Mapping Techniques

[Snakes]

So, after completing my DM map for 32in24, I officially took a hiatus from Doom to catch up on other games as well as to try and build a fresh perspective on mapping and how I should approach it. Going back and looking at everything I've done, I came to realize that I was constantly experimenting and tweaking the way I map.

So my question is this: do you have a set mapping technique?

For the longest time, I'd create an are and plan everything on the spot right down to architecture, detail and thing placement. Towards the end of it though, I started creating a layout before doing anything (even texturing) and would follow up with general architecture and thing placement. Oddly enough, the map where I tried this initially (Map22 of UR) wound up being one of my more interesting maps from a layout standpoint - the interconnectivity is off the charts.

Even moreso, I often found myself choosing the music to a map before I even started it - sometimes I just had an image in my head and thought "action-packed slugfest" or "linear and haunting."

Do you have a set order to how you create a map? Have you ever bucked the trend? Do you find any advantages to your approach or are you constantly try to find that perfect order to things?

[Doomkid]

Snakes said:

I started creating a layout before doing anything (even texturing) and would follow up with general architecture and thing placement.

This is my approach and has been for a few years now. If i'd start detailing a map before the layout was finished, I'd find that it always ended up way to small or short.

I often pick a MIDIfied tune I like, and listen to it a couple times as I start building the map. I find it gives me good inspiration for the visuals, odd as that might seem.

Great topic!

[purist]

I get best results by sketching a layout including some setpiece ideas first, drawing out the layout and then arranging architecture, and adding things, detail and lighting after. This way usually gives me my best results fastest. Problem is that first sketch. If I can't think of what to sketch I sometimes just start laying down sectors in Doombuilder and the map is created more organically. This method I find requires more revisions and just generally takes longer to be something I'm satisfied with. It's also more likely to just not work out - but I still do it if I'm struggling to think of something to sketch.

[Phobus]

I'm quite partial to the "slumped at my desk, slightly to the side my mouse is" approach, personally :p

[warning, feature-length memoir ahead]

To actually answer the question, I've been meddling with my approach as long as I can remember.

When I first started out all those years ago, it was always "auto-create 32-sided polygon, then fill with stuff". Eventually I came to realise there was a bit more to making maps than that, so I started linking shapes together and dragging vertices around to create corners and other constructs. Luckily the saga of my mapping was then influenced by Hank Leukart, whose book taught me about creating shapes in DEU from the vertices upwards, gave hints and tips as well as tutorials on the technical aspect. This informed what, to this day is still the basis of most of my mapping, where I basically turn up with an idea I want to use (back in the 90's this was often just a specific line action, but these days it's either whole concepts or several key encounters) and build a map to build up to and continue on from that.

Initially this was achieved with room-by-room stuff, which you can see a lot of in Scourge, where each room was built to completion and tested before the next one was made. Drawbacks here were linearity (even branched paths or hub layouts were basically just added linear paths) and making testing quite dull on longer maps, so I'd often keep them short and easy. The best part was that, basically at any point, you could just drop the exit in and that'd do. I found a continuing narrative made the obvious linearity make a bit more sense, but it was a clear weakness.

I was also mostly placing monsters where they'd have some sort of story-based reason to be, like Zombies in built environments, having imps basically serve as vermin, Cacodemons living high up and giving Hell Knights a cave full of corpses to live in and one Baron to serve as an alpha. Obviously this is in no way conducive to solid gameplay, but it's how I thought in my early teens!

Now, the room-by-room approach was a weakness I picked up on fairly early, so I did change to mapping areas instead. I still try to do this today, as it's by far and a way the easiest approach to having nonlinear and open areas with plenty of interconnection. I used to go for having all the architecture done first and then putting in the monsters to fill the space, but with Warpzone that didn't work out very well, so I've since gone for making the maps around the intended encounters. Sometimes the visual is either the main aim or will win out if it's particularly good but doesn't play well. I have noticed, however, that building architecture first makes losing interest very easy, as you end up with an empty map that may seem daunting to populate... Probably the number one cause of me canceling maps, with simply losing interest coming a closer second.

So, I'm mapping out areas at a time to allow for more complex design and flow, I'm planning out the fights and thing placement in advance, but not letting testing those two occur until after I've got a full map. Sounds like a recipe for success, right? Well, the DNF-rate is remarkably high for me working like this, so I've done some experiments over the years.

For example, ZPack E1M2 was built entirely as a flat layout first, then heights, flats, textures, architectural embellishments and other theming were added, then functionality and finally thing placement. I've done this a few other times and it works wonders for small maps that have a very open layout. It's painfully slow for big maps though, as you end up doing multiple passes and you end up focusing on some areas and trying to manage the aesthetic of the whole map at once. I've also noticed it can make getting the visuals right a PITA, as I don't think about those at all until I'm trying to force some semblance of aesthetics into a layout designed purely for flow. In one case (the grind map that can be found in my grab bag) I also was so focused on good flow that I didn't actually put in any intuitive start point. It's a fair technique for DM mapping, if my 32-in-24 experience was anything to go by, but it's easy to overdo the layout in speed mapping sessions with this.

I've tried mixing up what I map first before, going straight for the key idea or even the exit and then working back to the start. Mixed results here, as I'm often then happy to have made the idea and realise that was all I wanted to do, so I end up with a scrap that often gets deleted when I finally admit I won't be finishing the map - you need a few interesting ideas with this or you either get filler or no map, it seems!

One other approach I had was making little areas and then sticking them together or using them to finish a map I didn't have the enthusiasm to get all the way through but didn't want to waste. Most notable example of this is CC4 MAP07, where the western base was all just a direct insert. The map feels disjointed with this though, plus any weaknesses inherent in the approach used for that section can really be glaring if you were using a different technique to that point, as I was there.

All in all, I'd say I've covered most approaches by now, sometimes all in one map. I think having a couple of ideas you're really keen on, mapping start-to-finish area-by-area with gameplay in mind but not implemented until the end feels like a reliable workhorse approach after a this time. I think that's how all of my last 6 most recent releases and community project contributions were made and I'm pretty happy with them.

EDIT: Oh, and get a basic texture theme and light level in early, as applying it from scratch when you've got an entire layout already built had proven to be decidedly tedious in my experience. Even if it's mono-textured with flat lighting, you're already closer than if you're at the defaults for your editor and need to change literally every line. This is worth catching whatever your approach is, even with the layout-first tack.

[Phml]

Starting with a sketch, like many. Although I find it interesting that my sketches don't look anything like the few I've seen posted on the forums. Those of you who shared such stuff tend to draw things neat, with accurate dimensions and precise shapes, it almost looks like an automap shot.

Mines are rough scribbles on graph paper, with lines overlapping each other as I have new ideas, line actions jotted down with a number next to the activator and a number next to the affected sector, different areas connected by letter codes as I run out of paper, monster types written out, sometimes entire paragraphs of random musings written on the sides and the occasional 3d doodle to try to think of cool architecture.

But ultimately I find I need to fail and take a break before I get anywhere.

i.e. doing the first room, drawing sectors, moving vertices, picking textures, I very quickly get to a point where I feel like I'm not getting anywhere and want to throw my hands up in disgust. This is dumb, I fucking suck at this, this editor is stupid, why is it so hard to translate my super cool ideas from thought to reality, god I hate this, this is bullshit, etc.. You know, standard stuff. ;)

At that point if I sulk and keep going, I won't make any progress and it will eventually discourage me. Instead, getting up and going for a walk inevitably gets me to figure out a better way to do things. Then once there's one small piece of map architecture I'm happy with, everything else snowballs from there.

So it's all about getting to that initial failure state as fast as possible so I can overcome it and move on. Probably a terrible way to do things, but whatever works. Kudos to those who can just do a whole layout in prototype mode or even fully draw the whole thing on first go (that Ribbiks timelapse video is mindblowing).

[Phobus]

I figure I'll muse on what Hank Leukart said too, as this is a topic that interests me. A lot of his hints and tips worked in the form of pitfalls and how to avoid them. For example, linear room-door layouts are, generally speaking, not exciting to explore, so how do you get around it? Well, think about how else you can get between rooms, then think about what you'd need to change to allow for stairs, drops, lifts, teleports, etc. Often the answer is height variation, so that's a good place to start. If one room is higher than the other, you have to have more than a door to get between them, so it really helps to keep changing heights and keep thinking "What can I do here that isn't a door?" - I've taken that to heart and I'd like to think it shows in my mapping. Plenty of ledges, pits and stairs to be found, with a few doors to break up areas too.

As for linearity, he suggests "the webbing technique", where every room has at least three entrances. It's a crude approach and not entirely necessary, but it's a good start to get you thinking about open layouts. Combine that with the prior paragraph about changing heights and avoiding having too many doors and you've got three simple rules of thumb that will quickly make a map that is interesting to explore.

For monster placement he had one rule of thumb that is probably safe to discount now (ensure a player never has more than 5 projectiles coming his way at once), which was intended for fairness, but he did also suggest two key techniques that are worth thinking of when you're designing maps. Basically, do you want a steady flow of enemies or a pulse? Mixing between the two also works, as does making up your own approach, but as a simple distinction it is remarkably helpful.

For instance, in an open map with many ways of egress, you could sprinkle it with monsters, have groups appear when certain areas are entered or objectives hit, or perhaps have one large wave advance on the player from one side. Personally, I'm a fan of sprinkling enemies around corridors and on ledges (or at distance) to keep up a steady threat, but then also hitting hard with groups at key moments. The odd gauntlet against an entrenched load of enemies is also fun, but generally speaking I consider monster placement to be all about a combination of background noise and crescendos, so I map appropriately. I'm always looking to have enemies surprise the player or have a vantage point, so I map those in. It's pretty basic stuff, but you do need to learn it.


Kind of blurring the line between technique and design there, but that's how I approach these things. I think taking an integrated view of mapping creates a stronger result than thinking of it as the battle between gameplay and visuals.

[mouldy]

While working on my megawad I completely changed the way that I map. What I do now is construct the basic shape of the rooms and corridors, and then get it functionally working with doors/lifts/etc and at that point I add the monsters and items, making the whole map playable while it still has only default doombuilder textures. I will then playtest it to death and get the gameplay finished before adding textures, details, lighting and decorations.

There is a reason for doing it this way, rather than finishing the visuals first. I find that if I am building a room at a time, fannying about with details and lighting without knowing exactly how the whole map will work, it can take as long as it takes and its easy to end up getting bored, which is not what you want when you have 32 maps to create. However, making a map basically functional and playable is a lot faster, and when its done you essentially have a finished map, which gives more motivation to complete the visuals. This way I can get a map done in a couple of days and move straight on to the next one, then after doing ten I can go back and make them all look pretty, which is a lot more fun when you know you don't have all the tags and monster placement to look forward to.

[40oz]

My mapping technique is constantly evolving as I challenge myself to do things I was never any good at.

[Obsidian]

Wing it.

[purist]

Phml said:

Starting with a sketch, like many. Although I find it interesting that my sketches don't look anything like the few I've seen posted on the forums. Those of you who shared such stuff tend to draw things neat, with accurate dimensions and precise shapes, it almost looks like an automap shot.

Mines are rough scribbles on graph paper, with lines overlapping each other as I have new ideas, line actions jotted down with a number next to the activator and a number next to the affected sector, different areas connected by letter codes as I run out of paper, monster types written out, sometimes entire paragraphs of random musings written on the sides and the occasional 3d doodle to try to think of cool architecture.

I'm guessing mine more resemble yours than the ones scanned onto the forums that you mention. I just have no desire to make my madman scribbles public viewing. My guess is that others with a similar style think the same and that is why only the neat and tidy ones get seen...

[Tristan]

Official recipe for ErisWads

1. Open GZDB.
   
2. Draw the start room. Ignore risk of 153° interior angles.
   
3. Detail this room, until it looks nice.
   
4. Continue to next area.
   
5. Attempt to build next area.
   
6. It looks like shit, delete it.
   
7. In fit of rage, delete the rest of the map
   
8. Repeat above steps, in a perpetual cycle inevitably resulting in nothing ever getting done.
   

[Breezeep]

I just use my (slightly handicapped) mind to come up with a level, or I would get some inspiration by playing someone else's level.