Does extending the USB cable add latency for controllers? Or

No, it's a connection

>>376406287
As long as you make sure the cable is straight you should be ok.

In my experience using a wired 360 controller with an extension cord years ago - you should be okay

>>376406350
This. Any kinks in the wire will stop up the connection

>>376406287
as long as your extending it by literally hundreds of thousands of metres you're good

>>376406360
That has like a 6ft cable reach. Why would you need to extend it?

If it is under the max length for a UBS cable (5m) you should be fine. You would add more latency using wireless.

>>376406287
USB has fixed 1ms data transfer intervals, it'll just stop working if the cable is too long.

>>376406287
depends on the quality of the cable and it's resistance

repeaters for digital signals aren't usually required unless you're talking dozens of meters

so unless you're playing on a cinema screen, I wouldn't worry about it

>>376406539
>>376406541
That's good then. I was using a 6ft cable but the connector bent when I dropped the controller. All I have left are 3ft cables but it's too short for my needs. For some reason cables cost a lot to buy on Amazon.

>>376406380
>>376406350

this is PATENTLY FALSE

ANY extension cable serves as a buffer for the initial connection syncronization process. you need at least one, preferably two TIGHT knots to maximize the buffer.

Length is meaningless when it comes to latency. I studied this. You need to make sure the cable is elevated off of the ground at all times to reduce any interference.

>>376406680
Don't forget the loop to add velocity to the particles going like Sonic

so if usb works by moving little electrical impulses

who is the one moving them?

>go to friend's place
>his usb wires are a mess, it goes from his pc and loops around his monitor 3 times before going into his monitor
>sigh and un-spool my 40m quartz-studded usb type a cable
>run it through his fridge and toaster to keep it's shielding active, tie a loop around the toaster's switch to keep it running
>connect the other end to my over-clocked car radio
>he now gets an extra 20 hertz per second on his fps

>>376406487
Because my PC and bed were at other sides of the room and I wanted to play video games in bed

I actually had to remind myself why I needed an extension cable in the first place

>>376406989
Your computer literally does it for free for you. Once it becomes custom for each computer to have their own AI then things'll change.

>>376406989
Laplace's demon.

>>376406989
Gravity. That's why PC on desk fags use wireless controllers

>>376406989
A demon.
Corporate mages summon and trap them inside the devices.
Same thing with rockets. We trap a demon inside them, then cast a spell that ignites the demon.

USB is digital. It either works or doesn't.

>>376408853
>Digital
>When everyone can see it's an analogue cable
top kek

»376406287
Computer scientist here.

Yes, it does, The maximum length of a USB cable (standard) is 5 meters. The reason of this is that after this length, the signal last more than 1.5μs. If a USB host command is unanswered by the USB device within the allowed time, the host considers that there is nothing connected, so it terminates the connection until the device tries to connect to the USB host again, trying and retrying with no answer.

The average delay of a USB cable is 5.2 ns per meter, with a maximum total delay of 26 ns.

Also you have to think about the time that the delay while the device is processing all its logic to answer the USB host command, but the thing is that the longer the cable is the more the signal will last. The average of the USB device processing time is about 400ns. So, in total, the signal lasts 2 · 5.2ns/m + 400ns.

Having a time out of 1500ns, the maximum length of a USB cable could be:
1500 = 2 · 5.2x + 400
1500 = 10.4x + 400
1100 = 10.4x
x = 105m

File: 899-00.jpg (50KB, 1200x900px) Image search: [Google]

50KB, 1200x900px

>>376406287
it's called a USB line

i'll show you later

>>376406287
You're not going to be able to connect enough USBs together to impact the latency, but a wireless connection will be slower because it needs to change the signal into radio waves. Always use wired.

>>376409590
this conversation is the most scientifically retarded thing I've ever heard

>>376410219
yo computerscienceanon

apparently windows 10's 'game mode' somehow decreases button-to-pixel delay by 5ms, any idea how it does that, or why it only does it when it's enabled in the OS but disabled in game?

https://youtu.be/mVNRNOcLUuA?t=597

>>376410307
what was the name of the girl? asking for a friend of mine.

>>376406287
Connect the controllers USB to a splitter and run 2 USBs in parallel to your computer for a faster connection

>>376410675
>he doesn't use quad-channel USB

pleb

tfw wigig is coming soon

tfw wigig will remove wires

tfw wigig will remove connections

tfw wigig will make your computer into a set of parts each with their own separate cooling enclosure and all communicating via wigig at sub-nanosecond latency


future computers will be cool. it always makes me laugh when i see cyberpunk shit with wires everywhere. so uninformed, the authors

>>376410563
The OS does not understand what is receiving from the USB. To make it to understand you need drivers to translate all the raw and analog signals into binary and digital data that the OS can understand, and viceversa in monitors, speakers, etc.

The 'game-mode' is basically lowering the level of understanding of the OS, making it to understand basic data instead of processing it to Application level.

In other words, it simplifies the data processing by making the OS smarter.

When this mode is disabled, the OS uses the default driver to manage all the receiving data.

>>376411009
so when you use game mode, the data given from your mouse is never transformed to standard usb traffic, it's just read straight up?
why would they confine that ability to 'game mode' and not use it for everything?

>>376411161
No, the data that the OS receives is also processed to binary data, but the default driver has to encapsulate it into hundreds of bits of data to make it easy to understand to the OS.

Making the OS to understand raw data also lowers the amount of information that the driver has to add, but also increases the consumption of the OS, because all the services that makes it smarter have to process it. Can you use the game-mode driver in regular use? Of course, but the consumption may increase, and considerably.

>>376411585
>>376411161
I've come up with a way to explain it better. Imagine the default driver as the normie way to express ideas, and the game-mode driver as expressing them by memes.

Making somebody to understand memes can be tedious, and when searching for the pictures, for example, of feelsadman, you can consume more energy than explaining that you are sad, but it's way faster.

>>376406287
Regardless of everyone memeing ITT, if you have a 2M micro usb, even the official one, for your Xbox one controller, adding a 1-2 foot look USB extension caused some weird input affects. Tried multiple extansions and multiple cables too.
This was when trying to make my wife's xbox controller cable longer so we could run it under the rug because I / our guests kept tripping over it.

Wireless controllers are a joke.

>>376406287
It would add to the natural latency caused by the speed of signals traveling through the wire. It's not something you can notice.

>>376412137
elementary my dear watson

>>376406287
Unless your cable is 50 feet fucking long you're fine OP.

Cable length across like, a room wouldn't add a perceptible amount of delay at all.

>>376408228
wow I always knew we were living in a discworld

>>376406287
Yes it increases latency, but not by a perceptible amount.

>>376410923
Yeah, no. There's a reason why a humble Ethernet cord is still considered superior even to those fancy 802.11ac setups with theoretical link speeds in the thousands of Mbit/s, which on paper is far beyond what GbE can achieve.

>>376412740
>There's a reason why a humble Ethernet cord is still considered superior even to those fancy 802.11ac setups with theoretical link speeds in the thousands of Mbit/s, which on paper is far beyond what GbE can achieve.

wigig exists right now, in non-theoretical form, even for consumers. you need a $500 router and private chinese network cards to use it, but it exists. it has higher tested bandwidth than ethernet and negligibly higher latency, like a nanosecond or two higher latency

like i said, it's not for internet, because of the range, it's for replacing wires, which it definitely will do within the decade (or another high-frequency standard like it)

>>376406350
>>376406467
Don't forget that all the 1s are more likely to get stuck around bends compared to 0s

>>376412928
wtf is wigig

>>376414026
Only when your mouse doesn't have an integrated binary rearrangor that combines two 1s into an aerodynamic arrow