Negative mouse acceleration, a bug or it works as intended_
#1
Posted September 04 2013 - 10:42 AM
I thought that it's a bug, but then I found some threads where it was stated that this mouse behaviour is intentional.
If that is true, then there is no choice for me but to uninstall this game and forget about it, because mouse behaviour in Hawken is absolutely the worst I have ever seen in my life and it makes the game unplayable.
#2
Posted September 04 2013 - 10:49 AM
My only real suggestion is to make sure that you turn off mouse smoothing in the settings. That can cause issues with mouse lag.
#3
Posted September 04 2013 - 11:35 AM
#4
Posted September 04 2013 - 11:51 AM
There have been many people who thought that mouse acceleration was present in Hawken, and after tons and tons of testing, it isn't there. A developer who hates it himself went ahead and looked through the code for us once, too, and confirmed there was no code for mouse acceleration in the game.
If you want to see that turn rate cap indicator, by the way, you should get on that, because I think the new hud in the upcoming release will no longer have it.
Ah ha! Found the post I was referring to.
[ADH]Deuy, on December 21 2012 - 04:25 AM, said:
A lot of people mistake this for negative mouse acceleration. In fact, when you flick your mouse to the side quickly you are indeed turning at the max speed cap, however you are only turning for a short moment because your mouse move was so fast (it ended quickly). When you move the mouse slower you can still turn at that same max speed cap, but you will be able to continue turning much farther because you're moving the mouse for a longer period of time.
You aren't being punished for moving your mouse faster, the mechs simply can't turn that fast and will stop turning when your mouse isn't moving. We tested having the mech "remember" your mouse movement and attempt to "catch up" to that motion, but this made the controls feel laggy and unresponsive. It was also a bit nauseating for a lot of people.
P.S.- I personally asked an engineer to check on this based off the feedback from older threads. He looked into the code extensively and the info above is based off his testing. The speed cap is hit right away with quick mouse movements and then drops off instantly when the mouse movement ends. If you try moving your mouse very fast for a longer period of time (like dragging it quickly across the whole length of your desk) it's very difficult to not mix in up and down aim movement rather than just turning. I think this gives the impression that you just can't turn as fast as moving the mouse slowly.
Edited by Saturnine, September 04 2013 - 12:15 PM.
digitalhughes: @s9 you stinkn' weasel!
Don't know which mech to pilot_ Check out my guide to the mechs!
Want to hang out with some awesome people, including ReEvolve, Ashfire908, Rei, ThirdEyE, and even the world-famous Loc_Tr_ Come to the #hawkenscrim IRC!
#5 Guest_f_error_*
Posted September 04 2013 - 12:12 PM
When you reach the maximum you get
(maxRot/time) / (mouseEvt/time) wich evaluates to
maxRot / mouseEvt.
This means once you have reached the cap, you will get a smaller and smaller reaction ratio, the more events you deliver = the faster you move the less you get from it.
Many people seem to expect something like a non-uniform sensitivity. Like they could turn the sensitivity very high and then would be very "fast" while doing small movements and getting slower with higher turn angles. That is not correct. The turn cap is active all the time. Your mech does NEVER turn faster than allowed (aside from maybe special effects outside your control. maybe collision reaction etc., bugs, aimbots ...)
What you feel there is actually your own hand acceleration. If you have a low acceleration you will feel like having a big "high sensitivity angle". If you have a high acceleration, you reach maxRot earlier and hence earlier feel like your mech is sticking to some very strong chewing gum. This effect also happens when you work with slow movements and small rotation angles e.g. while sniping long distance. You dont hit the turn cap there and feel "free" therefore.You can apply your normal accustomed slow movements and because they dont hit the cap, the reaction is as expected and does not make you feel limited.
Edit: Thats Saturnine for you. Fast as ever.
Edited by f_error, September 04 2013 - 12:14 PM.
#6
Posted September 04 2013 - 12:35 PM
because bow ties are cool
#7
Posted September 04 2013 - 12:55 PM
#8 Guest_f_error_*
Posted September 04 2013 - 01:31 PM
sorry to say this, but implementing a speed cap IS "negative acceleration".
An object moving with constant speed has zero acceleration.
Hence if you have a (constant) acceleration on your mouse, but the cursor/mech/etc moves at constant speed, your acceleration gets countered. And to bring a positive acceleration to zero means, that a negative acceleration of the same magnitude is added. This is effectively what a maximum turn speed does. But its not a negative acceleration formula or anything.
The problem is all this pseudoscientific and religious mumbo jumbo that gamers carry around and only back up with the
"move your mouse across the pad slow and back fast and see if you are back at your starting point" test.
Its a sport to bash games for "negative mouse acceleration" like bashing it for being "pay2win" while there is no clear definition or formula or referential implementation for both.
"Positive" mouse acceleration is the desktop feature that helps you relax your arm. You measure the incoming rate of mouse events and apply a speedup factor to the reaction. E.g. 100 "left events" result in 100px movement to the left. Problem is, you have a HD-screen and want to go over to the other side. That is a lot of pixels and a lot of mouse travel you need there.
To move that distance "fast" and without acceleration you have to increase the movement factor by either sending more mouse events per distance (dpi) or by moving more pixels per event. The problem is, that a high factor like 500px/cm will get you across the HD resolution in about 4cm, but will it make really hard to hit a 64x64px icon (1,3mm mouse movement).
Try doing such things with a trackpad where you dont have the physical travel distance ...
In these cases mouse acceleration is a good idea and was therefore implemented. If the mouse events come at a very high speed, then the user does not care for very precise movements and instead just wants to "get over here" fast. So you measure the speed of mouse events and increase the pixel/event ratio. (I do not know if current implementations go even further and use acceleration instead of speed factors. I would guess so. Even more complicated)
Now negative mouse acceleration would be, that you get a higher result (more pixels, more angle) by slower movement. If that was applied consistently, your cursor would fly all over the screen when you come back from the toilet. So IF this is implemented, it needs a lot of work with thresholds etc. It would also mean that if you move your mouse incredibly fast, the cursor would not move at all. At some point the limited precision of floating point numbers would make that very small factor go to zero.
The typical gamer test, is too simple. It is clear that games in one way or another have a limit on mouse events. They can handle only so much events per time. Usually events get queued and stored, since they come in too fast for a full evaluation. Modern games run different parts of their calculations at different intervals. Game logic and reaction to events can/needs to be processed faster than rendering, AI can be evaluated slower ...
To make the test better, you need to test multiple settings. First you need to check your mouse and see if it really keeps delivering all those events. A simple mouse-event counter is enough. Move the mouse along the same distance with different speeds and check if the event counter is always the same.
When you know that your mouse is really delivering the events and they really get to the software, you can test your game. Set the mouse to highest sensitivity and set the game so that your defined distance equals the defined rotation. (180, 360 etc). Make your tests at different speeds. If at higher speeds the angle is reduced, then events get lost/thrown away.
Change to lowest sensitivity on your mouse and adjust the game so that the distance again results in the same rotation. Make your tests again. Now your mouse delivers a lot less events for the same distance and the same movement speeds. If the game is choking, then it should deliver correct rotation at higher speeds with this configuration.
Problem is, that you now have the lowest physical resolution and highest ingame factor on rotation. So you raise the mouse sensitivity step by step to find the highest value that your game can handle without loosing events at maximum hand speed. From that point onward, you have to do the rest ingame if you dont want to get the "distance is not the same" effect.
Why do i write all this stuff ...
Edited by f_error, September 04 2013 - 01:38 PM.
#9
Posted September 04 2013 - 01:51 PM
I also found post quoted by Saturnine too. So it seems it works as intended indeed.
Quote
Quote
*Sigh* I think I can do "optimal 90 degree turn" macroses in AHK to circumvent this [censored] "feature".
Thank you all again for your willingness to help.
P.S. And I beg your pardon for my imperfect English
#10
Posted September 04 2013 - 02:40 PM
#11
Posted September 05 2013 - 06:47 AM
#12 Guest_f_error_*
#13
Posted September 05 2013 - 10:43 AM
Indrik_StompHowler, on September 05 2013 - 06:47 AM, said:
Yeah, what_ Just because there is an issue or something prompting a suggestion, it doesn't mean one should personally attack one another, in this case, developers.
Why do some people feel the need to draw the line for players VS developers_ The devs are there to create a better final game for us to play, and in turn, we're in beta giving constructive criticism_ If you're only looking for some free gaming hours and attack those who are giving you said game, you can kindly GTFO.
EDIT: I do apologise for my tone. But I wasn't referring to you as a user/player, or the fact you started this thread. I was directing at the quoted post and the post itself only.
Edited by neo0031, September 05 2013 - 10:45 AM.
...Nicely and politely.
#14 Guest_f_error_*
Posted September 05 2013 - 11:54 AM
The issue is that some people seem to believe their personal views on FPS games are supposed to carry over to any other game, for example hawken. Just because all those "pro gamers" talk ratsh_it about negative acceleration without being able to deliver a single physical formula, doesnt mean a speed cap is an "issue". It is a good decision and it is there to stay.
Now if that is a deal breaker for anyone, by all means dont be stupid and quit! Dont waste your time on a game, that is not aimed at you or your specific needs. Play other games that deliver what you ask for. If they dont exist, do something better with your life.
If there is nothing better for you and you absolutely need that one game, nobody is able or willing to develop ...
then use your pro gamer wisdom and get started with developing yourself.
1 user(s) are reading this topic
0 members, 1 guests, 0 anonymous users