 GHz band, all the while with the adapter one floor above the router and about 20 feet away. since i've never gotten a "kernel panic" or anything remotely close with any of my USB based adapters in the past, i cannot recommend something like the EA-N66, especially in light of its terrible reviews. wireless bridges aren't worth the trouble to set them up, as far as i'm concerned.

moriz wrote:

5x as fast as the powerline adapter that i tried previously

You're talking about bandwidth, which has very little to do with latency.
Yes, I can get around (real-world) 140MBit/s over my wireless and basic powerline stuff is more like 45-80Mbit, but when you're on 20MBit internet, does it make any difference to gaming?

From Valve's dev blog:

most routers are designed for reliability and high bandwidth scenarios (good for downloading files and streaming movies where video and audio can be buffered) but not for low latency gaming applications. Some wireless routers can also periodically pause or take a second or two to switch from a low power to high power mode. Even if you have a good router, your wireless network may be congested with chatter from other overlapping networks or even your microwave oven

Powerline is not without issues either, YMMV! It's entirely possible to have a huge amount of noise on the mains power circuit in your home too, from a faulty appliance, or something ancient with passive PFC - any open system can be susceptible to outside interference.

Anecdotally whenever I have a LAN party at mine, people on my dual-band 5GHz wireless network complain of lagspikes but the people on wired connections are fine. I've certainly experienced several wireless sessions elsewhere with those "second or two" lagspikes that Valve mention. Like the OP I live in a flat rather than a detached house and yes there are other networks everywhere. My opinion is not worth more than anyone else's, I'm just pointing out that I've had to defer to Powerline in situations similar to the OP's having already tried to upgrade wireless coverage/routers and failed to net any gains.

Chrispy_ wrote:

You're talking about bandwidth, which has very little to do with latency.
Yes, I can get around (real-world) 140MBit/s over my wireless and basic powerline stuff is more like 45-80Mbit, but when you're on 20MBit internet, does it make any difference to gaming?

when the powerline adapter turns my 40/3 MBit/s internet connection into 2/0.5 MBit/s? yeah, there's a cause for concern. also regarding latency: i get about 25 ping with my wireless connection with 0-1 ms jitter, vs 23 ping and 0-1 ms jitter on the computer downstairs connected directly to the router (with pingtest.net). my wireless connection is not going to affect my gaming in any measureable amount.

and as you said, YMMV. to the OP: take advantage of a store with a good return policy, and explore all your options. sometimes, the best solution is the simplest, and you can't get a simpler solution than a good wireless router and usb adapter. if that doesn't pan out, THEN look for more esoteric solutions like powerline.

Chrispy_ wrote:

None of these bode well for flawless WiFi performance.

That's only your subjective opinion, not an objective fact I have 14 wireless devices (everything from TV's, tablets, smartphones, XBox One, PS4, laptops and including my main gaming PC) at home and more than a dozen of other sources of WiFi signals outside of the house, I have absolutely 0 issues with "lag spikes" or increased latency regardless of who uses these devices and in whichever way. If you personally do have issues - you're doing something wrong or you need better hardware with better firmware

EVERY network connection technique has its issues, including powerline adapters - their bandwidth rating tends to be greatly exaggerated, especially over long distances (not an issue at all if you have all devices nearby but can still affect other people with other setups), some of them also (just like some WiFi routers or majority of USB adapters) tend to overheat during heavy usage (you can find some cases in an online product reviews) which shortens their overall lifespan and most of them don't work well when plugged into UPS and other "power conditioning" devices (so if you have short brown-out/blackout - your connection will be lost, which never happens for my WiFi setup because everything is connected to UPS units (Verizon's ONT has its own battery backup) and there is also a generator outside for long-term power loss which can be started before UPS batteries lose their charge). Plus if someone already has few WiFi devices, the most logical step is to simply use one single WiFi router for everything which occupies very little overall space and only a single power outlet in a single room (and keep the older WiFi router as a working spare)

Edit: here's an average powerline adapter FAQ from a typical manufacturer (other manufacturers have similar wording):
http://www.tp-link.com/en/article/?id=406 (yes, it has a horrible translation to English but it's good enough to understand the meaning).

"Q: Why can not my powerline rate reach 200Mbps or 500Mbps?
A: 200Mbps and 500Mbps are the highest powerline rate in theory. However the power line circuit is very complex, there may be much interference. Electrical equipments with electromotor, like washing machine/air-condition, can generate interference may even cut off your powerline connection. If possible, please avoid such devices"

"Q: Can Powerline adapters work if they are separated by UPS?
A: No. Usually, UPS has filter function. It can cut the powerline connection."

This is a latency thread. Reliability aside (and it's not as if wireless is any better) you haven't addressed the only issue I've raised: Latency.

At the moment I'm the only one who's posted anything objective related to latency: Valve showing multiple lag spikes up to 100ms on even a good home WiFi connection, The same article saying that powerline can be good and then immediately explaining how most wireless routers are not good for gaming. I'm sure there are bad powerline implementations and I'm sure there are good wireless combinations for gaming, but if you are going to make a general argument you can't afford to cherry-pick min/max examples that suit a desired outcome.

I figured Valve's hardware dev team would have a shred of credibility but this seems to be a tough crowd. If all you can do is discredit Valve then I think that speaks for itself.
Are they omniscient deities?
- No.
Would they have sunk hundreds of millions of dollars into this SteamBox/OS project and not bothered with proper network testing across multiple routers/cards/dongles in a range of different environments?
- Doesn't seem very likely to me.

moriz wrote:

take advantage of a store with a good return policy, and explore all your options. sometimes, the best solution is the simplest

This is very good advice.
Whilst I apologise for cutting your quote off there, I reckon that inserting two plugs into two power sockets is by far the simplest solution

If you have a second wireless router lying around (or can get one for cheap on CL or eBay) you can flash it with DD-WRT and then set it up as a client bridge. I'd recommend trying this before buying a PCIe card or USB dongle.

The thread is about latency and whilst you are not showing many WiFi networks nearby, there can be other transmissions (more on the 2.4Ghz band than the 5Ghz band) that can knock out your signal and cause a retransmission which increases the latency. The same is true for powerline networks as well as others have posted if you get an appliance putting a lot of interference on the line.

Whilst WiFi is capable of low latency with WMM, the LAN gaming traffic typically ends up being classified under the best-effort class rather than the voice class and so doesn't get any latency help there.

Without expensive test equipment to sniff the RF spectrum or measure the interference on the power lines, all you can do is try the cheapest option hopefully with a good return policy and work your way up more expensive until you get something that works for you.

Chrispy_ wrote:

blah, blah, blah, but Valve! Lord GabeN doesn't lie! ranom lines! spikes!

Once again, I don't know which hardware Valve used and in which particular environment and particular router configurations to collect the data for their pointy lines - I do not perceive ANY type of latency issues (especially spikes) while constantly playing games like BF4, it always stays at 25-100ms average values throughout whole gaming round on all nearest 64-player servers (unfortunately the in-game PerfOverlay does not show or log any detailed network data) regardless of what other people do while connected wirelessly to SAME router:

http://i.imgur.com/YwjkKhe.png

Same thing happens when I just run PingPlotter for a while using BF4 server IP as destination:

http://i.imgur.com/Wk2OFia.png

It can occasionally spike up to 120-150ms between different ISP hops but NEVER to the first hop (which is the router).

This was true for my previous router (Asus RT-N66U, an EXTREMELY common and well-rated model, which died after warranty expired) and is still true for my current Netgear R7000 router.

This is why I still use WiFi for everything even though I can still run wire to every room by myself through all floors (it would just be unnecessarily annoying and a waste of time). The ONLY time I have any latency issues is when the local ISP starts screwing something up again (either routing between their nodes or is doing some equipment maintenance), in which case everything is affected even when using wired connection directly with their supplied router.

So, once again, to answer the ORIGINAL question - there are no issues at all when using WiFi for FPS gaming with the right set of equipment (both good router with stable firmware and a good WiFi PC adapter with "compatible" chipset), regardless of fairy tales some people force themselves to blindly accept Is it cheaper than using powerline adapters or running wire directly from router? Probably not (especially if you can run the wire by yourself), but it's still much, MUCH more convenient and more reliable (compared to powerline adapters only, not to direct wiring)

My advice to original poster - buy several devices (powerline adapters and a couple of popular routers from different brands), test them out in YOUR particular environment during a couple of weeks to see which solution works better with your particular devices then return other devices for full refund (good retail stores like Amazon/Newegg do not charge restocking fees, especially if you pay for "prime/premier" memberships, which you should if you shop at particular store very often) or gift them to a friend/relative or other person who might need them.

Chrispy_ wrote:

If all you can do is discredit Valve then I think that speaks for itself.

It sure does - it speaks that I am a rational individual who likes to believe in his own objective experience instead of being a mindless follower of any particular brand or "consumer product review" website That's why I own smartphones (but do not obviously use them all at the same time) by different brands which run WP8, iOS and Android OSes and why I own both Nvidia's Titan and AMD's 290x graphics cards (as well as other products like TN/IPS/MVA monitors of various sizes and brands).

But seriously... I am going to get the power line network for the following reasons:
1. The router I currently have has the tendency to occasionally drop the signal.
2. One of the issues with powerline adapters is due to the plugs being on separate sides of the breaker box... Mine are not. Plus the big items like the oven, washer, dryer and dish washer are all the opposite side from the room circuits.
3. I can return them and get a wireless solution.
4. The darn things are cheap!
5. I DO NOT NEED SPEEDS GREATER THAN THE INTERNET CONNECTION I CURRENTLY HAVE! Unless the circuits are really bad this should be very doable.
6. If all else fails I still have my 50 foot cable.

With all that being said I find it amusing that people use internet pings to prove/disprove Wi-Fi latency. Anytime you leave the intranet your pings become meaningless. Ping from my router to any given server on the internet and you will likely ping 4-20 ms higher/lower on successive pings. It used to be different a couple of years ago but now it is not very stable when you ping.