History Lessons: Betrayal At Krondor (part 1)

New readers may wish to read the History Lessons Introduction first. The rest of the posts about Betrayal at Krondor are here: part 2, part 3, part 4. Previous History Lesson posts can be found here.

It’s been too long since my last History Lesson post. Nearly a year! Time to get back in the swing of it. This time I decided to tackle a game I’ve been meaning to play for some time: Betrayal at Krondor, originally released back in 1993. It is fondly remembered by fans as one of the earliest attempts at an open-world role-playing game, and for its strong ties to Raymond E. Feist’s Riftwar novels (Feist later novelized the game, officially accepting its events as canon in his fantasy world of Midkemia). Given the huge popularity of Skyrim and the other games in the Elder Scrolls series, known for their open-world role-playing design, I thought it would be interesting to look back at one of the first attempts at this type of game.

I actually tried to play Betrayal at Krondor a couple of times in the early ’90s. My first attempt was foiled by an insufficiently powerful computer, which could barely even load the game before crashing. Later, I borrowed a copy of the game from a friend to try on a newer machine, and managed to get a little ways into the game before hitting a game-stopping bug, probably due to some hardware incompatibility. In 1994, the game was re-released on CD-ROM (instead of 3 1/2″ floppy disks), but I never had a chance to try that version. Now, it’s conveniently for sale on GOG.com, bundled with its less popular semi-sequel, and pre-configured to run on modern machines using the DOSBox emulator. A good opportunity, then, to take another look.

The version provided by GOG runs fine, with one minor exception. Betrayal at Krondor uses a resolution of 320×200, a common resolution in those days (it would be several years before “high resolution” graphics using 640×480 resolution would appear in PC games). Displays at the time were cathode ray tube (CRT) monitors, and virtually all of them used a 4:3 aspect ratio. Since 320×200 is not a 4:3 resolution, it would display on the 4:3 screens with non-square pixels — something that was easily done with a CRT display but impossible on modern flatscreen displays which have fixed pixel sizes. Game designers and artists would therefore create the visuals based around these rectangular pixels. When running a game at 320×200 today, however, the default behavior is to keep the pixels square, resulting in a slight horizontal stretch of the image. This can been seen in the screenshots on the GOG.com page for Betrayal at Krondor; note that the circles around the character portraits at the bottom of some screens have been stretched horizontally, appearing as ovals. Fortunately, the makers of the DOSBox emulator anticipated this problem and included an easy fix. Simply open the configuration file in a text editor (for the GOG version, this is called dosboxBAK.conf and is found in the install directory), find the line that reads “aspect=false” and change it to “aspect=true”. This activates a built-in aspect ratio correction for 320×200 resolutions which GOG should have turned on by default.

That fix is all that’s needed to get the game running as it did originally. But I decided to apply a few tweaks to the audio, to fix some shortcomings of the CD-ROM version of the game. Be warned: I’m about to go into a mini-history of game music. If necessary you can skip the next few paragraphs and resume reading when I get to the point.

Betrayal at Krondor originally used MIDI music; that’s Musical Instrument Digital Interface. MIDI was invented in 1983 to provide a way for computers and synthesizers to communicate when composing or performing electronic music. The genius of MIDI is that it does not send any audio signals. Instead, it simply sends information about which notes are being played (e.g., which keys on a synthesizer are being pressed), and what sound they are supposed to be making (e.g., piano, trumpet, etc.). The actual sound is then created by whatever device is receiving the MIDI signal. Let’s say, for example, that a touring musician has two synthesizers she regularly uses. Without MIDI, she’d have to transport both of these heavy synthesizers to every gig, and play on each one with its own keyboard. With MIDI, she can take whichever one has the nicer keyboard, and swap the other one out for a rack synthesizer — a much smaller synthesizer that contains only the electronics, and no keyboard. The first synthesizer can send signals like “middle C, grand piano” to the rack synthesizer, which then creates the sound. Continue this train of thought and we realize that the musician only ever needs one actual keyboard, which can be used to “play” many other synthesizers through MIDI cables.

Image credit: Taken 1/31/05 by user Pretzelpaws (English Wikipedia) with a Canon EOS-10D camera. Cropped 2/9/05 using the GIMP.

The other big advantage of MIDI is that the actual information it transmits is very small. So if you were, say, composing music for a computer game, and you needed to fit the computer game onto a few floppy disks that could only hold 1.44 megabytes of data each, then you probably wouldn’t want to actually record a bunch of digital audio to include with the game. Why not simply write the music in MIDI? The disks would contain the information about which notes to play at which times, and the player’s sound card would actually generate the sound. Needless to say, this is how all game music was handled until the advent of the CD-ROM, which was large enough to store digital music. With MIDI music, however, the actual sound depends on the sound card used to play it. MIDI was standardized in 1991 through something called General MIDI to ensure that all MIDI devices would use the appropriate instruments and sounds when receiving MIDI data; a piano would always come out as a piano, instead of a trumpet or a guitar. Even so, different sound cards still created their own unique sound when playing MIDI. There were many popular options, including the AdLib and the Sound Blaster line, but one of the most highly regarded for MIDI playback was the Roland MT-32.

I never owned a Roland MT-32. Like many players, my first sound card was a Sound Blaster (I forget exactly which model; a Sound Blaster Pro perhaps?). Upon researching the MT-32, I was surprised to learn that it’s not really a sound card. Rather, it’s a full-blown synthesizer module (albeit the budget model in the line), intended for musicians. I was also surprised to learn that it actually used sample-based synthesis for the attack sounds, before switching to more traditional subtractive synthesis for the sustain. This is why it could produce far more convincing MIDI music than most sound cards, with (comparatively) realistic drums and string plucks. As a music synthesizer, however, it could not play back any pre-recorded digital audio, which meant a separate sound card was needed in order to hear the high-quality sound effects that were becoming common in games. This is where the Sound Blaster cards excelled, and they could play MIDI music too, even if it didn’t sound that great. Add a built-in gameport used to connect peripherals like joysticks and other controllers, and the Sound Blaster cards were obvious choices for most players. If they couldn’t afford a Roland MT-32 as well, then they simply had to settle for subpar MIDI music.

But then CD-ROMs arrived. With more than 650 megabytes of storage space, CD-ROMs made pre-recorded digital music a possibility. For the CD-ROM re-release of Betrayal at Krondor, the original music was played through a MIDI device (I believe it was a Roland MT-32), recorded, and then included on the disc as digital music. Now, everyone would hear the same, high-quality music regardless of which sound card they were using. Perfect, right? Well, not quite. It was early days for CD music, and the tracks feature some jarring pops and abrupt changes. With MIDI, it’s easy to program fadeouts or crossfades, but the CD audio simply stops abruptly. Worst of all, it doesn’t loop, leaving the longer in-game battles without any music at all. What to do? The answer lies in soundfonts. Soundfonts are exactly what they sound like: a set of sounds that can be applied for synthesizing music, much like a typecase font is a set of symbols that can be used to represent text. Some soundfonts are aimed at realism, upgrading a MIDI track from the dated General MIDI sound to something akin to a live orchestra. But others are aimed at recreating the sound of classic synthesizers like the Roland MT-32.

Setting up a Roland soundfont to use with Betrayal at Krondor is actually very easy. There’s a helpful thread with full instructions in the GOG.com forums, which I used as a guide. A (possibly) interesting sidenote is that the soundfont actually emulates the Roland SC-55, a later sound module that can in turn emulate the original Roland MT-32. In Betrayal at Krondor’s case, things are slightly more complicated because the original game did not support the use of two separate devices for MIDI music and digital sound effects. Fortunately, this was fixed with a fan-made patch, also discussed in the thread. The patch was updated since the thread was first posted, however, so you need to use updated instructions to apply it. Once everything was in place, the music quality was significantly improved over the standard Windows General MIDI, if not quite as nice as the CD audio. But it also faded and looped properly, and my sound effects were working perfectly as well. Excellent! Now, on to the actual game.

The first thing a player will notice when playing Betrayal at Krondor is the close ties to Raymond E. Feist’s novels. In fact, remembering that my friend’s knowledge of the books helped him find some secret rewards and otherwise enriched his experience with the game, I decided to read them myself before playing this time. If I’m honest, they’re not particularly great novels. Set in a universe ripped straight from Tolkien, complete with elves and dwarves in the standard mold, they feature the overused “farmboy grows up and saves the world” plot. They do have some interesting ideas, however, that place the world in a larger context and introduce some alternate cultures to shake things up a bit. There’s also a heavier focus on actual medieval societal structure and warfare, which sets the books apart from other high fantasy. By the end, the world is in a good state for a game to take place, and I was interested to see how Betrayal at Krondor would fare. If you’re thinking of checking out the novels yourself, the relevant ones are Magician (later split into Magician: Apprentice and Magician: Master in the United States), Silverthorn, and A Darkness at Sethanon. Feist would go on to write many more novels in the same setting, but these mostly take place after the events of Betrayal at Krondor.

Straight away, it’s clear that Betrayal at Krondor was intended to capture the feel of a classic fantasy novel. There’s an introduction in the game manual (and the manual is required reading, for a game from this era) by Raymond E. Feist, describing the challenge of marrying the two forms of media, and his full support for the game that resulted (it’s a common misconception that Feist wrote the game himself; he did consult with the team and granted his approval to the final product, but the story and dialogue were written by the development team). Then there’s a second introduction by the lead designer, confessing his love for Feist’s books and how he reached out to Feist to collaborate on a game that would capture the feel of those adventures. Both point out that most role-playing games, for all their debts to Tolkien, are very different experiences from fantasy books. Table-top Dungeons and Dragons, with one player acting as the Dungeon Master, could provide an experience closer to that of a book, but the computer games it inspired soon went down a different path. Full of statistics, hit points, and loot, they were more often games of combat and dungeon exploration with thin or nonexistent stories. Those that did have an emphasis on story could end up sounding bizarre: the player would usually ignore some world-threatening evil to gallivant through dungeons and caves filled with monsters and treasure with little justification for their existence. Characters began as pathetic wimps and ended up as demigods who single-handedly slaughtered armies of demons. That’s not the kind of story we find in a fantasy novel.

So, Betrayal at Krondor is different. The player does not create characters at the beginning; instead there are pre-made characters, some of whom appeared in Feist’s novels, with their own pre-set statistics and skills. This lets the writers give each character a distinct and consistent personality, which comes across in the game’s text. Text! In these days of fully voice-acted dialogue, written text in games is becoming a thing of the past. I think the last major role-playing game to rely heavily on text was Morrowind, and the switch to full voice-acting is at least part of why the sequels Oblivion and Skyrim fell short in the writing department. Text affords distinct advantages, not the least of which is that it’s much, much cheaper and faster to implement than recorded voices are. I appreciate well-voiced dialogue as much as anyone else, but for a large and complex role-playing game, once voices are recorded one can’t simply go back and re-write a scene. Not to mention that when games are voiced poorly, it can ruin what was otherwise good (or at least serviceable) writing. Imagine if all the resources and time that went into recording voices for modern games was instead all put into the writing? I think we’d see much more interesting games being made. Or at least much cheaper ones.

Then there’s the fact that text can fill roles that recorded voice cannot. It’s not only the character dialogue that’s shown as written text in Betrayal at Krondor — the game is absolutely bursting with flavor text. If I try to buy an item at a store when I don’t have enough money, instead of a simple “not enough gold” message I’m treated to a full paragraph describing how Locklear re-counted his coins and then sheepishly admitted to the shopkeeper that he didn’t have enough. If I search a body, I read about how Owyn gingerly turned the body over, trying to resist his revulsion as he looked for useful supplies. If I examine a sword, I’m told how Gorath took a few experimental swings with it, testing its balance and heft, before pronouncing it fit for combat. All of these are written in the style of Feist’s books.

The game systems are designed to mimic fantasy novels as well. There are no character levels; instead each character simply has a set of skills that increase with practice. Characters’ strength can increase too, but it’s a rare event, and their speed will never improve. If a character is slow, he’s slow. He won’t magically become faster just because he killed a bunch of goblins. In lieu of standard hit points, we have health and stamina ratings. Stamina is reduced first when characters take damage, but once it’s gone their health will take a hit. When that happens, their statistics and skills are reduced according to how much health they have left. If they only have half of their health remaining, then their strength, speed and skill ratings will all be half of their normal values. Being wounded has serious repercussions, just like in the books. Also like the books, players will spend the majority of their time traveling. The team must make camp at night, and be sure to stock enough rations to keep themselves fed. They also must care for their weapons and armor, keeping them in decent shape. Camping in the wilderness will never fully recover a character’s stamina, so the player is just as relieved as the characters are when they stumble across an inn. Combat happens frequently enough to be engaging, but not as often as many other role-playing games, and there’s always a decent reason for it. One of my favorite details is that if the team spots the enemies first, they can try to avoid them or attempt an ambush, with success depending on the character with the lowest stealth rating. A successful ambush catches the enemies off-guard, granting the player the first moves in combat and placing the enemies in poor formation.

Then of course, there’s the overarching story, of which I’ve seen very little so far. I’m still in Chapter 1 (there are nine Chapters) and my current goal is to take a poorly-disguised political refugee to the city of Krondor. A simple enough task, and the title of the game gives me a hint of what will happen when I get there. But I haven’t gone there yet, due to the other striking thing about the game: it has a fully open world. Today, if a development team wanted to make a game that mimicked the feel of an epic fantasy book, with a big focus on a gripping story, they would design a linear narrative filled with cinematic set-pieces and exciting battles. But back in 1993, such a restrictive narrative was out of the question. Role-playing games at the time were prized for their openness, and the team was not willing to sacrifice that. In fact, creating a freeform interactive experience that still featured an intricate and exciting plot was the main design challenge for the team, as discussed in the designer notes in the manual.

My memory from playing the game long ago was that each chapter took place in a small section of the world, with the player free to wander anywhere and do anything within. Advancing to a new chapter might open up new areas, and would also change the areas the player had already visited. Upon playing now, however, I’ve slowly come to a crazy realization: the chapters don’t just feature small sections of the game world. They each feature the entire world, with me free to wander the length of it if I please. OK, that’s a slight exaggeration; for starters, the game takes place only in the central part of the Kingdom, with the world of the books being larger than that, and I found that a few places were blocked off. But usually I was shocked to discover that I really was free to wander halfway across the world in the wrong direction if I wanted to. The game manual brags that there’s nothing stopping the player from exploring Dimwood Forest in the first chapter, and if the player returns in Chapter 2 there will be entirely new monsters there. I took up this challenge, heading into the woods when I should have been heading for Krondor. The forest was tough going, and far, far larger than I anticipated, but it was also full of treasure and a few mysteries. I’m sure I’ll return there later as part of the game’s main plotline to unravel them. When at last I’d scoured the entire place and left, packs bursting with loot, I tried a little experiment: instead of heading towards Krondor, I wondered if the game would let me go in the opposite direction, towards Sethanon? Well, it did, and I realized that I now had the entire eastern half of the Kingdom before me, promising days and days of exploration before I would eventually loop around to Krondor.

Ultimately, I decided to turn back. I want to keep some areas unexplored, so I’ll have new places to go when the game’s main story takes me there. Also, it just didn’t seem right to go running off when my mysterious companion needed to get to Krondor with the utmost urgency, guerrilla assassins hounding him at every turn. Maybe there’s some cool stuff in the eastern parts of the Kingdom that can only be found in the first chapter, but I doubt it. Besides, I want to see exactly what this “betrayal” I keep hearing about will be.

I’d like to go into more details about how the open world is designed, but this post is already really long, so I’ll leave that for part 2. I have a feeling there will be quite a few parts this time. Stay tuned!

Read the other parts of this series here: part 2, part 3, and part 4.