Turtle Rock Studios Interview

Interlopers.net had the chance to ask Turtle Rock Studios a few questions regarding their role in the creation of CS: Source maps and content.

Firstly, thanks for taking time out to answer some questions, I’m sure everyone will be interested in what goes on behind the great CS: Source releases we have had lately.

To start off with how many people make up Turtle Rock Studios and how long has the company been working alongside Valve on Counter-Strike content?

(Mike Booth)

Turtle Rock Studios is starting its fourth year of working with Valve on the Counter-Strike franchise. We have been growing steadily from late 2002 when I was contracted by Valve to build the Official Counter-Strike Bot. The studio then expanded to assume development responsibilities on Counter-Strike: Xbox and Counter-Strike: Condition Zero in 2003, and then transitioned onto Counter-Strike: Source.

There are currently eleven of us at the studio. Our philosophy is that a small team of talented and collaborative professionals can be more effective in many ways than giant teams of hundreds of people. This is proving to be true and very effective for us.

Valve has been extremely supportive throughout this process, providing us with their expertise and guidance as game industry veterans while simultaneously allowing us the creative freedom to excel on our own.

From a quick glance at the Turtle Rock Studios website you can immediately see an impressive roster of Source maps. Once a map is chosen for development what are the usual steps the team will go through to bring a map to completion?

(Chris Ashton)

1. Relearn the map: So far our mapping focus has been bringing the best maps from 1.6 and CZ over to CS: Source, so the first step is always to spend some time familiarizing ourselves with the old map (even if we already know it). We play it, we look up old map guides, we discuss the flow of the map, etc. We need to make sure that we know what areas are fun and why, what areas could use some work, etc.

2. Pick the environment: We like our Source maps to look like real environments. Not only does the scale of the world need to be correct, but also every room needs a purpose and every wall needs to be there for a reason. So the second step is to decide on a theme or setting that will work with the layout/design of the old map.

3. Research: Once the setting is decided, we spend some time gathering reference. Most of the time this involves a lot of Internet searching, trips to the bookstore and in some cases photography trips to a real environment. We need to know how to build it (architecture), what materials are used (textures), how it’s lit (lighting) and what objects can be found in that environment (models). For de_nuke, I have over 100 reference photos of nuclear power plants saved on my hard drive. For de_inferno I purchased some books on Tuscany and villages in Spain, and also took a number of trips to some Southern California Missions.

4. Build it: Then we get to the real meat and potatoes, building the map. We start by importing the old map into hammer. If we don’t have the source files, the .bsp is decompiled and then imported. Then the entire map is rebuilt using the old geometry as guides. Our mappers make their own world textures, so geometry gets textured as it is built. While building, the mapper figures out what models will be needed and assigns them to our prop modelling team. Lighting is done along the way.

5. Test it: As soon as there is a playable version of the new map, a nav mesh will be created for the bots and we begin playtesting with humans and bots. We’re looking for any red flags here. How is performance? How is gameplay? How does the map look? Testing will continue until release.

6. Polish it: Final props are added. Final lighting is added. Effects are added. The soundscape is added. The map overlay and mission text files are created.

7. Bug fixes and optimizations: Time to fix the last of the bugs and optimize the map. Optimizing involves different things and different lengths of time depending upon how much frame rate is needed. Prop fade distances are set, portals are added, extra textures are removed, LOD’s are added to the models, etc.

How long does it take to reach a final product from conception? Will there be various versions and ideas tried or do the team work to one clear cut vision of what the map will be?

(Chris Ashton)

Map time varies a lot depending upon the mapper, environment and challenges. Since we are working from existing maps, there usually isn’t a lot of experimentation to be done. We know what we need to do from the word go. To keep everyone on track and headed in the same direction, we have reviews every Monday.

Having multiple level designers work on a single map would pose a difficult situation to most teams. If this is the case how is the workload usually organised?

(Chris Ashton)

You are correct. Having multiple mappers working on one map would complicate things, especially for us, so it’s pretty rare for one mapper to touch someone else’s map.

Your latest releases are heavily integrated with prop models and skilful texture use. How closely do the Modellers/Level-Designers/Texture Artists etc work with each other? Is a similar ‘cabal’ system used as utilised by Valve?

(Chris Ashton)

Good questions, I love it! As mentioned earlier, our mappers all make their own world textures while they are building their maps. We do not have a dedicated texture artist here. On the last few maps that we made, we have had an awesome duo of modellers (Sean Keegan and Todd Williams) working full time along side the mapper. The mapper works as the art director, finding visual reference for the model, assigning it to one of the modellers and signing off on the final model. If the model is shaped specific to the environment (like a set of I-beam roof supports), the mapper will rough it out in Hammer and export the geometry to be used as a guide.

Have you had to overcome any major obstacles on any projects while using the Source engine? Are there many occasions of ‘back to the drawing board’ with map ideas?

(Chris Ashton)

The Source engine is quite robust. Over the years, I’ve learned that you can do just about anything that you can think of. There is so much stuff in there that the challenge has been figuring out HOW to do what you want to do. As of late, I’ve been getting a lot of answers directly from the Valve Wiki website. Occasionally we do run into little items here or there that we’d like to add, and either Valve or someone here will code it up.

Counter-Strike has long been known for its competitive and non-competitive player base. When planning to convert maps popular in clan matches (de_inferno for example) are specific considerations used to keep them true to their CS 1.6 counter-parts?

(Mike Booth)

With few exceptions (cs_militia being one), all of the classic maps we have remade for CS:Source have been built directly from the underlying skeleton of the original map. The dimensions of the basic world geometry are often identical, and the routes are usually the same. In fact, one of our biggest design challenges has repeatedly been to “make sense” of many of these map layouts – to make them believable spaces. With the Source engine, you can no longer get away with out of scale objects, or rough suggestions of detail. This means creating a huge amount of content, from coffee cups on desks, to garbage in trash cans, to toothpaste and toothbrushes in the bathrooms, and so on.

What always amazes me is after our process is complete how very different the map often feels. de_inferno was a good example of this. Although I knew the original like the back of my hand, the Source version has a unique setting and far more visual density. Although I could navigate by remembering the previous map’s routes, it felt like a wholly new place with new tactics to explore.

How closely do Turtle Rock work with Valve during the creation of a map for Source? Is their much creative control or will Valve specify guidelines per map?

(Chris Ashton)

Valve always knows what we are working on, but we have creative control over each map that we build. So long as we’re doing great stuff, Valve will back us. It’s really the coolest set-up I can think of. Our success guarantees our creative freedom.

Are there plans to re-make the entire official CS map line-up or selected maps only? If so what is this decision based upon?

(Mike Booth)

We looked at the most popular “stock” maps from Counter-Strike and Counter-Strike: Condition Zero based on actual gameplay statistics, and generally used that data to decide which maps to re-make and in what order. cs_militia was an exception to the rule: Although it wasn’t quite as popular numbers-wise, I had a great time playing that map “back in the day,” and we had some pretty exciting ideas on how to take it to the next level in the Source engine.

There are of course hundreds of budding developers looking for a slot in the game development industry. What advice could you give to someone in this position?

(Mike Booth)

First, continue your formal education. The demands of modern games require more depth and breadth of knowledge than ever before to create. That said, don’t wait to start working on your game project! There are more freely available tools with more creative power than ever before as well, such as the Source Mod SDK. The best way to break into the game industry is to have a game project that you have built on your own. It doesn’t have to be a huge project, but having something that demonstrates your talents and your commitment to gaming is invaluable.

What can we expect of Turtle Rock Studios in the future? Will you be branching out into other engines or projects?

(Mike Booth)

We are quite happy with the Source engine, and we will be announcing our next project soon.

What feature implemented into a CS: Source map are you most proud of and why?

(Mike Booth)

That’s hard to pin down. What I’m most proud of is how our team has continually pushed to make each map a unique and believable environment, and that everyone at Turtle Rock has really collaborated in a team effort to make that happen.

We also have very detailed soundscapes and other environmental effects, such as our interactive fish. While such things are not necessities for gameplay, they do add a great deal of depth and believability to the environment. Much like the physics simulation of the Source engine, these details create a richer overall experience.

Big thanks to the Turtle Rock guys for taking the time to answer the questions. You can find more information about them by checking out the Turtle Rock Studios website.

- By Blink 17/03/06