I had expected E1M8b and E1M4b to make news in the gaming scene, but I never could have expected how long of a tail DOOM had. Gaming, like all technology sectors, moves at the speed of light. New games come out and are forgotten in a heartbeat as players move on to the next new thing. DOOM fans, on the other hand, visit the old levels as if they are coming home.
In late 2017, DOOM fans were still uploading playthroughs of my recent DOOM maps to YouTube, still commenting on videos, still tweeting at me about them, and still asking questions about DOOM’s development and legacy. It is such an incredible and dedicated community. Amid all the comments and video uploads, some players posed a good question: “When will you make a full episode?” In DOOM, modders can build and release a single map, or they can release a pack of maps, called a megaWAD. Other fans went on doomworld.com and asked the same question in forum threads. It just so happened that I was being asked that question over and over in late 2017, when I realized DOOM’s twenty-fifth anniversary was one year away.
The timing seemed perfect. I decided to celebrate a quarter-century of DOOM by releasing a megaWAD. I could build a megaWAD in my spare time, when I wasn’t working on company projects.
There were lots of questions to answer. What did I want the megaWAD to be about? What should it look like? What theme and gameplay features should guide its design? At that moment, none of those questions had answers. There was a single idea: Make a new episode of DOOM—eight regular levels and a secret map.
As I thought about the episode, it started to take shape. E1M8b and E1M4b had been techbase levels because those fit the theme of episode one. Since I was making a brand-new episode, I wanted to design it using a theme I’d never used when we were making the game: hell. Sandy Petersen designed every map in episode three, “Inferno,” while I was laser-focused on building DoomEd—Sandy was fast! I’d dipped my toe into hell in E1M8b, painting the glowing cracks along the floor of the first room right through to the boss room where you fight the Bruiser Brothers. This time, I wanted to take you to hell and keep you there for nine levels.
Choosing hell as a theme helped me decide where to place the episode within the other episodes. Back in 1995, we made episode four, “Thy Flesh Consumed.” It’s by far the toughest chapter of DOOM, but I set out to change that. My episode would fit between “Thy Flesh Consumed” and DOOM II. I chose SIGIL as the title. A sigil is a symbol of magical power, and I wanted to extend that definition to serve as the theme of these levels. When I was thinking about where the episode could take place, I went back and played the final map of “Thy Flesh Consumed,” Unto the Cruel. I noticed that the final teleporter has some weird texture issues. That was something we didn’t catch before shipping The Ultimate DOOM. I wove that glitch into my story: When you step into it, it doesn’t transport you back to earth, where hell has invaded. Instead, a mysterious sigil causes the teleporter to malfunction, transporting you to the deepest, darkest, hottest pits of hell.
The thing about DOOM’s third and fourth episodes is that even though they take place in hell, the satanic imagery we used was kind of light. I wanted to push the envelope, so Brenda and I hunted for artwork that evoked Satan and eternal damnation. She found artist Christopher Lovell’s painting of Baphomet, a bipedal deity with the head, horns, and legs of a goat, and the wings of a demon. Lovell’s painting, particularly the colors and textures he used, would look right at home on the cover of an Ozzy Osbourne album from the ’80s. But the best part of it is a detail you might miss unless you look closely: In the center of Baphomet’s forehead is “666,” the mark of the beast, but the numbers are twisted so they almost resemble the symbol for atomic energy. It evokes terror, heavy metal, and pure evil—just the cocktail of elements I wanted to imbue SIGIL with. I worked Baphomet into my story: The teleporter in Unto the Cruel did not malfunction by coincidence. Baphomet sabotaged it by placing a sigil inside it.
We reached out to Christopher to see if the piece was available. As it turned out, he was a huge DOOM fan and was as excited about his piece being selected as I was about Brenda finding it! For the cover of the box, I decided I wanted the artwork to appear without a logo. I didn’t want anything to obscure Christopher’s work, which was, I felt, the perfect complement to what I had in mind for episode five. Everything was coming together.
The word “sigil” resonated with me in other ways, too. I liked that it was one word, like many of my other games: DOOM, Quake, Daikatana. It was also versatile. I could build an entire brand around it: SIGIL, the name of DOOM’s unofficial fifth episode, and then, later, SIGIL followed by a subtitle—a way to let fans know that what they were seeing was part of the SIGIL brand. In order for the brand to generate excitement, though, I needed more than a cool name and a wicked piece of cover art. I needed a fresh gameplay hook.
When SIGIL entered development, DOOM had been around for nearly a quarter century. Players had made tens of thousands of maps. Coming up with a style of gameplay no one had seen seemed impossible at first. Then I thought of the evil eye symbol we included in DOOM and DOOM II. It’s an eye floating in midair and set in the center of a pulsating green triangle pointing to the right, almost like an arrow, and ringed by a green circle. The artwork was awesome, but we didn’t use it much. That gave me an idea: What if the evil eye was Baphomet’s eye? It could be more than just a piece of décor. Normally, players make things happen in DOOM by opening doors, flipping switches, and walking over line triggers. However, there’s another progression mechanic in the DOOM engine called a gun-action trigger, and we didn’t use it much, either. The first instance can be found in the secret maze of computer terminals in E1M2. The player comes to a corridor with a light that fades in and out. Two imps guard a green armor vest, and shooting the wall behind the vest opens another secret area—a secret within a secret—that leads to the chainsaw.
My gameplay hook was coupling Baphomet’s eye and gun action trigger to make something happen when you shoot the eye. From a game design perspective, it’s important to teach the player something, and verify that they get it before moving them on to the next mechanic so they don’t get frustrated later on. In E5M1, Baphomet’s Demesne, I taught the player about the eye—the sigil alluded to in the megaWAD’s title—right away. The level has the player coming in hot: They start in the middle of a pentagram surrounded by imps and a spectre. The immediate task is to kill the demons and look around. There doesn’t seem to be a path forward. Then, the player hopefully notices an ammo clip near a wall with a window in the middle. Looking through that window reveals the sigil, and the ammo clip functions as a clue for what to do with it—I gave you bullets, so you need to use them. Players shoot the sigil, and the wall raises to reveal a cavern filled with lava and bordered by bodies and severed heads impaled on spikes. As they walk forward, part of the wall in front of them slides open to reveal another sigil. Players blast that and raise a bridge to their right that takes them further into the cavern.
The next couple of sigils are harder to find: They’re set back within walls and visible through vertical slits. The player has to position themself just right to shoot through them to hit the sigils. Each eye is a breadcrumb in a trail players follow to the exit, a massive symbol of Baphomet that ends the level when players touch it. By the time they enter E5M2, they are equipped with more than weapons—they have a firm grasp on how the eyes work.
E5M2, Sheol, builds on what players learned in the first level. As they play, they may notice an eye set back in a wall just enough that it’s easy to miss. Players don’t need to shoot it to finish the level, but if they do, it reveals a secret. Now they knew two things: that shooting the eye made something happen, and that something could be a way forward or a secret. Interestingly, the gun-action trigger only responds to weapons using bullets or shells. That means that players have to use the pistol, chain gun, or shotgun to activate them, which meant I had to give them a steady supply of bullets and shells. The goal was to balance each map with Ultra-Violence difficulty mode and pistol start in mind, the same goal used as we made the original DOOM. Because it was the fifth episode, I knew it had to be difficult, just as if the player had rolled into it from episode four in 1993. As I designed SIGIL, I updated a progression list detailing which maps would get certain weapons and which ones would have a weapon hidden in a secret area.
I was feeling devious when I designed E5M1, so I created a triple-nested secret—a secret within a secret within a secret—the first such sequence of that type I’d made. In summary, after shooting the first eye and entering the cavern, players shoot the wall dead ahead and raise two bridges (an obvious one in front and a secret one behind them).
E5M1 shows off another design hook. We designed DOOM’s levels to be playable in single-player as well as in deathmatch and co-op games. That’s why there are spawn points, weapons, and ammo that only appear in multiplayer. The problem is that some of those maps were too big for deathmatch. Shooting friends and strangers with rockets is fun, but having to hunt down an opponent between frags ruins the excitement. With SIGIL, I wanted each map to be a Trojan horse—a single-player level and a multiplayer level, but not the same terrain. Load any SIGIL map into an editor, and you’ll see that it’s two maps in one. There’s the area players access playing solo, and an entirely separate area built with multiplayer in mind. I came up with this idea on February 18, 2018, and was convinced I was the first DOOM mapmaker to do it—until I found REKKR.
REKKR is a megaWAD by Matthew Little, known as Revae, that replaces all four of The Ultimate DOOM’s episodes with four brand-new ones. The day REKKR came out, I was reading reviews and learned that Revae had built a separate deathmatch arena into each map. I couldn’t help laughing. REKKR ended up releasing first, and I gracefully surrendered the honorific “first” title. It made me happy to see others still trying to innovate within DOOM, too.
I needed to add something new to the deathmatch experience to see if I could one-up REKKR, and figured some players might like to play deathmatch in the single-player maps. I solved that by adding a way for players to connect the single- and multiplayer sections of each map as they played. In a deathmatch session, each player must flip two switches, which are set far apart, at the same time. After throwing the switches, both players must go through teleporters that transport them to the single-player section. This implementation forced both players to agree to expand the terrain where they’re playing. If one player wanted to stick to the deathmatch area, they just didn’t flip one of the switches. Also, knowing that some of those single-player areas would be too large for a good one-on-one deathmatch, I added some new walls to constrain the space a little on each map where needed.
E5M7 was SIGIL’s first designed map, and originally, I intended it to be its fifth. But by the time I finished with it, it was, to that point, the largest map I’d ever designed, and it ended up being the largest in the episode. It’s got a nest of secrets to discover, and it’s brutally difficult. To open the episode in E5M1, the first room I made was also not the starting room. It was the lava cavern players entered after shooting the sigil to leave the first room. Players were on bridges bordered by lava. If they fell in, they would have to scramble to climb out. Full of shotgunners and imps, players had to be mindful of their footing as they dodged fireballs, or they would strafe right off the bridge into the lava.
I approached the design of levels for SIGIL the same way I approached levels for DOOM, DOOM II, and The Ultimate DOOM. First, I imagined an opening scene. The first thing the player sees or hears in a level sets the tone and pacing for the rest of the level. Next, I created an area that players would visit multiple times—a landmark they would come to recognize as they opened up parts of the map and discovered how everything connected. Finally, I kept the design of each map tight. I don’t do sprawling levels. They get too chaotic and distracting. I want my levels to feel like roller coasters—high points of excitement broken up with exploratory lulls for the player to get their bearings. While I am making a level, as always, I play through it hundreds of times. Even the smallest change necessitates a replay to see how it feels.
I listen to music while I work. During the development of DOOM, I was at my keyboard with the tunes of Queensrÿche and Lynch Mob flowing into my ears. On DOOM II, it was Alice in Chains and Vince Guaraldi. While I made SIGIL, I listened to Buckethead’s music nonstop. Buckethead is one of the most creative musicians around, and I’ve been a fan since he started in the late ’80s. I mentioned to Brenda once that I wished I could use Buckethead’s music in SIGIL. Having listened to his music constantly throughout development, I just couldn’t imagine SIGIL sounding any different.
“Why not ask?” she said.
She had a point. The worst that could happen is he would say no, and that would be no different from where I was without his music. Why not? I sent him an email, and after a couple of months, his representative answered. It turned out Buckethead was a DOOM fan! This had happened once before, when Trent Reznor and Nine Inch Nails told us they played DOOM on their tour bus between shows and would love to write the soundtrack for Quake. Buckethead was gracious enough to let me license any songs I wanted, but the cherry on top was he wanted to write a brand-new song for SIGIL. I had no idea what it sounded like or what it would be called until the day I got a link to it. As a huge Buckethead fan, words fail to describe how I felt listening to a song that the virtuoso had composed for my game. Even writing about it now, it feels unreal. I was honored by his generosity and made his new track, “Romero One Mind Any Weapon,” the background music for E5M1. For the free version of the game, I worked with musician James Paddock. Paddock was already well known in the DOOM community, and his work is excellent. Like Buckethead, I selected scores from Paddock’s existing work that best fit the levels and the feeling I was going for.
Licensing to Buckethead’s music was one reason I needed to consider all the source ports of DOOM that have sprung up since John Carmack released the game’s code in the late ’90s. The larger reason was that SIGIL’s map design was incompatible with certain versions. All versions of DOOM can be sorted into two categories: those that adhere to the same rules and limits of The Ultimate DOOM, and those that have bells and whistles, such as allowing for more line segments. In DOOM parlance, this is known as limit removing—literally removing the limit of how many lines you can have in a level. Some older DOOM source ports support them, or let you enable or disable them, but many don’t. I released two versions of SIGIL, one for each generation. The “older” generation complies with the original game’s compatibility, such as a MIDI soundtrack by James Paddock, while the “new” generation’s release uses MP3s and supports added complexity in the map design.
Really, I was working on three WADs at once. There was the SIGIL mega-WAD, the music WAD, and another WAD called “SIGIL_compat” for the old generation. As an example of compatibility, the old generation of DOOM cannot register any episode beyond “Thy Flesh Consumed.” Running “SIGIL_compat” on any old-gen release will replace episode three, “Inferno,” with SIGIL. (You can still play “Inferno,” just not while you’re running the SIGIL megaWAD.)
There are all sorts of technical issues in DOOM’s engine. The only people who know about them are people who have analyzed the source code, and people like John Carmack and me, because we wrote it. That knowledge gave me the advantage of understanding DOOM’s tech at the deepest, lowest level. It’s like the difference between a high-level programming language like C and a low-level language like Assembly. If you’re fluent in Assembly, you can get down into the guts of a computer’s hardware and manipulate it in ways that add incredible amounts of performance to your software. Looking at DOOM’s source code reveals a hybrid of C and assembly language. That combination made DOOM fast, sleek, and cutting edge.
Another example of compatibility challenges can be found in E5M6, Unspeakable Persecution. The map’s defining feature—the landmark architecture—is a maze patrolled by a Cyberdemon. I’m thrilled with how the area came together. There are monsters all around the maze, and the player’s goal is to climb to higher ground above it while avoiding fireballs thrown by the Baron of Hell, camped in a nearby tower. The Cyberdemon stomps around the maze like the minotaur in Daedalus’s labyrinth from Greek mythology. It is utterly terrifying as a player to know the Cyberdemon is seeking you out while you try desperately to escape.
But as cool as the maze turned out, I also enjoyed crafting the area before it. Like many regions of SIGIL, E5M6 is a dark level—it is the depths of hell, after all. There are places so dark that the best way to fight monsters is to wait until they walk across a lighted area or a light source, such as a torch. The area before the maze is a river of blood that inflicts damage as the player wades through it. Three Barons of Hell teleport down there to hunt them, and the only way out for the player is by finding lifts that raise them onto safer terrain. Those lifts are deceptively complex. They ascend and descend, but they’re constructed with lots of vertices and line segments. The challenge was that DOOM’s engine was written to process a maximum of eight line segments intersected by the player’s hit box—a rectangle. That meant I had to simplify the line segments that made up those lifts, or they would be unusable.
I didn’t mind striving for compatibility at all. The reception to E1M8b and E1M4b assured me that millions of players welcomed more DOOM levels, and I was determined to meet their expectations.
Before releasing SIGIL, it needed to pass through its own gauntlet. Even though I’ve been making DOOM maps since the beginning, my maps are never perfect out of the gate. Every game goes through testing, and SIGIL was no different. Boris “dew” Klimeš, Xaser, Michael “UberGewei” Schaap, Flambeau, and Keyboard_Doomer were my play testers for SIGIL, and their feedback was invaluable.
Even before play testing, my process was iterative. Sometimes I’d come up with an idea for one map while I worked on another, or I’d think of a hook for a deathmatch arena that I felt was so cool I had to go back and make sure it was in all of them. In designing a game, I expect my early maps to be altered during the development process as I come to understand the aesthetic of an episode. It’s why I make my first level last. It’s the first thing players will see, and by then, hopefully, I’ll be at my best.
Most times, my iterations come from seeing the level as a player. I work on a level for a minute or two, play it, and explore how it feels. Designers can’t know how the adjusted hallway, or the change they made to a weapon’s damage, will affect the flow of their level and the player’s perception of the space or a game’s systems without testing it over and over and over, long before anyone else weighs in with feedback. So often during SIGIL I moved through areas before adding any monsters because I had to know how an area looked, how it felt to interact with and be in that space, before thinking about encounters. By the time a level was truly finished—no more adjustments, no more play testing—I had played it hundreds of times. I code the same way: code, test, code, test.
My launch plan was staged depending on what players wanted. As a collector of big-box PC games, I wanted SIGIL in a big box, too. It was for that reason that we had licensed Christopher Lovell’s Baphomet. I planned to release the megaWAD for free online with James Paddock’s excellent MIDI soundtrack. If people wanted the licensed Buckethead soundtrack, they had to purchase the version that included it. Many players knew SIGIL would be released for free, and most of them were content to wait for the free version. I knew that, too, and was fine with it.
Limited Run Games was enthusiastic about entering into a manufacturing and distribution relationship with us. Up to that point, they had specialized in curating physical editions of retro or indie console games and games that had only been published digitally. SIGIL would be their first PC game. We devised a roadmap that offered two physical versions of SIGIL. The Beast Box was positioned as a top-shelf item for collectors, featuring an oversize box inspired by the id Anthology box and boasting Christopher Lovell’s artwork. Each box was individually numbered and signed by hand. Inside the box there were some great collectibles, including a USB drive shaped like a 3.5" floppy disk, an art print signed by Christopher Lovell, a large pewter statue of my head on a spike as a wink and nod to DOOM II, a T-shirt with Baphomet on it, and lots more. The second box, called the Big Box, was much less expensive and came with fewer collectibles, but would satisfy any DOOM fan who experienced sticker shock at the Beast Box’s price tag. It also featured another of Christopher Lovell’s pieces on its cover.
We announced SIGIL on December 10, 2018, the day DOOM turned twenty-five. That same day, we announced a two-week window to order the Beast Box and Big Box physical editions. We sold thousands of boxes, making SIGIL an instant success for us and for Limited Run Games. On May 31, 2019, we finished manufacturing and shipped all the products to our customers. When everyone had it in their hands, I released the megaWAD for free online. It has since been downloaded more than a million times.
Fortunately, positive reviews and comments flooded in. Every fan who tweeted their review to me, streamed the game, or took the time to share their opinions about the game on forums meant the world to me. As Daikatana taught me, there are a million ways for things to go wrong, and I was truly feeling emotional seeing people say things like, “Romero’s still got it,” or “A typical Romero level,” followed by a discussion of how SIGIL was punishing and diabolical, both of which I took as high praise. A highlight was spending hours on Twitch watching players experience SIGIL for the first time. For game creators, the opportunity to watch your audience experience your product in real time is invaluable. I ended up altering a couple of things because of those plays, too. Among the first was the E5M4 crusher maze. It was too difficult for players and ruining their experience. So I made it easier. I also addressed most of speedrunner extraordinaire Zero Master’s exploits.
At the end of the year, I was honored when doomworld.com named SIGIL a runner-up for a 2019 Cacoward, annual awards given out to the honor the best DOOM levels, WADs, and total conversions released that year. In their write-up for the Cacoward, they described it as “the most anticipated, previewed, played, pored over, replayed, analyzed, praised, and shat-on release of the year.” In 2016, they had awarded E1M8b the best of the year. But the review that meant the most to me didn’t come from critics. After Bethesda and id Software rereleased DOOM and DOOM II in 2019, I got an email from Kevin Cloud. I hadn’t talked to him since QuakeCon in 2014, and it was great to catch up, but he wasn’t emailing just to say hello. I was floored when he explained that the team had curators looking for megaWADs to release as free content, and they would love to feature SIGIL. SIGIL’s availability through an official DOOM release felt like coming home. It had been welcomed as an unofficial/official fifth episode of DOOM, one twenty-five years in the making.
The release of SIGIL provided me with an education of sorts. The positive response let me know that people still had as much fun playing my levels as I did making them. Fans started asking for a sequel megaWAD to SIGIL right away, and I plan to release SIGIL II for DOOM in time for DOOM’s thirtieth birthday on December 10, 2023. A megaWAD for DOOM II is also underway, and I’ve already published a teaser. As a show of support to Ukraine following Russia’s invasion in 2022, I released One Humanity, a single-level WAD for DOOM II that is part of an upcoming thirty-two-map megaWAD. One Humanity costs five euros, and 100 percent of it goes toward the Red Cross and UN Central Emergency Response Fund. As of the date of this writing, just over €40,000 has been donated.
As for the rest of the upcoming megaWAD for DOOM II, I’m going to take a page from id Software’s playbook and say, “When it’s done.” All of these projects are made in my spare time, when I’m not designing my main project.
In July 2022, Romero Games announced that we partnered with a major publisher to develop my next first-person shooter. It’s not a megaWAD, it’s not Blackroom, and it’s not a retro-style shooter. It’s a modern shooter that pairs modern tech with a combination of old-school and contemporary design sensibilities.
First-person shooters have undergone major evolutions since we created and popularized the genre in the early 1990s. For most of that decade, the “FPS genre” did not exist, at least under that name. Everyone called them “DOOM clones,” and that’s what they were: imitators that looked to us as the pioneer in tech and game design. Now there are more subgenres than I can count: military sims, tactical, arena, battle royale, competitive, retro, multiplayer online battle arena (MOBA) shooters, hero shooters—the list goes on. Developers now have access to tools Carmack and I would have killed to have when we were getting id Software off the ground. AAA shooters have deep analytics that tell developers everything they could ever want to know about who’s playing their game, and how, and when, and what they like most and least. Live-service FPS games require around-the-clock maintenance and a constant flow of new stories, characters, weapons, and mechanics to keep players invested.
Today’s technology is light years ahead of what we had at our disposal in the ’90s. As technology advanced, player expectations rose. For an FPS to succeed, it has to go beyond the minimum effort of meeting those expectations. It has to do what id Software did with Commander Keen, Wolfenstein 3-D, DOOM, and Quake—it has to find and define a space.
When will my next shooter be released? When it’s done. Until then, the response to SIGIL has made one undeniable statement: The Icon of Sin is back, back to making games in the genre I pioneered, where I feel most at home, and where the community is the best. It will take more than a few rockets to the head—or a spike impaled through my brain—for me to leave.