|
I've published a map that measures the ingame ping.
It started a few hours ago to test a way to measure the ingame ping and someone in TL's irc told me that it might be useful for players to test.
Screenshot:
![[image loading]](http://i.snag.gy/ixBDC.jpg)
My values for the Korean server.
Features:
- real time latency
- game time latency
- skipped game frame amount
- multiplayer support
Battle,net data:
- AM, EU, SEA, KR
- map name: "Ping Test"
Map Links
Americas
Europe
Korea & Taiwan
Southeast Asia
How does it work?
1. When you move your mouse, it starts a timer for each tested aspect. Also, it's making a huge, fully transparent image react to your mouse.
2. Another trigger detects that you started to hover over that image and pulls the displayed values.
Some notes:
- If you spam mouse movements, it might increase the latency as the "mouse moved" event might cause lags. But more tests are required regarding that.
- I used the mouse move event as it makes sure you are focussed on the screen with the mouse. The system doesn't work when your mouse cannot touch that transparent image. I might use something else in future.
The map can have multiple players, so you can even test the measured latency when you are playing together.
Hopefully that is useful for someone.
|
Netherlands110 Posts
|
Excellent, I have been waiting ages to test out the actual latency in-game and not just some theorised crap. Thanks a tonne!
|
I get the same score on both EU and NA (playing from Finland)... I can't imagine that's correct.
kr at least show a significant difference
|
Maybe you get the same ping on NA and EU because of the build in delay? I think it makes you have the same ping on every server unless you play on KR/SEA from EU.
The tool may be useful with all the network spikes problems there are these days though, it should show on the real time delay, and confronting it with a pingtest I guess you could determine it's not coming from your internet..
|
My real time delay was 177 - 222 on NA, 300 - 400 on EU, 480 - 600 on KR and 550 - 600 on SEA. Is it that bad?
|
first of all really nice to have this, ty!
I get 177 or 222 yet nothing else. it's rather weird to have just those 2 fixed values, isn't it?
|
curious as to how this works, as others have mentioned getting specific values that alternate... ranging frmo ~3.1 to 4.5 on KR
|
Australian here
_______delay(realtime)_____frames
America ___.31--.35 _________8
EU________.44--.48 ________12
KR _______.35--.57 _________11
SEA ______.31--.35 _________8
Heres my results, Kr was the most unstable for me, with pretty erratic spikes
|
|
|
|
|
I see. I was confusing it with the actual net ping. Ty, still very nice thing to have 
|
Might as well test when i'm there.
From doha (Qatar) :
NA_Ping : 0.2666_Frames : 6
EU_Ping : 0.1777_Frames : 5
KR_Ping : 0.4443_Frames : 10
SEA_Ping : 0.8442_Frames : 12 (unstable)
|
How do I read the results? I have 0.35 Delay (real time) and 0.5 skipped game time. How does that translate to ping?
|
On September 28 2013 07:53 Garnet wrote:
How do I read the results? I have 0.35 Delay (real time) and 0.5 skipped game time. How does that translate to ping?
0.35 Delay would mean 350 ms ping
|
On September 28 2013 09:17 lysergic wrote:Show nested quote +On September 28 2013 07:53 Garnet wrote:
How do I read the results? I have 0.35 Delay (real time) and 0.5 skipped game time. How does that translate to ping? 0.35 Delay would mean 350 ms ping
That's surprising. When I do a tracert to the server's IP it's only 150ms.
|
United Kingdom20326 Posts
we are playing with faster which is 1.4x the normal speed.
It's 1.38*, not 1.4 mate 
I've got 5-6 frames on EU consistent, which is pretty nice i think
|
|
|
On October 06 2013 00:49 Cyro wrote:It's 1.38*, not 1.4 mate I've got 5-6 frames on EU consistent, which is pretty nice i think
It's 1.4, if you only consider what the game runs in locally. It's the value the game would reach in a perfect environment. 1.38 was an approximate value measured ages ago from times a human stopped himself when the game. I don't think that value is still accurate.
Ingame, on battle.net, it actually measures 1.4065 as a factor, if you start two timers for game and real time and keep displaying the calculated factor between the two elapsed times.
5-6 frames delay seems to be a good connection. So far, I don't know if anyone has better values than constant 5 game frames.
edit:
Also, the most important value is the amount of skipped game frames or the game time. Since the game progresses in intervals, the input just needs to arrive before the next game frame execution.
|
United Kingdom20326 Posts
Really? Everyone's been saying 1.38 for four years! D:
|
Imho this map doesn't work. I got 0.1777 too, and stable, except i move my screen, then it jumps to 0.2666, if i stop moving it's back to 0.1777. Doesn't make any sense
|
On October 06 2013 02:13 Chemist wrote:
Imho this map doesn't work. I got 0.1777 too, and stable, except i move my screen, then it jumps to 0.2666, if i stop moving it's back to 0.1777. Doesn't make any sense
Moving the screen is input, too. Maybe that amount of additional input has an impact on the delay with your connection. It has no impact on the connection I'm currently using.
Also, this map is just measuring. The only time it breaks is when you move your mouse on the menu buttons or tab out the game making your mouse not being able to fire the trigger to stop the measurement.
If you keep moving the mouse it can actually increase, too. The event used for that can cause that problem. For example, I can make my connection to KR worse when I just keep moving the mouse. But I will update the map in 1-2 weeks when I get back to my appartment to not use the "Mouse moved" event and instead to measure the latency immediatly after receiving the mouse reaction.
|
I've updated the map to automatically start the tests after a small random delay. I've fixed the game loop delay value, too. It was displayed one too high.
The connection seems to appear to be more unstable on my home region with automatically repeating tests, but I can't freak out my connection to KR anymore because of the mouse move event. I don't know why it's unstable now, but I assume it's not battle.net having a bad day. It might have something to do with the buffers the bandwith demanding trigger events use.
|
|
|
|
|
|
EPT
