I’m obsessed with the use of voice in games. I find it one of the most fascinating aspects of the medium. Most related discussions you might encounter, certainly as an audio practitioner, focus on processes relating to getting a good acting performance and how to best capture it – casting, auditions, directing, ensemble performances, mo-cap, union versus non-union talent (in the USA at least), which microphone is the best for a particular kind of sound and other tricks of the trade. Then there are the production and in-game implementation issues of managing tens or even hundreds of thousands of assets, localisation fun, batch processing of files, the pros and cons of temp dialogue, real-time versus off-line manipulation and the joys of synthetic speech. All of which are important, none of which address the fundamental issue of why most game dialogue is excruciatingly painful to behold.
There’s been an obvious improvement in the general standard of writing and acting in games over the past decade. And as time goes on this can only get better – actors, directors and home-grown writers who understand the medium and the unique challenges it presents will naturally continue to hone their craft. But there are a couple of stumbling blocks:
- There’s a general trend for more dialogue.
- Integration between story/writing and game design is minimal or non-existent in most games and their development processes.
This first issue is just so face-palmingly stupid. If we want the quality of the experience to go up we need less dialogue. Fact. If you don’t believe me just watch this:
I get the marketing polemic here. Fine, whatever. But which comic genius thought it would be a good idea to start this video with a performance by one of the 20th century’s finest actors? This only serves to highlight how bad all the dialogue is in this bizarrely awesome train wreck of a video.
The second stumbling block is even more of a challenge. There’s only so far we can take this with “better acting” and “better writing” – the biggest issue is integrating these aspects with the game design itself which needs to use voice more appropriately instead of relying on it to communicate anything and everything. Whilst dialogue and its performance should of course be as awesome as possible, it is the role it plays and the information that it is burdened to convey which seals its fate.
Despite my criticism of Bioware/Lucas Arts’ approach in the voice abhorrence that will be Star Wars: The Old Republic, there are glimmers of hope to be found in Bioware’s other products. Dragon Age: Origins mesmerised me for a week over the past Christmas holiday period. Mrs. Kenny was working so I was able to indulge in some epic RPG action, which isn’t something I normally have the time for – I refuse to play this kind of game in small chunks and much prefer being totally engrossed. It was totally sweet, sitting there in my pants, refusing to wash unless absolutely necessary. Eating and sleeping became thoroughly irritating necessities. The only way I can justify this to myself is that I know the experience will eventually come to an end, and I can pick up my real life from my last save game. It’s for this reason that I’ve never played World of Warcraft – the idea of a never ending RPG gives me the fear. Looking past the fact that the Dwarves have American accents (WTF?!), some of my favourite experiences in Dragon Age were driven by its use (or lack of use) of voice.
In contrast, I found Mass Effect, Bioware’s previous IP, to be a bit of a chore – all RPGs are intrinsically a bit grindy for sure, but Mass Effect didn’t give me a story or universe I wanted to see more of and the user interface didn’t make life any easier during combat. As soon as I started playing Dragon Age another thing became dazzlingly clear – the dialogue tree system in Mass Effect (which they’ve retained for Mass Effect 2 and SW:ToR) is not to my tastes. The way it works is it presents you with a list of general directions/responses your character can take and, once you have made your selection, you listen to what your character has to say along those lines. I have a few problems with this:
- Frequently, my character will say something which I categorically had no intention whatsoever for them to say, in a way which just doesn’t suit the character I’m trying to be. I’ve been forced to choose from a small selection of directions which are compromised abstractions, the result being frustration with my character and the game.
- I’ve got to listen to the mouthy bugger, and if I skip this I have no idea what they’ve just said because of the limitations of the aforementioned abstractions which are only vaguely representative of my character’s actual response and not the entirety of the rambling speech he then goes on to make.
- I am my character (this is an RPG, no?), so why do they do things and say things which I have little control over, and know a whole bunch of stuff which I don’t? I mean, I’m meant to be them, but I’m having it rammed down my throat that I’m quite clearly not them. They are themselves more than I am them. If that’s what I was looking for I’d watch a film, a really good film that has a century-long legacy of perfecting this kind of storytelling.
- In summary – why give me a choice, the illusion of control, only to immediately remind me who’s really in charge? I don’t get this kind of frustration, certainly not to the same degree, playing a game with purely linear cutscenes.
Dragon Age, however, uses a different system whereby you are presented with several verbatim options for what your character could say and, then, as soon as you click on one of these phrases it is as if your character has already said it and you immediately hear and see the other party’s response. This works beautifully for several reasons:
- Having read all the options, considered whether it fits with the character you have established and any potential outcomes, there is no need for you to hear your character speak this information out loud again (a trap fallen in to by earlier games, such as Ion Storm’s Deus Ex) because you’ve already just “heard” it in your head when reading it. And so, the act of clicking replaces the act of speaking.
- To highlight how awesome this is, compare it to what happens when you select an action for your character to perform rather than a phrase to speak – you generally have to watch your character perform the action. Why? Because if you didn’t see your character perform the action and yet you instantly saw the results of said action, this discontinuity would require a mechanism which explains the passage of time. But we know our character has said something aloud when the other characters present respond appropriately to our chosen selection, so there is therefore no need to hear it – it has clearly already been said. It’s as if the time spent reading your options replaces the time spent talking and communicating your thoughts to the other parties.
- If your character were to speak out loud, ignoring the redundancy of hearing it all again, who’s voice is this we are hearing? It certainly isn’t mine or my character’s – it’s some poor bugger who’s been in a recording studio for weeks, where everyone in the recording session has zoned out because it’s the end of another long day of the same monotonous pap, and the director has long since given up trying to get every line perfect. There isn’t even the time for that, never mind the will. And it’s not that the character is mute – this is not the same as Gordon Freeman, the silent protagonist of the Half-Life series, where the player is never given the option to “speak” – it’s that this interface paradigm bypasses the need to hear the character speak. But similar to Gordon Freeman, by not hearing a prescribed character voice the player isn’t bumped out of the experience and is empowered to fully inhabit their character.
- The experience becomes less about communicating information via voice, and more about communicating via the written word. This opens the door to a whole new world of immersive experiences that voice and dialogue can never get even remotely close to. You can certainly get quite close using sound and the moving image, with judicious use of voice, but you will never have the time or a big enough team to realise this in a game the size and scope of Dragon Age.
- Less time and money needs to be spent on voice records and localisation. And the experience is better! Low hanging fruit or what?
So, in playing, or trying to play, Mass Effect and then coming to Dragon Age, the differences in these two systems was made so abundantly clear that Dragon Age felt like it had invented a brand new, revolutionary interface paradigm. In reality, this is the same old tried and tested dialogue tree system of Neverwinter Nights and Knights of the Old Republic. So why did Bioware change this for the Mass Effect series of games?
Well, there are things that you can do in Mass Effect that you could never do in Dragon Age – talking to a group of people for example. Check out this scene (taking note of the interactive music which manages to score the scene rather nicely despite the dialogue trees doing their best to mess up the timing):
Pulling off this interactive cinematic in Dragon Age would be impossible given that your character spends no time speaking – this scene is all about your crew focussing on what you have to say, being inspired and motivated (it’s also about motivating the player, which is interesting given that it’s the player’s character doing all the motivating). So, perhaps the story the writers of Mass Effect wanted to tell couldn’t be done using the standard Bioware model. It’s just a shame that the method they’ve settled upon has knock-ons which make the experience less personal, and more frustrating for me as a player. I’m really interested to see what model their next new IP ends up using.
Whilst Dragon Age is more to my tastes it’s not perfection incarnate. For example, there’s some contradictory behaviour. At the beginning of the game you create your character by choosing their sex, race, specialisation, class (or caste), appearance and voice. This voice is heard when your character is engaged in combat, which is useful because these vocalisations can be a really important aural tool which supports the animations for fighting, exertion and death (however, there are generally enough sounds going on during a fight that these events could arguably be described as surplice to requirements). But what I find particularly inappropriate is that when you move your character or send them to pick up an item they will sometimes respond with an “as you say” or “your wish is my command” type phrase. This is odd – I am the character, not a 3rd party, so why are they addressing me as if this is a commander-led situation as in a Real-Time Strategy or God game? This identity crisis is made especially apparent precisely because you never hear your character speak at any other time. Interestingly, all the Non-Player Characters in your party will respond in a similar way when you give them such an instruction – this makes sense, because it’s as if they are responding to an “unspoken” command issued to them by your character (i.e. you).
One aspect of Dragon Age which I really enjoyed was those rare occasions (I only recall a couple of them) where voice was dropped altogether in favour of communicating what was going on via text. Check this out:
I don’t know how that views/feels for you as a non-interactive experience (my impatient clicking certainly makes it quite hard to read all the text – sorry!) but for me this was a most magical encounter. I mean, just compare the special feelings generated when reading a line like:
“The Presence in the gem is at first alarmed when it senses your touch. It recoils in fear, and the images that rush through your mind are ones of imprisonment and loneliness”
…with the awkward spoken dialogue and exposition near the beginning of the clip:
“Is that blood in there? Who’s I wonder? You’d think it would be all dried up after so long. There must be magic involved!”
Was this dialogue meant to be Scooby-Doo bad as part of some in-joke at Bioware that they’d all rather be making intelligent text adventures than spoon-fed talkies with hastily written and recorded dialogue that treats the audience for their adult rated games like children? Who knows? But I found this “text adventure” sequence magical – the voices, sounds and images I experienced whilst reading about and interacting with that spirit were better than anything else I found in the entire game. Point is, this wasn’t just a simple text adventure – interacting with the visuals, music, sound (that ambiguous yet suggestive whispering voice!) and text all added up to something which cannot be experienced in any other medium (including most games).
And whilst text is rarely used for story or exposition, as in the example above, Dragon Age isn’t shy in using text descriptors to heighten the experience in a rather subtle way. For example, the game is awash with “Ominous Doors”, none of which look remotely ominous in the slightest, but their text descriptors do a fantastic job of hinting at what might lie beyond, allowing your brain to fill in the blanks and make the experience more sophisticated than what mere pixels can convey. That is a powerful tool. And Bioware seem to know it – Dragon Age is packed full of information which is locked away in the form of text inside ancient books and scrolls.
So if Bioware aren’t shy in using text all over their experiences, thus making the “oh, people don’t read things”, “people don’t want to read”, “lots of people can’t read” argument somewhat weak, when are they going to do the right thing and cut back on their crummy spoken dialogue so that whatever voice acting they do have stands a chance of being totally awesome and contributes more to the experience of playing one of their games than it does take away? I’d just luuuuuuuuuuuuuuuuuuuuuurve to see the telemetry for how much of their dialogue gets skipped. Go on. I mean, prove me wrong kids, prove me wrong…
I’d hate to give the impression that Dragon Age is devoid of interesting, or more sophisticated uses of voice. In the ‘Anvil of the Void’ quest the character of Hespith, encountered in the Halls of Bownammar in the Dead Trenches, offers up one of the most intriguing examples in the game. As you explore the Halls you hear her chanting a dark and disturbing verse:
First day, they come and catch everyone.
Second day, they beat us and eat some for meat.
Third day, the men are all gnawed on again.
Fourth day, we wait and fear for our fate.
Fifth day, they return and it’s another girl’s turn.
Sixth day, her screams we hear in our dreams.
Seventh day, she grew as in her mouth they spew.
Eighth day, we hated as she is violated.
Ninth day, she grins and devours her kin.
Now she does feast, as she’s become the beast.
All of which sets the grisly, eerie mood rather nicely. The deeper you delve, the more of the verse you hear which gives you a palpable sense of approaching doom (i.e. a kick-ass boss fight). Now, throughout Dragon Age you encounter many demons, the voices of which have been processed with the prerequisite reversed-reverb effect to make them sound suitably otherworldly. And yet here we have a spirit who you can hear speaking through solid rock and locked doors from several hundred yards away and her voice is totally unaffected or processed. The cool thing is it doesn’t need an effect – a disembodied voice chanting such disturbing material gives it an otherworldliness that all the fancy-pants processing in the world cannot impart on any source material.
I say this all the time, but I think it’s worth emphasising and reiterating – the use of a sound/voice is just as important (perhaps more so) than the sound/voice itself.