View Full Version : [Devblog] Fine-tuning Aiming and Controls

July 25 2013, 06:50:08 PM
The feel of aiming is a complex area and there are many things in a game that have an impact on it. The feedback we have received in this area since May has been very useful but also very varied. In some cases this feedback was conflicting and in others it was just plain confusing. We have conducted an audit of our aiming systems and identified a number of areas for improvement as well as finding changes/problems outside of the control system itself that are affecting the experience. In this blog I'm going to talk about some of the problems we found, what we're doing about them and hopefully answer some of the other questions along the way.

Performance has a significant impact on how the aiming and controls feel and no amount of tuning and tweaking of the controls themselves can change it. Poor frame rates can increase input latency. Unstable frame rates can make the control sensitivity feel inconsistent. Performance is an area we have had some problems with and is one that the whole team has been working very hard to improve. This has been a major cause for a lot of the problems we've experienced, and the positive feedback we've had since 1.2 was released with performance improvements bears that out. This is something we remain focused on improving further.

Client/server de-syncing and projectile weapons
In 1.2 we had a couple of fixes that improved projectile weapons. First we found that the way we check for splash damage was prone to "missing" partially occluded targets. We have now improved the system to solve this problem. We also found issues with projectiles and client/server de-sync. In poor network conditions the projectile can "miss the mark", making the weapon feel unreliable. We have made improvements to this as well and judging by resurgence of the Mass Driver and popularity of the Flaylock this seems to have worked (perhaps a little too well in the case of the Flaylock, but we're dealing with that separately: see this dev blog).

Controller & mouse
There's been a lot of discussion surrounding controller performance and what needs to be done to improve it. With the 1.4 release the core control settings for speed and sensitivity will be updated and improved, and it should feel tighter. The biggest update, however, is that we plan to reintroduce aim assists that we previously disabled. Don't sneer! Aim assists are part and parcel of aiming control on console!
The assists in question are magnetism and adhesion. Magnetism works by helping you maintain your aim as you strafe a target. Adhesion adjusts your rotation when an enemy is targeted to help you stay on target. Why did we disable them in the first place? In testing, we found that in a limited number of circumstances they could interfere with aiming instead of assisting it. We made the decision to disable it until we could iron out the kinks in the system. Since it was disabled, there has been feedback that staying on target, particularly at closer ranges has become harder. We believe this is in large part due to magnetism and adhesion being disabled. So we overhauled the system, and so far internal playtests have been extremely promising. The improvement in accuracy is significant.
http://failheap-challenge.com//web.ccpgamescdn.com/dust/news.control/64950/1/Amarr_Assault.JPG (http://failheap-challenge.com//web.ccpgamescdn.com/dust/news.control/64950/1/Amarr_Assault.JPG)
We have also been asked if we can add an option to disable acceleration. First I would like to explain quickly how acceleration currently works on the controller. What we have in DUST 514 is best described as "edge acceleration". What this means is that at 95% stick input (that is, when the thumbstick is 95% of the maximum distance away from the center position) your aim will start to accelerate to an extreme speed; before that point speed scales linearly to a set point. This is to allow for finer aiming at low input levels without sacrificing the ability to turn at higher speeds. The benefit we have found with the linear "curve" is its predictability when ramping up/down makes tracking speed changes easier to judge. This is pretty important in a game like DUST where the high HP model requires you to better maintain aim on moving targets if you want to score a kill. We have been experimenting with different curve setups but so far have not found another one we particularly like.
So, the answer is we could disable it, but you probably wouldn't like it. If we were to just turn it off you would be left with a very low maximum rotation speed compared to those using acceleration. Controls would feel pretty sluggish unless you jacked the sensitivity very high at which point you would lose a lot of low level control. So really it's less a case of turning it off and more a case of creating and tuning a separate aiming scheme that is fully reliant on a curve. It's possible we could do this at some point, but right now we believe it's better to focus on just one set up for the controller.
That was the controller; how about the mouse? Well, the mouse had one big problem: speed. It was suffering from having its speed capped quite low which made it feel very flat. In 1.4 that cap has been removed. Its sensitivity levels in and out of scope have also be re-tuned to provide more responsive, finer control. The fine aiming "wobble" that has been in previous builds should now be a thing of the past. As a result the mouse is feeling much better and we're continuing to work on it during this release.
That's it for this quick update. The change really is significant so it's very important for us to get it in to the wild and get feedback on it. 1.4 can't come soon enough!
CCP Wolfman

original link (http://dust514.com/news/blog/2013/07/fine-tuning-aiming-and-controls/)

July 26 2013, 01:30:06 AM
Oh crap there goes my mouse muscle memory again.