Login | Register 

Urban Terror Blogs

Zenity's blog


pencil Movement Update

bullet_none posted Saturday, 23 January 2016
Over the last few weeks I have been working on some movement related features.


In my mind, Urban Terror movement is primarily defined by two words: "Quake" and "Parkour".

The Quake heritage is obvious, but what exactly does "Quake like movement" really mean? When boiled down to basics, what makes Quake movement special is that it's heavily physics driven. Quake never had any sort of fancy movement system. There's no dodging, no double jumps, nothing but WASD and a jump key. Yet the simple whacky physics create gameplay so complex, that Quake movement is considered the most skill based of all. In game design, this is called emergent behaviour and the result is the holy grail of "easy to learn, impossible to master". So when I add movements to Urban Terror, I try to keep this in mind and follow the same pattern. Movements which are easy to execute (no extra keys required) but are driven by physics so that advanced players can bend it in unforeseen ways.

"Parkour" is the perfect real world equivalent to our crazy ass movement, but it has always bothered me just a little that actual Parkour is in fact more agile and advanced than ours. Vaulting and Wall Running are two movements in particular that make perfect sense in a Parkour environment, but I also wanted to implement them in a way that doesn't feel scripted but heavily physics based to fit in with the Quake style. I am very happy how this turned out, and I think that both will become very natural additions to Urban Terror once players had a chance to get used to them.

I am also experimenting with additional legacy Quake physics (similar to Challenge Pro Mode), like ramp jumps and bunny hopping / air control. You can see both in the video, but notice that they are extremely toned down so they don't overpower the usual Urban Terror movement but instead create subtle additions to it. Bunny hopping in particular still needs a lot of testing and feedback so there is no telling if it will actually make it in the end. But ramp jumps are pretty subtle while still being a blast to execute, so I think that those will be fairly uncontroversial.

I appreciate the irony that Urban Terror on Unreal Engine could well become more Quake-like than it ever was on the actual Quake engine.

What About Jump Maps?

Of course we will try to maintain 99%+ accuracy for basic movement (this includes circle jumping, wall jumping, powersliding), but the reality is that most jump maps will have to be updated for HD either way due to minor differences in the engines which are not worth changing (we will only port the bugs which are actually fun :)). Furthermore we cannot stay stagnant if we want HD to be a true resurgence rather than just a monument to days long past, so jump maps will have to evolve along with us. Of course this is not an ideal situation (especially for high score tracking), but we will work with the jump community to make the process as smooth as possible. If you are working on a jump map, please don't hesitate to reach out to us!

That said, one reason I wanted to put a lot of work into movement early on is that hopefully we can get it to a point where we won't feel the need to make significant updates to the movement system for a long time to come.

What About Action Mode?

In a previous blog I hinted at the possibility of an alternate "action mode" that would increase the overall speed and drop all pretences of being tactical. However, after experimenting with this it just doesn't feel right, mainly because the existing maps are not suitable for it. In the end, I think that adding subtle amounts of pro mode style movements to stock Urban Terror is the better approach. Instead of an "action mode" however I could imagine to make infinite stamina a proper option for all game modes. Right now stamina is the one thing that reigns in all the crazy movements and maintains a tactical element to the game. So playing without stamina is very much like playing "action mode", just one that still feels like bona fide Urban Terror. I do think this option would be justified since sometimes we just feel like mindless fragging, whereas other times we may feel like deploying a slightly more deliberate approach.

Anyway, I hope you enjoyed this update and like I said in the disclaimer of the video:


This is still early days, and if there's a time to make radical experiments, then it is now. :)
comments: Feel free to discuss this on our forums
linklink bullet_sharethisshare

pencil New Year's Development Update

bullet_none posted Saturday, 02 January 2016
Things have been rather quiet from me the last few months, which is mainly because I've been unusually busy. But lately things are getting better and HD is progressing really well, so it's about time for a quick update.

I have nothing big to talk about this time, as most of the work has been on details and putting things together. Turnpike is coming along really nicely and is probably the best example for a classic map that received the full HD treatment. FrankieV is doing a cracking job on it.

Here's a quick rundown of some of the new stuff. Note that the gun has no recoil in this demo, that's obviously going to change. Everything else is work in progress as well. The weapon itself is a stock model, but I like it a lot so I expect that it will make it into the game. Also this is medium quality since my PC is not the beefiest. Don't mind the music, just having some fun. :)

Surface Effects

One thing I've been working on is integrating the obligatory surface effects including customizable footsteps, impact effects, bullet holes, and impact sounds for each material. UnifiedTone has been recording the footstep sounds. I tried to show some of this in the Turnpike video, but this video gives a more detailed overview of the current capabilities:

Activatable Objects and Prompts

Another new feature you can see in the Turnpike video is activatable objects that are hooked up to the game's Activate button. This is accomplished via a simple Blueprint interface and also allows the mapper to set custom prompts, which will helpfully appear on the screen as you get in range of the item. Combined with the virtually unlimited potential of Blueprints, there is very little that you can't do. Want to make a Duke Nukem style flushable toilet, helpfully prompting you to "take a piss"? Knock yourself out!

I expect that we will release this and other map objects together with a mapping guide in a simple mapping kit fairly soon. This definitely wouldn't be the worst time to begin learning UE4 and start mapping for HD. :) Glory and fame awaits!

Quo Vadimus?

2016 is poised to become a fantastic year for Urban Terror with 4.3 around the corner, and HD shaping up nicely. I've had a few personal epiphanies myself, and have never felt more motivated to work on Urban Terror or more optimistic about its future.

I absolutely believe that there is a demand for a no-strings-attached shooter that is simple enough to get into to be a great casual game, while offering a high enough skill ceiling and consistency to be a viable esports title. The kind of game you get hooked on because of its simple yet satisfying mechanics, and then stick with it as it gradually evolves year by year. No subscriptions, no pay to win, no yearly $60 update plus DLC. Just fragging with style, pure and simple.

This is Urban Terror.

Happy new year!
comments: Feel free to discuss this on our forums
linklink bullet_sharethisshare

pencil HUD, Domination, Menus, and More

bullet_none posted Monday, 27 July 2015
It's been over two weeks since my last post, so I figure it's about time for an update.

Progress has been very fast, but I've been focusing on getting a hang of the engine in various areas while laying the groundwork for future development. As a result there hasn't been anything in particular to showcase yet, since most of the features are still in half-done state. That said, here's a quick rundown of what I've been working on:


I have focused on adding some missing combat features like fall damage, limping, locational bleeding, and a primitive bandage function. Not much to say here really. Fall damage and limping behaviour are almost identical to Q3UT, whereas bleeding and bandaging will likely be changed a little bit.

Also thanks to HappyDay's measurements, I have been able to tweak various other aspects of the physics to make them more identical to Q3UT. This includes the exact dimensions you can fit through, or steps you can climb up on.


Domination is the first game mode I have been implementing. Most of you probably know what it is, but for those who don't here's a quick description:

You fight over control for three flags on the map. If you control more flags than the opponent, then every five seconds you get points based on how many flags you control more than the opponent. Your team also gets a point for every kill. When you die, you quickly respawn randomly in the quarter of the map which has the least enemies.

It's a very simple mode, and I think it will be one of the best modes for new players to make their first steps with Urban Terror. It's very easy to understand (go to nearest flag, kill people) and the punishment for death is manageable. That said, simplicity is not the only thing the mode has going for it. It also creates very balanced match ups on practically any type of map (which is very useful for Urban Terror's diverse range of maps), and in a competitive setting it can offer a surprising amount of depth which mainly centres on strategies to hold two of the flags at the same time. Due to the random respawns it also provides quite dynamic gameplay and does not suffer from the spawncamping issue.

It's not quite done yet, but I am very happy with how easy and fast it was to implement the basic game rules in UE4, and we'll be able to finish it quickly once we are ready for testing. Another reason I started with Domination is that I believe that it's the best mode for early testing, since it focuses the action and does not require advanced teamplay.

New Player Model

The art team around FrankieV and BladeKiller have done a fantastic job working on a new high quality player model (starting with Sarah) and I've been working with Frankie to get it into the game and hook up the animations. There is still a lot of work to be done of course, but it's coming along really nicely as you will see in the video further down. :)


I have also done some work on the HUD, mainly re-creating the health/stamina bar of 4.x. This is probably by far the most complex element of the URT HUD, so I wanted to get it done early. My goal for the design was to stick with the current design but give it a bit of an HD treatment. I think it turned out very nicely, although I'm sure that it will receive further touch-ups. You could already see it in action in HappyDay's jump video, but here are some "close up" shots:

Menu System

Finally, I've started working on a new menu system and the basic framework for our customisable character models. This is still extremely early work in progress of course, but at least we can choose maps and characters already.

This video provides a quick glimpse of most of the new stuff:

Action Mode?

One thing we discussed is the possibility of an "action mode", which would essentially be to Urban Terror what CPMA is to Quake 3. Urban Terror 1.x began as a much faster paced game than what is is today, based on the spiritual ancestor Action Quake 2. Right now there seems to be a lot of support for increasing the pace of the game within the team, so this is something we may want to explore.

An action mode could have features like no stamina, faster base speed, CPMA style air control, and ramp jumps. In other words, it would lose all pretences of "tactical" movement in favour of mad skill based carnage.

Of course there is the concern of playerbase fragmentation, so we are not going to add such an option lightly. But this seems to be a good time to give it a whirl and simply see how it goes. If it looks like it could cause a serious split in the community, then we'll probably want to go back to the drawing board and see if we can come up with a better compromise.

Right now this is just a thought experiment so it may not happen after all, but I think it's certainly interesting to mull over. I also appreciate the irony that UTHD on Unreal Engine could become more Quake-like than Q3UT ever was. :)

What's Next

One thing we are still missing is a test server and the ability to connect to it, so I will prioritise this as well as fixing any missing bits that stop us from having a decent game. This shouldn't take too long, so hopefully we will soon be able to show some actual gameplay and then we can build on that.

As usual, I will keep you updated!
comments: Feel free to discuss this on our forums
linklink bullet_sharethisshare

pencil Hits and Hops

bullet_none posted Friday, 10 July 2015
Since my last post, I have been working on implementing spread and recoil. Here's what it looks like so far:

(Of course all assets in this video are either placeholders or work in progress)

I've tried to stay close to the current behaviour, without any major experiments. A couple things of note though:

- In Q3UT, increasing recoil will make your crosshair jump towards the left and right, so it never actually points at the position you are aiming at. While I do think this is a useful feature to "feel" the increasing inaccuracy of spraying, and you would not be able to hit anything by directly aiming at it anyway once it gets so bad, it doesn't feel really good to me. So I tried to replace the left right movement with a camera roll which still conveys the lack of accuracy, but does not actually move the crosshair from where you are aiming. This feels really good to me so far.

- I have also added the option to set a random seed for each weapon. This means that spray patterns are always identical and while this still won't allow you to perfectly control a spread like Counter Strike (which wouldn't be a good fit for the game IMO for various reasons), it will allow some sort of control and reward experience. By and large though, you are still much better off burst firing.

In the end of the video I hop around a bit to show the state of wall jumping and powersliding.


Also some people mentioned the issue of "hits" in general, so I thought I'd quickly share my thoughts on this. I believe there are generally three areas which can potentially lead to faults with hit detection. From least to most significant:

1) Random Spread

Sometimes you just have bad luck... Urban Terror has extremely accurate weapons, so as long as you tap or burst your shots, you will hit what you aim for. Spread escalates quickly though, so you can get the situation where you just barely miss your first couple of shots, and then your spread already becomes large enough to make you miss most of your shots. This can lead to the odd situation where you aim is almost perfect, yet you don't hit anything. Hit detection in Urban Terror can be extremely unforgiving due to the relatively small hit boxes coupled with a relatively fast movement speed.

We can counter this by making it more obvious when the accuracy goes down (especially for players who don't use an expanding crosshair).

2) The Hit Mesh

ARIES was way ahead of its time, and there's no wonder it had some growing pains. These days I believe it works very reliably, and that's quite remarkable considering the technology it's based on.

That said, UE4 makes life a lot easier for us in this regard. Instead of per-poly hit detection on a low poly model, we use the physical asset for hit detection, which uses simplified shapes and is of course very well tested. The physical asset follows the reference model perfectly, as you can see here:

3) Netcode

Finally and most importantly, network code is always fickle and things can easily go out of sync. This becomes even more true with small and accurate hit models. No matter how well designed the network code is, there is always a chance for inaccuracies caused by flakey network conditions or server performance issues.

Fortunately, there is an option which sidesteps all those headaches: Client side hit detection.

Now before you throw eggs in my face, hear me out. Client side hit detection has become more popular in modern games, for one simple reason: When the client is compromised, then all is lost already. Allowing a compromised client to mess with hit detection sounds scary, but the truth is that this kind of hack would be the easiest to discover, not to mention incredibly boring for the cheater. With every random hack featuring perfect aimbots and wallhacks, there is just no reason any more to be concerned about this. Of course servers still do basic verification of each hit and any obvious tampering would be very easy to spot. The bottom line is that when in doubt, the client is trusted that a bullet was actually a hit.

One commonly perceived downside is the effect of getting hit behind walls, but this is also an issue with server side lag compensation, and it pales in comparison to missing shots. It is also up to us how lenient we want to be with this, e.g. whether to reject hits if the target has moved behind a wall by the time the hit arrives at the server.

The giant advantage of client side hit detection is simply this:

Perfect hits. Always. Without exceptions.
comments: Feel free to discuss this on our forums
linklink bullet_sharethisshare

pencil Zenity Checking In

bullet_none posted Monday, 06 July 2015
Hello everybody!

I thought I'd quickly introduce myself. My name is Daniel, I'm from Germany (currently living in the UK) and I'm 33 years old. I have spent the last three years or so working on and releasing my first indie game in full time and am now joining the team of Frozen Sands to help out with Urban Terror HD. My primary skills and interests are programming, game design, and UI design.

Coming from Action Quake 2, I have been following Urban Terror even before beta 1.0 was released. With 1.27 I started to get really into it, and 2.0 was the release that made me a huge fan. I went on playing it very competitively for some time, after which I have been playing on and off and following its development loosely.

I believe that HD is a chance for Urban Terror to truly live up to its potential, as long as we focus on our strengths and go back to basics. The game's core movement and shooting mechanics are still unmatched in terms of being extremely skill based and yet highly accessible. If we combine this with a modern engine, modern community features, and smart refinements of the gameplay, I am absolutely convinced that this game is still relevant. I see its target audience mainly as the competitive gamer who is looking for an old school PC(!) shooter that is more fast paced than Counter Strike but more accessible and refined than Quake Live.

My goal is to get HD into a playable state as soon as possible, so we can inject some life back into the community. If we are to make this happen, we need all of you to cheer us on! :) I know that things have been tough for loyal fans during the recent years, and I shouldn't make any empty promises, but what I can promise is that I will work as hard as I can to make this new chapter of Urban Terror's story a success.

The first thing I have been working on is to help Barbatos finish the port of the basic movement physics. This is coming along really nicely, with the most important movements (especially strafe jumping and powersliding) now being near 100% identical to the original. There is still a lot of tweaking to be done, but rest assured that it's going to feel like genuine Urban Terror, not any other Unreal Engine game.

Just now I have been working on the hit detection:

I am now planning to work on basic weapon behaviour and shooting mechanics. I'll keep you posted!
comments: Feel free to discuss this on our forums
linklink bullet_sharethisshare


Copyright © 1999-2017 Frozensand Games Limited  |  All rights reserved  |  Urban Terror™ and FrozenSand™ are trademarks of Frozensand Games Limited

Frozensand Games is a Limited company registered in England and Wales. Company Reg No: 10343942