BSP Builder

Here is a small collection of my twitter/meme single page info busting pages (originally inspired by @radiatoryang )
There was plans for more, but time, energy and enthusiasm got the better of me, maybe one day worth returning to.

 

View attachment 1834

Sketching on paper vs. working in 3D is a long-debated topic in architecture as well as level design. I think it is a very valuable skill to be able to not necessarily design in 2D, but to be able to communicate the idea of a layout through 2D sketches. That is to say, the sketch is a jumping off point, a way to work through some bigger picture design problems, and the layout can later be refined and evaluated in 3D software.

Typically just a plan sketch is not enough - you may need multiple plans or section cuts / elevations to communicate height variation. You might also choose to learn how to sketch in 3D, although this can become complicated quickly.

I bring this up because I think being able to communicate a 3D design through 2D means is a cornerstone of the architectural discipline, and it is a valuable skill to have. It's less important with level design, since your 3D software builds the final result, rather than in architecture, where your 3D model exists to produce construction drawings. Regardless, the ability to sketch builds some useful skills:

1) Ability to communicate a 3D environment through 2D means (which even includes things like how to take good screenshots)
2) Ability to understand plan drawings more completely
3) Ability to work through design problems and take notes when you aren't at your computer

tl;dr: I agree that beginners who design on paper are less likely to understand how to utilize height variation, and you should be doing the majority of your design work in the editor. I just want to also advocate for sketching as a useful tool to have at your disposal.

 

Sketching on paper vs. working in 3D is a long-debated topic in architecture as well as level design. I think it is a very valuable skill to be able to not necessarily design in 2D, but to be able to communicate the idea of a layout through 2D sketches. That is to say, the sketch is a jumping off point, a way to work through some bigger picture design problems, and the layout can later be refined and evaluated in 3D software.

My view on this is that it's mainly good - as you identified - for communicating ideas, which is very important if you have multiple people working on a level. That's rarely the case when it comes to hobbyist mapping however: it's usually just one person making a complete level on their own, and if there is a team collaborating on a project, each mapper still mostly maintains ownership of their own level.

Personally, I aspire to create layouts that can only be fully understood and appreciated in 3D, and to favour the player's perspective over a plan or out-of-bounds view. Where translating between 2D and 3D is an essential skill in architecture, in level design it creates an unnecessary barrier between the intention of the designer and the experience of the player. Because of that, I'm sort of resistant to 2D layout sketches on principle. Some people find sketches of individual areas to be helpful, but I find that sketching in Trenchbroom is so fast that it has limited use to me. Where I find it's most applicable to Quake is for planning and structuring huge maps (on the scale of ad_sepulcher, for example). I'm yet to attempt anything like that personally.

I'm loving the insight that you as a source mapper and architect are able to bring to this community, btw.

 

I agree with Idolon. I don't sketch much myself... but when teaching level design to beginner students, I found (small, fast) sketches were important, as well as pitching their sketch to others. It makes the designers try to write / say what they think the level should be about.

Experienced level designers can probably do this without a drawing, but imagining space and player experience doesn't always come naturally to people. They might just stare at the level editor screen with anxiety and no clue about what to do.

Will the layout sketch ever survive reality in the editor? Of course not. But even that lesson -- the limitations of layout drawings -- is something the beginners don't really feel / understand until they draw layouts and see their blockout diverge. It's hard to even get them to playtest and observe honestly, which is also maybe a core level design skill we don't emphasize enough.

re: the last drawing, I want to note -- even "multiplayer 3D shooter level design" is a huge genre with different styles? It depends so much on the game. For example, Counter-Strike maps are mostly flat with very limited use of verticality, some good bomb sites are like 50% dead-end-ish which is what makes them interesting, and a lot of good CS mappers plan initial travel times and pacing on paper. And there's maybe more CS-Valorant players / maps than CTF players / maps today. Shouldn't this be the new template for a "default multiplayer game mode" then? Is this "evolution"? Is Quake CTF style thinking kinda old and obsolete? (Maybe?)

idk, this is part of my own guilt for that original level design post too -- there's a bit of truth but there's so much missing nuance and edge cases, which doesn't work for our type of level design (old school, room-design) -- our work is all about nuance, details, and edge cases?

tl;dr: it's complicated

 

I really wonder if there any level designers who don't blockout but making actual levels right away? Or is it completely banned prohibited practice in real gamedev?

I am wondering because for me makin' blockouts is close to impossible. They always lead to stagnation and lack of ideas. While when I start seeing actual design of a place, exact visual idea - fantasy starts working in both directions: visuals & gameplay. For clear example - I see that in this particular room bricks are cyclopian gigantic, and right away I see an opportunity to make some destraction, remove some of those gigantic bricks and put some mobs in there. etc. Exact visuals are opening gameplay opportunities for me personally. Just blockout - doable of course, but really painful & uninspiring

Just wonder if I do it wrongly and such practice is really-really bad one. Or there are some boyz & girlz who work similar way?

 

I really wonder if there any level designers who don't blockout but making actual levels right away? Or is it completely banned prohibited practice in real gamedev?

I am wondering because for me makin' blockouts is close to impossible. They always lead to stagnation and lack of ideas. While when I start seeing actual design of a place, exact visual idea - fantasy starts working in both directions: visuals & gameplay. For clear example - I see that in this particular room bricks are cyclopian gigantic, and right away I see an opportunity to make some destraction, remove some of those gigantic bricks and put some mobs in there. etc. Exact visuals are opening gameplay opportunities for me personally. Just blockout - doable of course, but really painful & uninspiring

Just wonder if I do it wrongly and such practice is really-really bad one. Or there are some boyz & girlz who work similar way?

Blocking out is another thing that's really good when you have other people working on the project. It allows you to make changes to the layout before art goes on, which is important because professional game art is expensive and time-consuming nowadays, and it's often necessary because the art and layout are being done by different people. Apart from that, it's also not as important for Quake because the geometry doesn't have to be hugely detailed, and can be easily modified whether using final or prototype textures. I know Lunaran's "blockouts" are very similar to the final map, just with less detail such as 16-unit high steps instead of 8. My process for mapping involves doing whatever I feel like at that particular moment; I find it helpful to start in prototype textures because it means I don't have to start picking out textures right away, but I'll rarely complete a full blockout before beginning art, either with textures or just architectural shapes. I do this for the reasons you touch on - it's a lot easier to stay motivated when you can get a steady stream of inspiration from both art and gameplay.

To answer your question, in "real" gamedev, it's generally a very bad idea to do art from the beginning, because it makes it a lot harder to change and iterate on things, and often you have entire levels scrapped during development, or moved to different places in the story, which suddeny makes all of that work obsolete. You'll also have people desigining levels before the story is complete or the environment concepts fully developed, so doing level art at that stage removes that flexibility. Modding is nice because it's free from all those pressures and constraints.

 

Interesting read.
I am still new to mapping so can't really contribute much here, but I initially did start out on paper for ideas, but it wasn't long before I started to just build as I imagined.
I think familiarity with the tools, in this case Trenchbroom was an important factor in ditching the pen and paper.
The 3d window gives a immediate view of the results and you feel a bit like a sculptor, spinning the work in front of you and zooming along as if moving through on foot.
I have tried blocking out, and just finishing individual areas as I go, and both work, but I tended to find with blocking out that the maps got sooo big, when detail was added later the compiling time went up significantly. Whereas, finishing smaller areas and doing a lot of detail in them allowed for faster compiles, and an ability to work around design elements that were going to lead to slow framerates. I have tended to use everything I make, so unlike real game developers I doesn't make much difference in terms of the detail.
All that said, the pen and paper just came out again to flesh out some new ideas.
In terms of detail I would like to see more maps chocked to full with detail, the new capabilities of sourceports like Ironwail should surely result in maps of a different type to the years of empty corridors we have grown so used to.

 

I've started using the Townscaper "toy" to get a rough idea of level layouts. You have to look past the "everything is a house" visuals and it can be a bit confusing, but it's still pretty useful in quickly getting a sense of how a level might be built. You can also export it to .obj format and then walk around it in first person using a browser tool. It's not ideal, but I find it very helpful to get that first pass of the general shape and scope of a level.

In terms of detail I would like to see more maps chocked to full with detail, the new capabilities of sourceports like Ironwail should surely result in maps of a different type to the years of empty corridors we have grown so used to.

^^ That's my intention. I'm also doing something different with textures, working at 4x or 8x normal Quake resolution whch gives a very interesting aesthetic. I discovered Quake in Feb this year and have spent 6 months playing a ton of mod and noting the changing styles once powerful source ports became available. As a result, I think there's a lot of potential to increase the visual quality of maps: whether players like it or not remains to be seen....

 

I've started using the Townscaper "toy" to get a rough idea of level layouts. You have to look past the "everything is a house" visuals and it can be a bit confusing, but it's still pretty useful in quickly getting a sense of how a level might be built. You can also export it to .obj format and then walk around it in first person using a browser tool. It's not ideal, but I find it very helpful to get that first pass of the general shape and scope of a level.

^^ That's my intention. I'm also doing something different with textures, working at 4x or 8x normal Quake resolution whch gives a very interesting aesthetic. I discovered Quake in Feb this year and have spent 6 months playing a ton of mod and noting the changing styles once powerful source ports became available. As a result, I think there's a lot of potential to increase the visual quality of maps: whether players like it or not remains to be seen....

There's definitely a core of fans who prefer to stick to the predefined format, or to merely opt for small enhancements that don't overall change the mechanics of the original, but personally, after 26 years, and thousands of maps, there's some real boundaries being pushed. Personally I wish I had coding skills as I would attempt to create some role playing elements and go for something along the lines of the Elder Scrolls, but in the Quake engine.

 

I really wonder if there any level designers who don't blockout but making actual levels right away? Or is it completely banned prohibited practice in real gamedev?

I am wondering because for me makin' blockouts is close to impossible. They always lead to stagnation and lack of ideas. While when I start seeing actual design of a place, exact visual idea - fantasy starts working in both directions: visuals & gameplay. For clear example - I see that in this particular room bricks are cyclopian gigantic, and right away I see an opportunity to make some destraction, remove some of those gigantic bricks and put some mobs in there. etc. Exact visuals are opening gameplay opportunities for me personally. Just blockout - doable of course, but really painful & uninspiring

Just wonder if I do it wrongly and such practice is really-really bad one. Or there are some boyz & girlz who work similar way?

I've been trying out different ways of building maps recently as I really want to start releasing some maps rather than ending up with a lot of half-finished maps. I find that if I try to build actual levels straightaway and detail as I go along, I run into problems when certain aspects of the architecture don't match up and it takes a lot of effort to change what's already there and get it right, which bears out others' arguments for doing blockouts. But when I've tried the full-on blocking out approach, I've tended to stagnate quickly and lack inspiration because the blockout is typically too far removed from the sort of final product that I want.

It's clear to me that I need to find a happy medium in between the two. My latest attempt is to loosely follow dumptruck_ds's suggestions in his YouTube videos, starting with a low detail pass that focuses on floors first, and then walls, and then gameplay elements and ceilings, I'm currently seeing how that evolves. But I imagine that the blocking out approach is more likely to be optimal and in some cases necessary if several people are working on a map, as is common in "real" game development.

 

I personally make the map along the way with full detail. I actually see it be born in front of my eyes and discover it gradually like the player will. I just have to listen to my feelings to know what's next, what feels right. It's intuition driven.
Only once (after having watched dtds' vid) I tried to block out the place first and it instantly inhibited my inspiration totally. I had to ask a friend to help me and that's he who made the map out of the blocking out because I had stuck myself creatively and couldn't simply move on. Never blocking out again!!!