Voice acting is crucially different from other art assets in game dev, either visual or audio, in that there's no such thing as "off the shelf" voices.
Every project has its own dialogue, and if a developer wants to add voice to their game, it's an individualized, bespoke project. If you're a solo or small-team game creator, voice acting may seem out of your reach—and your budget—but enthusiast, semi-pro, and freelance voice acting has flourished in much the same way indie game development has.
Now, getting great voice acting in your game is often simply a matter of learning how to connect with freelance artists who are within your means.
Ashe Thurman (of Austin-based one-woman development studio Pixels and Pins) has experience on both sides of the mic, as a freelance voice actor and game developer of original English-language visual novels (OELVNs).
Here, she takes time out from development of her upcoming game, The D (Stands for Demon), and talks with Gamasutra about what she's learned from her own experiences commissioning enthusiast and freelance voice talent -- and how you might apply those learnings to your own small-team or solo-developed games.
Questions and answers have been edited for brevity and clarity.
Thurman: Any game can benefit from voice acting. It helps us connect to characters and can make settings feel more dynamic. Though, obviously, this applies to certain genres more than others. Narrative-heavy games (like visual novels and RPGs) definitely see a bigger boon in including voice acting over, say, a platformer or FPS. But it's not a sort of linchpin element. We've all played games without a lick of voice work, and enjoyed them just the same. It's also important to note that bad voice acting can ruin a game.
"The implementation of voice acting doesn't have to be an all-or-nothing approach."
The implementation of voice acting doesn't have to be an all-or-nothing approach. The voices should enhance—not get in the way of—the gameplay, and there's not really a one-size-fits-all answer to that. Play games in the genre you're working in that have voicework and see what people do, see what sounds right to you.
For puzzle or strategy-based games, having voiced cutscenes and unvoiced gameplay sounds fine. It's an easy way to break those portions of a game apart. In a more action-heavy game, having voiced cutscenes but not including voiced efforts during gameplay can sound strange and inorganic -- you're better off just having in-game voice efforts and carrying them into the cutscene. With visual novels (my developmental metier), it's not so easy to break up "gameplay" because it's just the same thing throughout the whole game. If you want to shoot for partial voice in those scenarios, you have to break it down thematically; a voiced scene in an otherwise unvoiced game is going to carry more narrative weight than other scenes.
As long as you maintain internal consistency and it doesn't feel "weird," you don't have to abide by some hard and fast rule. In my game The D (Stands for Demon), for example, I chose to use emotional barks, but only from the main love interests, and have certain pivotal narrative scenes between the female leads fully voiced. However, I was willing to pull the full voice acting scenes if they didn't work. It played well in the demo, so they're staying in. My first narrative game Amaranth was fully voiced. Despite it being a short game, it was an incredible amount of work. That's part of the reason I scaled back for The D.
Pixels and Pins' The D (Stands for Demon)
When to go the extra mile for voice work comes down to if the team can physically do it, and do it well. Including voice acting comes with casting actors, directing, line editing, voice implementation, and about a thousand other little nitpicky things you don't even think about until you're already knee deep in it. It is a gigantic amount of extra work. A lot of pro developers outsource it. There are entire studios and companies specifically centered around casting and recording for video games. That's how big of a task it can get.
You're only really ready for the undertaking if you have or can bring in someone (or multiple someones) that will be able to provide that combination of hard and soft skills, and you have the time in your development cycle to not half-ass it.
Something else super important to consider beyond just capability is budget. You and your team may be working "for free," so to speak, but you've gotta pay your voice actors if you're planning on putting out a paid product. If the money's not there, the money's not there.
Budget early, cast late. In theory, voice actors are some of the last people you should be bringing in, but by the nature of voice acting, you're generally going to be paying them upfront in lump sums. You have to make sure you have room for that in your budget from the very start. And if you're budget's going to inform how much voice acting you can actually hire, you'll have to make sure that's accounted for in your game design.
"Budget early, cast late."
Pro studios might offer between $200-300 per hour for a 1-2 hour minimum session, and it's not unreasonable to expect to be paid between $500-$700 as a main character in a fully voiced, full length visual novel.
At the indie scale, per line or even per word is the most common break down. [It's difficult] to put a hard number on it, but to give an idea I've done commercial work for $5 per 100 words. But that is the absolute lowest I'm willing to go, and that was only as part of a bigger project where the final payment was much higher.
Talk to fellow indie devs. DM a couple friendly looking VAs on twitter, explain what you're researching, and politely ask what their personal per word rates are. Things like Fiverr and People Per Hour, despite their downfalls, can at least give you an idea of what actors are willing to come to the table for.
Even in the OELVN market, you're going to have a hard time framing voice acting in and of itself as anything beyond just a neat feature. There are a lot of visual novel players, for example, who just turn off voices. Demos and trailers are gonna be spiffier sounding with voice acting, and the game will feel more professional, which will get more people interested, but that's more of a subconscious thing.
The way you want to use voice acting from a marketing standpoint, is as more of a hype machine. Your voice actors are going to promote the heck out of games they're in if you give them something to be excited about "Oh look at my character in this! She's so cute!" Casting announcements generate buzz. "Ohh! Who's gonna play the villain!? What are they gonna sound like!?" Then you're going to have the fans of specific voice actors who look into a game just because that voice actor is in it.
Then you're sort of looking at more of a long term marketing scheme. Having a voice creates a deeper connection to a character. That connection turns to devotion, accelerated, possibly, by that voice's pre-existing fandom. Devotion turns to obsession, and suddenly it's the only thing Tumblr's talking about. So it's not so much a matter of "hey this game is fully voiced" as it is "hey this game stars John Johnson as a green pigeon. Come listen to him make cooing sounds for ten hours."
There're are a handful of professional casting options, generally for a price you're not going to be able to afford. But, yeah, if you're new to casting and don't have a pool to choose from you're going to be doing a lot of just hoping and waiting. Finding voice actors on your own is a bit like fishing. Your project's the bait, and you throw it out there to see who takes a nibble.
Social media and word of mouth is your friend, in this regard. If a voice actor sees a neat project, they'll drop it in every discord server and group chat they have. They'll pass it on to their friends. Hopefully you've already got some kind of presence on Twitter, Tumblr, or Facebook, so posting a link to that casting call as a pdf or Google Doc is a good place to start.
You also want to go to the places where the voice actors are. The big three, so to speak, are voiceactingclub.com, Casting Call Club, and the casting call section of Behind the Voice Actors. They've all got their pros and cons, but these are all places a lot of really talented voice actors go to find projects that are hiring. There are a couple other sites like Newgrounds and Lemmasoft that have casting call areas, but I haven't found them particularly useful in recent years.
[Unless] you have the budget to pay for someone else to cast (which, let's be real, you probably don't), you're gonna need to be ready to sell the potential of your project. [Likewise,] I've seen a developer just post "VAs send me your demo reels, I'm casting for a thing," but I don't find that a particularly effective method. You have to have a big enough reach to get enough useful submissions, and it doesn't really tell you if an actor can nail the specific character you're looking for.
"You may feel an instinct to provide an exact example of the voice you're looking for. Don't."
Start with a casting call. In about two paragraphs you need to be able to describe your studio, describe the game (including possible explicit content), succinctly explain the amount and kind of voice work involved (eg, partial, full, efforts, singing, etc), how much you're paying, and mention your intended distribution platform. Also provide links to other completed projects.
After that be very specific but simple about the technical aspects of audition submissions. One file per character, 3 takes max, mono, delivered as an mp3, labeled "Actor Name-Character Name", and e-mailed to xyz with Subject "whatever" is pretty standard.
You want to make it easy on your actors to audition. They should be able to see everything they need in one place and not have to jump through hoops to submit for consideration. If you're drafting this in a document, all the above information should be a page, at most. This tells me right away, as an actor, who you are and how I can expect you to present yourself as a creator. That's your cover letter.
After that, in this same package, are the audition sides. A side is everything you need to audition for a character. It'll have a description of the character and lines that you would like your actor to perform to audition. A good character description will have maybe one or two sentences about the character, the vocal pitch (high, low, medium), the vocal quality (nasal, bassy, feminine, tomboy, etc.), and vocal age. Is this a teenager or an adult? Having art available is not required, but it is going to help considerably.
Then you need to pick your side lines very carefully, because you have 3-5 lines to encapsulate everything this character's going to have to be able to do. If this character's going to speak French, you better have a French audition line. If they're going to need to cry, include some crying.
You may feel an instinct to provide an exact example of the voice you're looking for. Don't. It doesn't make things easier for anybody and it tells an actor that you have a specific voice in mind, and nothing but an impression of that voice is going to work for you. That will stilt everybody involved creatively.
You're going to get what you pay for. The newer your actors, the more likely you're going to run into lower technical ability, lack of professionalism, and just an overall lower quality deliverable and experience. It's the same as any profession.
"You're going to get what you pay for."
Because it's not exceptionally difficult to acquire half-decent equipment to build your own home studio, the online voice acting community has become this kind of wonderful hodgepodge of pro, semi-pro, and amateur talent all mingled in together, and doing more than just games. I've done games, yes, but most of my work so far is actually in pre-lay animation [voice acting done before the voiced scene is visually complete]. I've also done audio dramas, ADR [post-production re-recording], and weird stuff that defies traditional labeling. And that applies to a lot of actors I know. Very few people I know are just a "video game" actor.
Generally, the audition itself will at least reveal new actors. They'll make very dumb technical mistakes like plosives [causing popping sounds from breath hitting the mic, usually from the letter P], peaking [by exceeding the maximum volume of their microphone], or their equipment just doesn't sound great.
[After that, the difference is] less immediately apparent. Number one, because it's really easy to hide lack of practical experience behind good technique and equipment, and number two, actors are just equally good. Talent does not reliably translate into booking gigs, so you can have people who are very good, who are just not landing roles. You also have the opposite problem.
General professionalism is going to end up being one of the biggest indicators of their experience level. Their communications are going to be short sweet and to the point, respecting your time as a producer or director. And most active actors have some kind of social media of some sort. So a hop into Google is going to pull that information up for you.
You can start by just listening to the auditions top to bottom, and you'll find the perfect voice. Then you'll find another perfect voice. Then you'll find a third perfect voice. It's easy to get overwhelmed, so breaking it down into parts helps.
I personally start by picking out any submissions that don't meet the technical requirements I'm looking for. If their mic is bad or if there're a lot of plosives or peaking or room noise or what have you. And those get put aside. They're not out completely, but they get pushed toward the bottom of the stack for consideration.
Next I look at overall vocal quality per character. Oh this voice too young, too old, too high, trying too hard to sound low, whatever, and those get put in the maybes. This is also when I might look at a voice fitting a different character better. This happened on my last game. An actor I sort of knew had auditioned for Greg, but when I listened, I thought, "feels like it could be a good Kaliocanthus voice" so I e-mailed him and asked him to audition for that part if he was up for it. He ended up being perfect and that's who I ended up casting as that character.
Then it's all about the acting, and experience is really the best asset to tackle that task. If I have a stoic character, and I get a read that's not stoic at all, they're not grasping that character, so it's gonna get kicked to the "maybes." I'll listen to individual line reads, and if they don't seem to be providing the emotions that I'm aiming for with the lines I gave, they'll get kicked to the "maybes" because I need to be able to direct someone. If they have trouble picking up the emotional cues I'm trying to give them, that might end up being a point of contention that gets in the way of the final performance.
Then once I have my top choices for each character, I play them all next to each other, and try to look at it holistically for my final decision. I had a situation with a project a couple years ago where I had two female leads that were very close in age and talked to each other a lot. I ended up using my second choice for the older girl because my first choice just sounded way too similar to the actor I wanted to use for the younger girl. So having that last overlook of the entire cast helps considerably to make everything feel like it fits together.
And at this point, I've tried very hard to not actually look who the actors are. I want to come in as unbiased as possible. It's not perfect because I'm going to recognize some people no matter what, but I do the best I can. I use experience as my tie breaker. If I have three voices, and they're all perfect, that's when I'm going to go look and see who the actors are and that'll influence my final decision. And honestly, someone I worked with before might also get a little extra consideration if they also gave me a good audition.
The biggest sticking point you'll face in voice compared to screen or stage is a super compressed timeline. You're gonna have a handful of sessions at most to slam all this work out. Usually only one or two. There's not time to sort of sit with a character and learn them or languish over their characterizations. I try to give out scripts ahead of time for an actor to flip through to get a feel for their character, at least, but I'm also ready to just get that worked out in-session.
So as a voice director your primary job becomes parring all this information down into something easily usable for your actor. There's all these acting methods—Stanislovski, Hagen, Spolin—that can introduce you to some very interesting vocabulary, but you can also just think of the whole experience as basically a conversation where you're working out a scene together.
"The biggest sticking point you'll face in voice compared to screen or stage is a super compressed timeline...there's not time to sort of sit with a character and learn them or languish over their characterizations."
You want to provide your actors with enough background to know what kind of world their character is living in, but that doesn't necessitate these huge lore bombs. They don't need to know about the Succession War of the 10 Dragons and the Knights of the Templar Order of St. What'shisface. "Magical Medieval Europe in a country torn apart by a hundred years of war" tells an actor pretty much everything they need to know initially about a setting. That gives a time, location, and genre reference right there. And reference existing properties if it's appropriate. It's not creatively bankrupt to take little shortcuts like that.
For each scene as you're developing a character through the narrative, you can just provide additional information as you go. Start with the basics of the character (race, gender, age, profession, general attitude, and their sort of overall through line) then as events occur, you offer how that event affected that character. How they feel about certain characters as they come up. Then just update that information in real time as you go.
You can also withhold information to get the response you want for a scene. Because of the nature of it, a lot of voice actors don't really bother trying to get the whole story until the full release. Use that to your advantage. One time I let an actress go a whole episode of an audio drama without telling her her sister was a vampire and two more episodes after that before giving her the script where her fiance betrays her. It creates nice, natural drama.
I also like to play "weird/not weird" with the immediate environment the character is in. This person being a wizard? Not weird. A person wearing glasses? Very weird. A man with a robot prosthetic? A little weird because they're usually too expensive for retail workers.
After providing all that information, I narratively lead them into a scene with something like "It's a sunny afternoon in your manor, but a courier's just delivered the message that your brother Eric has died," then sort of just let them go and show me what they've got. If I'm able, I personally like to do it as a sort of table read where I'll sort of half-act out the other side with the delivery I've either already gotten or am going to be looking for. I find that it creates a more natural flow, and lays the skeleton of the scene out really well.
We won't do the whole scene that way in one go. We'll section it out. Then we'll go back and tweak. I want this line a little more angry. Let me get an emphasis on the word "everyone" here. Okay, do it again, but let me hear with an emphasis on "we" this time. Oh this line gave us some trouble. Let's run it a few times. More. Push it harder. Harder. Louder. Add some crying. Okay pull it back just a little. Start the crying on "John" instead of "baby." That was good, but I lost the words in the crying there toward the end. Let's sharpen that up a little bit. And you just adjust adjust adjust and that's all intuition and practice and being able to hear what you want in your head and trust your actors. You definitely have to be ready to hear a read you weren't expecting and end up liking it better than what you initially planned.
One of the bigger things I've seen creators do is really hard to explain, but it basically boils down to not knowing what you want, or at least not knowing how to explain what you're looking for. It's one thing to be directing an actor and decide hey can we got a little higher or a little lighter or a little quicker paced. It's an entirely different thing to hire an actor for a standard midwestern accent, then suddenly say, oh no we want Russian instead, can you do that? If you know the character like you're supposed to as a creator, finding a voice for them shouldn't be a guessing game.
Situations arise where you can't live direct. Time zones. Schedules. Project sizes. A lot of things get in the way. You have a lot of trust in your actors and your writing at that point. You want to provide written action cues so they understand what's happening. You want to do what you can to mimic the kind of direction you'd give live as direction notes and know that you're just not going to be able to play with deliveries or maybe get it exactly the way you're hearing it in your head. You can ask for revisions, but there's gonna be a limit to how long that back and forth can before it's just frustrating for everyone.
And all of this comes with a sort of tacit understanding that there's going to be multiple takes, hopefully not too many, but that's so hard to know for sure ahead of time. It's pretty common practice when you're doing non live-directed work, to give 3-5 different reads, let the director pick one, then see if you need to do a retake. So there's a sort of general idea of how long it's going to take to nail something. You both don't want to be there forever, and if you picked your actors right, you're already going to be in good shape in that regard.
As long as you respect their work and their art, freelancers are generally super easy to work with. It's rare you're going to have major problems. If you do and you have them consistently, the freelancers might not be the problem.
"When money's involved, especially, get everything down on paper. That covers both your butts."
You do want to avoid misunderstandings. When money's involved, especially, get everything down on paper. That covers both your butts and prevents any a possible he said/she said situation from arising, which can happen. [Similarly,] be easy to get ahold of and take on the burden of communication. An actor shouldn't have to come ask you when they're going to get paid, for example. If there's a delay, tell them.
Sometimes you're going to run into an actor you just personally don't totally get along with, and it can create just this little tiny bit of friction. If you're all professional, though, you can push right past it, no problems. If you do get into a situation where you never want to work with that actor again, you don't have to.
You might also get into weird schedule conflicts that you don't know if you can blame on poor planning or not because emergencies do happen to people. I had an actress that I wanted to cast as a lead, and when I offered her the part, she said sure, but she was going to be having major dental work in a month and be out of commision for six weeks afterward. And this was an episodic project that still needed to finish off the last two episodes, so there was just no way to know if it was possible for her to record all that material in time. So I kind of apologized, and explained the situation, and asked if she would be up for playing a smaller part instead that I knew she could finish, and I never heard from he again after that. That kind of thing happens, and you just sort of have a backup. For that particular project, I able to cast my second choice (who I was gonna offer that small part to), and played the small part myself.
To start, a big, unfortunately common issue, is bringing in actors too early in development. Scripts aren't ready. It's still in pre-development. You don't even have all the main characters solidified. The longer you have to make an actor wait around not doing anything the less enthusiastic they get and the more nebulous your ability to actually finish the project looks.
Be upfront. Don't surprise an actor with the amount of work you expect or the presence of illicit content or suddenly recasting them. These are all things that have happened to myself or fellow actors.
And as sort of an addendum piece of advice regarding mature content in particular, be very explicit about the nature of the content right off the bat and what each character is going to have to perform. Actors have limits, and they need to know if the project they're auditioning for might have them do things they're uncomfortable with. Additionally, you can't have minors doing explicit sex scenes. Period. A teenage voice actor is generally going to know that, but it's your job as an adult/creator to double check.
When you hire me as a voice actor, remember that's all you've hired me to do. I'm not your audio engineer. I'm not your editor. I'm not your writer. I should be able to just send you a raw file, and a creator should be okay with that because they're going to do everything else on their end. I don't mind cleaning a file of dead space a little bit, or breaking it up into smaller parts. Those are very mechanically simple. But I don't want to have to mix things for you because you don't how to level audio.
Be the kind of creator that leads an actor to want to work with you and promote the work their doing with you. I've seen content creators that put out good work, but don't necessarily conduct themselves in a way that I'm particularly fond of, which makes me reticent to seek out working with them.