[PROTOTYPE] QT hybrid movement mutator

[Harry Spanner]

QT is a C++ mutator that combines movement elements from Quake and UT within the official moveset defined for UT4. Specifically, it adapts strafing from Quake III, borrows the basic dodge from UT2004, and takes everything else from UT4 with a few tweaks to things like wall running.

In QT you can continue sprinting after you jump. You could view this as a toned down version of bunny hopping from Quake, though pursuing that comparison was not an explicit objective.

The main goal of QT is to strengthen strafe and jump and sprint relative to dodge so players aren't so reliant on dodge both in combat and for map travel.


Installation

1. Download QT here.
2. Move BP_Mutator_QT-WindowsNoEditor.pak to Epic Games\UnrealTournamentDev\UnrealTournament\Content\Paks. If there is no Paks folder, create it.
3. Move the Mutator_QT folder to Epic Games\UnrealTournamentDev\UnrealTournament\Plugins.

To play:

1. Open UT from the launcher. Go to Play --> Create a Game --> Create your own custom match --> Custom.
2. Move the QT mutator to the list on the right then click Play.

Note, if the game freezes upon launch, it means the mutator is not up to date. You will have to delete QT and wait for me to update the mutator for the newest release build.

LAST UPDATED FOR RELEASE BUILD: 2 FEBRUARY 2016

To run QT on a server:

1. Copy the Mutator_QT folder to <SERVER FOLDER>\UnrealTournament\Plugins.
2. Start the server through a command line from your <SERVER FOLDER> directory and include the argument Mutator=Mutator_QT

Example command: UnrealTournamentServer UnrealTournament DM-DeckTest?TimeLimit=10?Game=DM?Mutator=Mutator_QT


Strafe

Run speed in QT is about 6% faster than in UT4 and about 9% slower than in Quake III. Acceleration is about 7% faster than in UT4.

The values for ground movement in QT are derived from Stolid's excellent UT4 Strafing breakdown. Here, QT approximates Quake III strafing with max run speed adjusted downward:



Here are the ground movement values for comparison:

Value	UT4	QT
Run Speed	940	1000
Crouch Speed	315	500
Acceleration	6000	6400
Deceleration	500	1250
Ground Friction	12	8
Braking Friction	5	2

That crouch speed is half of run speed borrows from Quake Live, where walk speed is half of run speed.


Dodge

Long story short, QT adapts the values from Trev's UT2004 Port based on Sir Brizz's Prototyper to create the basic UT2004 dodge arc under the condition of higher UT4 gravity. Due to this disparity, the QT dodge is 6% faster.

For the long story, read on.

 

Spoiler

To achieve a specific dodge arc, you need only know the height and length of the dodge. To find the height and length, you start with the vertical and horizontal impulse values and gravity.

Based on Trev's port, these are the relevant UT2004 values:

Dodge Impulse Vertical (DIV) = 420
Dodge Impulse Horizontal (DIH) = 1320
Gravity Scale (GS) = 0.882079

First, find the time (t) it takes to reach the peak of the dodge, when vertical velocity is 0.

v = v₀ + a*t

Here, acceleration is simply gravity, so it is negative. The value for acceleration from gravity, as found by Stolid in the ini, is 2154. Here we're dealing with lighter UT2004 gravity, so we must furthermore take the gravity scale into account.

0 = DIV - GS*g*t

0 = 420 - 0.882079*2154*t

t = 0.22105284 ≈ 0.22s

Now it's possible to find dodge height (DH) with the equation of motion for displacement over time.

x = x₀ + v₀*t + 1/2*a*t²

DH = 0 + DIV*t - 1/2*GH*g*t²

DH = 420*0.22105284 - 1/2*0.882079*2154*0.22105284²

DH = 46.42109744 ≈ 46

Next, find the total dodge time (DT) by doubling the time it takes to reach the peak of the dodge.

DT = t*2

DT = 0.22105284*2

DT = 0.44210569 ≈ 0.44

Use the horizontal impulse and the total dodge time to find the dodge length (DL). There is no acceleration, so this is relatively simple.

x = x₀ + v₀*t + 1/2*a*t²

DL = 0 + DIH*DT + 0

DL = 1320*0.44210569

DL = 583.57951068 ≈ 584

---

Now find the vertical and horizontal impulses required to recreate the UT2004 dodge arc under the condition of higher UT4 gravity, where the gravity scale is simply 1. To find the vertical impulse (DIV), you can use the equation of motion for velocity given displacement, which is useful because we don't know the value for time yet.

v² = v₀² + 2*a*(x − x₀)

0 = DIV² - 2*g*(DH - 0)

0 = DIV² - 2*2154*46.42109744

DIV = 447.19356857 ≈ 447

Now let's save some steps by inserting the total dodge time (DT) into the equation for finding the time it takes to reach the peak of the dodge, which we know is half of total dodge time.

v = v₀ + a*t

0 = DIV - g*(DT/2)

0 = 447.19356857 - 2154*(DT/2)

DT = 0.41522151 ≈ 0.42

Finally, solve for the horizontal impulse (DIH) needed to travel the dodge length in this time.

x = x₀ + v₀*t + 1/2*a*t²

DL = 0 + DIH*DT + 0

583.57951068 = DIH*0.41522151

DIH = 1405.46550123 ≈ 1405

---

In summary:

Dodge Value	UT2004	QT
Impulse Vertical	420	447
Impulse Horizontal	1320	1405
Height	46	46
Length	584	584
Time	0.44s	0.42s

Note that both dodge arcs have the same dimensions but that because of the difference in gravity, the QT dodge takes 6% less time to complete.

Here's how the QT dodge stacks up against the UT4 dodge:

Dodge Value	UT4	QT
Height	59	46
Length	658	584
Horizontal Speed	1410	1405
Time	0.47s	0.42s
Height : Length Ratio	1 : 11.2	1 : 12.6
% of Jump Height	47.4%	37.5%

The QT dodge takes slightly less time, has a slightly flatter arc, and is farther differentiated from jump in terms of height. This means there are more situations in which jump is the better choice for clearing obstacles and navigating the map.

UT4 run speed is 66% of dodge speed, while QT run speed is 71% of dodge speed. Again, this increases the appeal of strafe relative to dodge in combat.


Sprint

At 1370, sprint speed in QT is 98% of dodge speed, which means sprint covers long distances significantly faster than dodge spam.

Moreover, you can continue sprinting after a jump, allowing you to surmount many obstacles and sprint for much longer distances. A jump at max sprint speed covers 59% more horizontal distance than a dodge in QT.

This change to sprint required that QT be built with C++, and all credit goes to RattleSN4K3 for designing the mutator. Without his incredible instructions, I'd still slamming my head against broken builds in Visual Studio.

A few points about sprint:

• You cannot continue sprinting after a dodge.

• You can continue sprinting after you run over a ledge.

• For sprint to continue when you land, your horizontal speed must be greater than max run speed.

• This means if you slow down too much in the air, you'll have to wait for sprint to reactivate the normal way.

• Sprint activates in 1 second, half the time it takes in UT4.

After sprint activates, it takes 2.0s to reach max speed compared to 0.97s in UT4.



In QT there is no immediate loss of speed when you land a jump during sprint.



This accounts for a huge increase in distance traveled by sprint compared to UT4.


Wall Run

You can wall run only when your vertical velocity is between a certain pair of upper and lower bounds.

Bound	UT4	QT
Upper	650	365
Lower	-180	-730

Check the spoiler for a detailed walk through of how jump interacts with the wall run velocity bounds.

 

Spoiler

The instant you jump in UT4, you are imparted with a vertical velocity of 730. As you travel upward, gravity steadily slows you down, and in the mere 0.04s it takes to reach one fifth of peak jump height, your vertical velocity has already dropped to 650, the upper bound for wall run. From this moment forward, it is possible for you to begin wall running, but for now let's assume you choose to continue rising through your jump.

At 0.34s you reach peak jump height (124) and zero vertical velocity, then you begin to descend and your velocity becomes negative. For now you are moving very slowly, and by 0.42s you have dipped only six percent below peak jump height to -180 velocity, the lower bound for wall run. Now it's too late to start wall running. You continue to drop and gain negative velocity, reaching -730 vertical velocity the instant before you land.

Wall run engages almost instantly after you jump in UT4, while if you are slightly beyond the peak of a jump, it is no longer possible to begin wall running.

In QT the upper bound is exactly half of jump impulse velocity. You slow to this speed when your elevation is exactly three quarters of peak jump height, and you reach the lower bound just as you land.

UT4 jump	QT jump

The UT4 dodge is not subject to the upper bound because its vertical impulse is below the velocity threshold to begin with. You reach the lower bound after dropping one tenth below peak dodge height.

In QT you reach the upper bound at one third of peak dodge height, but you do not reach the lower bound by the time you land. In fact, you can dodge off a ledge, dip below your starting elevation, and drop an additional 167% of dodge height before you reach the lower bound.

UT4 dodge	QT dodge

At one fourth of normal gravity, QT wall run gravity is 56% stronger than UT4 wall run gravity. You drop faster during wall run, but it still takes longer to reach the lower velocity bound.

Overall, QT greatly expands the range of application for wall run.

[Cunni]

The mutator does not appear in the list when creating a custom game :s
I've checked where the files were extracted and it seems ok

im looking forward to test it

[Stolid]

Good job getting this all done. Since the movement mutator isn't updated anymore it's quite a lot of trouble to do anything with movement. If you want to do anything more than the mutator allows for like you wanted to do you are forced to use a C++ mutator which makes it a real pain.

I don't know if this mutator is meant to be a suggestion on how to change the standard movement in UT4 or just meant to be an alternative mutator people can use if they want to.
I will give my views as far as this replacing UT4 movement. This is after testing this mutator offline and considering UT gameplay historically.

Strafing
I don't feel a huge need to up the default walk/run speed. It's not something that I found very obvious when playing either, so in that sense I'm not fully against it either.
Acceleration is actually much the same, I don't notice a huge increase, but that might be because the max speed is higher so the time it takes to reach full speed isn't very different.
The thing I definitely did notice is the change in ADAD movement. This is quite reliant on friction, which determines the FS2FS line. It's not that weird I noticed that seeing as the friction is much lower. The FS2FS is even slower than in QL with your value of 6. A friction value of 8 would match the graphs up better than they do now, but personally I'd like it to be closer to 10. For reference the current value is 12.

Dodge
I can be pretty short about this, I like it. My own movement prototype had very similar values: a DIV of 465 instead of your 447.
There is one problem however, the devs don't want to lower their DIV value any lower than it is now AFAIK. From what I understand it's because of collision issues, but they haven't really said anything about it in detail. I don't know if there's a chance this issue could ever be solved, assuming this is really a thing in the first place.

Sprint
In the current game sprint is almost useless, it needs a big change or it should just be removed IMO.
As far as your changes go I don't think you've even gone far enough. I'd like it to speed up even sooner, speed up even faster and go to an even higher speed than both current UT4 and your mutator do. This is the only way I can see sprint becoming viable.

Wall Run
Wall run is another thing that just isn't useful enough to have in the game at the moment. I like your changes, but for this too I doubt they go far enough. I'm not quite sure what I would personally change though.


As far as your main goal of strengtening strafing and not having combat rely on dodging as much, I actually think the opposite is needed at the moment in UT4. Strafing is probably more powerful than every other UT and dodge is probably the least powerful it has ever been. A lot of combat is happening with strafing only right now, which in previous UT games was rare.
Now if you just want to have a more Quake like experience, involving less dodging in general, then that's fair. For the default game I don't see that as the way to go though.

[Infiadux]

Very nice mutator, looks like you put a lot of work into this. I ran around with it a bit offline, and the most noticeable things are the shorter dodge and faster run speed. I like the faster run speed but I do think your idea of making strafing more powerful isn't the right direction to head in.. Stolid took the words out of my mouth:

As far as your main goal of strengtening strafing and not having combat rely on dodging as much, I actually think the opposite is needed at the moment in UT4. Strafing is probably more powerful than every other UT and dodge is probably the least powerful it has ever been. A lot of combat is happening with strafing only right now, which in previous UT games was rare.
Now if you just want to have a more Quake like experience, involving less dodging in general, then that's fair. For the default game I don't see that as the way to go though.

I like the shorter dodge though, it always felt to me that it was too long in UT4. Still, it's the same speed as the UT4 dodge which is just too slow, especially with the improved strafing. Also like the smoother feeling slide, I guess that's due to the lower friction.

[Stolid]

Current strafe speed is actually very close to UT99 strafe speed (depending what scale factor you like to use) and the dodge speed is exactly 150% of the strafe speed in both games. What makes it feel slow apart from taking 100ms longer because of the arc is probably it being slow relative to most projectiles. The reduction in speed of those is mainly what I like to see happen, though their experienced speed is also affected by the ping compensation.

[Infiadux]

Current strafe speed is actually very close to UT99 strafe speed (depending what scale factor you like to use) and the dodge speed is exactly 150% of the strafe speed in both games. What makes it feel slow apart from taking 100ms longer because of the arc is probably it being slow relative to most projectiles. The reduction in speed of those is mainly what I like to see happen, though their experienced speed is also affected by the ping compensation.

Hmm, I'm pretty sure UT99 has closer run speed to the 1050 used in QT rather than the default 940. Or, when you stay strafe speed, do you mean the acceleration and effectiveness of ADAD spam? I do agree I don't like the arc of the dodge in UT4. In UT99, it was a "sharper" arc and it allowed the dodge to be faster without making it longer. Also threw off people using hitscan so you could more easily get in close range of a sniper.

From what I played around with in the Editor, 10% faster horizontal dodge impulse and 10% stronger gravity feels closest to UT99. But again the arc is really smooth which might be making it seem slower than it really is. And I found no way to change the arc other than going into the C++, which I have no idea how to do :/

[Stolid]

Most things in the game seem to be scaled by 2.2 or less. A run speed of 1050 would give a scaling of around 2.4.
The arc is simply a result of the angle, speed and gravity. You don't have to edit anything as those are the only things involved in any UT game and they give the same arc if they match. That is as far as physics goes, the eye height during the dodge and the player animation is different though.

[Infiadux]

Most things in the game seem to be scaled by 2.2 or less. A run speed of 1050 would give a scaling of around 2.4.
The arc is simply a result of the angle, speed and gravity. You don't have to edit anything as those are the only things involved in any UT game and they give the same arc if they match. That is as far as physics goes, the eye height during the dodge and the player animation is different though.

I don't think UT99 dodge arc was calculated the same way as the UT4 one though.. either the gravity there works differently somehow or dodge actually has an artificial height limit that kills off the vertical impulse before it fades out by itself.

[Stolid]

I would be surprised. You'd have to put in an automatic sort of dodgejump, only instead an impulse going straight up it would have to go straight down at the apex. AFAIK The source code to describe the dodge in UT99 isn't available to check this, but it seems unlikely.

[RPGWiZ4RD]

Agree with Stolid and Barktooth here regarding strafing vs dodging. Personally I think strafing doesn't need any buff anymore whatsoever, it's already really powerful in this UT4. I kinda wished like Barktooth to give dodge impulse a slight boost and gravity increase to keep the dodge distance the same as it is. 10~15% or so increase on both would sound about right.

Personally I'd possibly even want to decrease acceleration and friction a tiny bit (max 10-20% or so, nothing major) so the strafing wouldn't feel so jerky and lead to a smoother transition from standstill to running. I think this is where movement differs the most to what people are used to in other games, the instant directional change will probably make it seem very fast compared to other games.

[Infiadux]

I would be surprised. You'd have to put in an automatic sort of dodgejump, only instead an impulse going straight up it would have to go straight down at the apex. AFAIK The source code to describe the dodge in UT99 isn't available to check this, but it seems unlikely.

Oh yeah, I think that describes it even better. A bit after the apex of the dodge, there seems to be a downward impulse applied in UT99. You can see it fairly clearly if you record a video in slomo 0.1 and watch it back frame by frame. It looks like after the dodge reaches its apex, it sort of evens out and you just barely start falling, at which point the speed with which you fall is suddenly increased.

Actually, an even easier way to show it is to use the Pulse secondary beam to draw on the wall.





As you can see, the UT4 beam trail is smooth, while the UT99 one has a bit of a hitch in it. Strangely, it looks like there isn't just downward impulse applied, but also more horizontal impulse right after that - as indicated by the Pulse beam angling down pretty sharply after the hitch but then straightening back out afterwards.

[Laambo]

To run QT on a server:

1. Copy the Mutator_QT folder to <SERVER FOLDER>\UnrealTournament\Plugins.
2. Start the server through a command line from your <SERVER FOLDER> directory and include the argument: ?Mutator=Mutator_MovementPrototyper

Example command: UnrealTournamentServer UnrealTournament DM-DeckTest?TimeLimit=10?Game=DM?Mutator=Mutator_QT

... you are referencing brizz's prototyper on line 2? ... is that right?

[Stolid]

Oh I know what the UT99 dodge is like, I've done slomo and pulse/link secondary tests and comparisons myself in the past.

The question is whether that dip is real, meaning that the player collision hitbox physically goes down at that point, or that it's a view effect much like the landing view shake effect you get when you land after a dodge. You can see that happen both in the video as well as the pulse decal drawing on the wall angling up sharply at the end. The view changes, but the hitbox was already on the ground and remained there during that entire effect and also remained the same size.
I would put my money on it being similar to the landing view shake effect.


Having said all that, it might be an interesting thing to try out in UT4 even if it doesn't actually happen in UT99 as a way to bypass the DIV limit (if it is in fact a limit). It could remain 500 while making the dodge artificially angle down faster using an impulse. The max height would still be as high, assuming you apply this impulse at or after the apex, but the distance you move and time it takes would be less just like this QT dodge.
The question is whether you could apply enough of an impulse so that it has the needed effect and at the same time doesn't make it feel bad like you're hitting your head against the ceiling or get shot from above with every dodge you do...

[Infiadux]

Oh I know what the UT99 dodge is like, I've done slomo and pulse/link secondary tests and comparisons myself in the past.

The question is whether that dip is real, meaning that the player collision hitbox physically goes down at that point, or that it's a view effect much like the landing view shake effect you get when you land after a dodge. You can see that happen both in the video as well as the pulse decal drawing on the wall angling up sharply at the end. The view changes, but the hitbox was already on the ground and remained there during that entire effect and also remained the same size.
I would put my money on it being similar to the landing view shake effect.


Having said all that, it might be an interesting thing to try out in UT4 even if it doesn't actually happen in UT99 as a way to bypass the DIV limit (if it is in fact a limit). It could remain 500 while making the dodge artificially angle down faster using an impulse. The max height would still be as high, assuming you apply this impulse at or after the apex, but the distance you move and time it takes would be less just like this QT dodge.
The question is whether you could apply enough of an impulse so that it has the needed effect and at the same time doesn't make it feel bad like you're hitting your head against the ceiling or get shot from above with every dodge you do...

Hmm, that could certainly be it.. maybe the eye height gets lowered down in the middle of the dodge to "prepare" for the landing view shake. If it's not that which allows the dodge to be shorter, then I guess UT99 could have two different gravity values for dodging and for everything else? Seems unlikely, but I just can't imagine it being possible to have the light gravity for jumping/falling and such a short and quick dodge at the same time.. Hmm, but then I think that would mess up dodging off high places since you'd basically land much faster than by walking/jumping off, so the UT99 dodge might just have a really low DIV and it's not noticeably snaggy because the player collision isn't a capsule.

As for trying the downward dodge impulse in UT4, as long as it's not too strong I don't think it will feel bad. Whether it's an eye height change or a physical character movement in UT99, I was never bothered by it there.

[Harry Spanner]

The mutator does not appear in the list when creating a custom game :s
I've checked where the files were extracted and it seems ok

Are you running UT on a Win64 client? Win32 and Linux should work as well, but I have no way of testing to be sure. If you're on a Win64 client, these are the two file paths that matter:

1. Epic Games\UnrealTournamentDev\UnrealTournament\Content\Paks\BP_Mutator_QT-WindowsNoEditor.pak
2. Epic Games\UnrealTournamentDev\UnrealTournament\Plugins\Mutator_QT\Binaries\Win64\UE4-Mutator_QT-Win64-Test.dll

I don't know if this mutator is meant to be a suggestion on how to change the standard movement in UT4 or just meant to be an alternative mutator people can use if they want to.

To me this direction for movement was appealing in theory, and I wanted to build it so people could try how it feels. Some of the changes, like the more flexible wall run, were made from conviction. Most of the changes were made out of curiosity, to pursue the idea of mixing Quake and UT.

The FS2FS is even slower than in QL with your value of 6. A friction value of 8 would match the graphs up better than they do now...

Yeah, I actually tried 8. You can get the graph to look closer that way, but there's a pretty big difference in feel when transitioning between forward-backward motion and side-to-side motion. That's why I settled on 6, the friction value from Quake III.

I'd like it [sprint] to speed up even sooner, speed up even faster and go to an even higher speed than both current UT4 and your mutator do.

The first values I tried were 1100 run speed, which I think is a pretty close approximation for Quake III, and 1400 sprint speed, which is over 99% of dodge speed. That run speed combined with instantaneous direction change from dodging made the game speed feel way overboard to me, and the sprint speed felt a bit silly too. Subjective, yeah, but that's why I reduced both speeds a bit.

As far as your main goal of strengtening strafing and not having combat rely on dodging as much, I actually think the opposite is needed at the moment in UT4. Strafing is probably more powerful than every other UT and dodge is probably the least powerful it has ever been. A lot of combat is happening with strafing only right now, which in previous UT games was rare.
Now if you just want to have a more Quake like experience, involving less dodging in general, then that's fair. For the default game I don't see that as the way to go though.

I like the faster run speed but I do think your idea of making strafing more powerful isn't the right direction to head in.

Agree with Stolid and Barktooth here regarding strafing vs dodging. Personally I think strafing doesn't need any buff anymore whatsoever, it's already really powerful in this UT4.

I certainly agree this is a significant departure from historical UT, but then again the nimble strafing in Quake is a big part of what has made it such a competitive classic.

I knew some of you would disagree with this change in theory, so I'm even more curious about how it feels to you guys, irrespective of historical UT and the current weapons. If you do take the current weapons into account, do you feel like you become harder to kill given the more nimble strafing? I gather a lot of people think it's too easy to get kills in UT4 right now.

I like the shorter dodge though, it always felt to me that it was too long in UT4. Still, it's the same speed as the UT4 dodge which is just too slow, especially with the improved strafing.

To faithfully explore the basic idea of QT, I can't make the dodge any faster without increasing the strength of dodge relative to strafe and sprint.

Also like the smoother feeling slide, I guess that's due to the lower friction.

I think that's part of it. Also, I increased the slide duration slightly, from 0.5s to 0.57s.

... you are referencing brizz's prototyper on line 2? ... is that right?

Oops! Typo. When I was writing the OP, I referenced Brizz's directions because he knows what he's doing, and I don't. Fixed now.