Hi, I was just watching your video of gameplay and I noticed something that has bugged me in a few games. And that’s the studio recorded sound. If you notice in a Movie they use a “boom” mic and this technique records actors voices but also because your using the mic positionally it also records the positional charecteristics of the scene. Meaning one actor may sound at a bit of a distance to the right and the other one sounds like they’re coming from the left and a thrid sounds like they are actually laying down. When you record video game narrative and don’t accurately rerepresent the scene everyone sounds like they are talking into a studio mic wihthout any positioning. Now you can add a convolution reverb to add a room or emulate a room but it’s not quite the same. Even if you were to record the positional sound in a wooden cabin you’d almost subtly hear the wood as opposed to say a cathedral. It’s these little things that make all the difference and I hope you consider this and it helps.
Thanks
Actually it’s pretty simple in theory to make it sound like a proper positional recording: All voices are monophonic recordings and if you place the source of the sound to the mouth of a person with the direction pointing outside the mouth, defining a range and angle, you can create a good positional playback, if you change the sound of the audio clip in relation to these parameters.
I know what you mean and it’s really disturbing if someone is yelling and the dry sound of the voice has a tiny reverb in it because of the reflections of the recording room. In general the recording is not the issue. Instead it’s the implementation.
5 Likes
First off all voices are not monophonic infact it is far more common to record the voices in stereo, even if they were there isn’t this magical software or plugin that easily postions sound or process or measuring distance and angles inputing these parameters and have a piece of software calculate the positioning in relation to this. It’s a great techy idea but not informed and the sound to the mouth of a person with the direction pointing outside of the mouth statement really goes to show you just sort of made this idea up. Like what are you talking about here? The sound source? (which is your mouth by the way) Or some mic that you stick inside of your mouth? (a mic is the sound source to the mix and your mouth is the sound source to the mic) A directional mic not an omnidirectional mic right? or something lol. Anyway if you record the audio from a “boom” mic like in the movies - ie a single sound source - All the positioning is done for you and its absolutely accurate and you don’t need to taint the track with excess processing keeping the signal clean and dry. Which actually is what you do normally - I know everyone hears the processing but when recording you want a clear unaltered sound source. And you’d be surprised at how the most expensive processing out there is designed to be transparent and not actually alter the sound characteristic or its balance so much as a “bringing it out” transparently. I could complain that is was overly compressed and equalized, that your trying to fabricate and manufactucre the sound charectistic as opposed to bring it out, but… so is everything you get from a below AAA producer lol It’s amazing all the charecteristic you’d recapture if you just bypassed these guys whole signal chain and subtly enhanced the master but anyway. ( And with the way sound has been trending lately these so called producers would be far better served if someone didn’t come up to them say WOW look how this can change and alter the sound but they instead told them Wo wait be careful this can actually change or alter your sound!)
I like the sound of the room and its not just the reverb of wood that has characteristic is also sound absorbtion whcih produces those little things you can’t just process or throw in the mix. Like think of how much people will spend for their instruments to made of specific types of wood grown in specific regions for this charecterstic. Anyway I don’t expect to see recording engineers pull out the tape measure with a protractor and use some processing to define to position when you can just mix it in the left channel 5.1 or .7.1 adjust the pan and volume. (Like the Waves S360 Surround Imager & Panner which uses imaging and distance panning to place sounds within a mix - but it’s not like you’ll be able to pan in a way that you can tell someone is lying down) But I thought I’d point out the absolutely accurate charactersitic of positioning you get with a single sound source boom mic like on a movie set and how recreating the scene when you record is the best way to achieve realism and would even a go a bit further and record this in a cabin or on a location and its technically very easy to do because the mic does it all for you. Position the Mic to where our charecter is and would hear the sound and position people and actors on the bed and around the room maybe even mix in a little close mic and you have the most realistic sound - processing can’t buy.or recreate
Voices are mostly recorded in mono, even in film production. For location sound in interiors mostly hypercardioid microphones are used, for exteriors it would be classical shotguns. Rarely I saw MS stereo shotguns to be used but mostly location sound guys try to get as dry signal as possible to give post production engineer most freedom in post processing. Boom operators can’t choose best place for the microphone anyway because of frame size, camera position, lightning in the scene or unwanted reflections. Most high budget films use ADR anyway. In studio only mono cardioid large diagraphm microphones are used (one at a time).
Sound perspective in film post production has two essential aspects when sticking to ADR: losing energy in different frequency bands differently as the sound travels trough space and of course reverbation. It becomes common practice location sound guys record convolution of particular space. Best possible sound positioning could be in theory achieved with recording sound with dummy head but for perfect reproduction it requires you to use headphones.
Conserning video games there are no standards and I noticed most dialogue in games is dry as hell. I’d like to go other way. We won’t use convolution reverb (although I record my own surround convolutions) but instead I’ll emulate sound perspective through dynamic EQ and reverbation (different EQ and reverb parameters for different distance between sound source and listener). It should be good enough :] I’d like to go as realistic as possible with the sound. All sound atmospheres that are mostly static were recorded in ORTF by me to achieve best first person immersion.
2 Likes
Source of the sound… sound source… Maybe I should have said “character” instead of person at that point.
You might want to take a look at how sound is recorded and how different it sounds:
It’s such a great sounding experience about fire you don’t wanna miss. And you might notice that the majority of mics are shotguns and all in different positions because you want to record the sound only in that position, not anywhere else.
Another nice how-to regarding sound recording is this:
Also many monophonic mics (except for the Zoom H4 he’s showing) and it all sounds great. To get high quality sound you need to record as much sound as possible and at the same time limit it to the sound you want to record as well which is I guess the most difficult part in sound recording.
This same quality is needed for recording one voice. That means you want to record the voice in its clarity, but you don’t want any other noises from that particular person, be it the rubbing noise of cloths as an example, or the noises of anybody and anything else surrounding the area of that voice.
I used a shotgun mic to record all my interviews at the Gamescom and I’ve made one big mistake when I recorded the interview with Daniel Vavra: I didn’t use a proper windscreen. Instead of his voice I basically had wind noises which ruined my whole recording. But some recordings in very noisy environments (the whole Gamescom is a noisy environment) were actually great because I pointed my shotgun to the interviewee and not to the unwanted sound. This is the first time I got to know how important directional mics are and how great their effect is.
Therefore in my opinion directional microphones with a narrow recording area are better for single sound sources like a voice. In games the positional audio playback relies on the implementation. And I already had my fun with FMOD Studio already, so I know at least the basics on how game audio gets implemented and how you can digitally process audio in relation to game parameters. That’s why I love game audio because you need to think of dynamic situations and environments where the same sounds needs to be processed via dynamic values.
LOL sorry I didn’t kow you were the sound guy my bad lol If your looking at sound perspective (to the listener through their speakers) there is this tool M360 Surround Manager & Mixdown. Using ORTF would create a static atmosphere as its designed to do so and cut down all room noise. I was going in the other direction and trying to capture the atmosphere, like where you said you don’t want to hear rufflng of clothes I would disagree and say that is what I want to hear and movement but then if you “close mic” it, all you’ll hear is bass. Distance is the key to a natural sound which is why boom micing on a movie sound like a clear voice not overly bassy like the guy narrating and action movie trailer, which forgive me is knida how the voice actors sound on the video. once again I’d use some distance, maybe go as far as getting an actor to lay down if thats what they’re doing on screen like you said a shotgun mic like on a movie set which is good because its very directional but being thats its placed overhead the sound is free to travel so you don’t get that excess bass.And like I said I might use a studio condensor in the position of where the character is supposed to be mix in a little close mic/shotugn but use the condensor for the main frame of reference to the character and positioning and provided you did this in a studio, or a cabin made to be the studio and there is reasonable distance from the mic and any of the actors one mic might even good enough or better. But like you said as well Wind is a killer and can ruin anything as delicate as that setup so then you shotgun boom mic too. I would kind of think of it like overheads on drums they’re used as a point of reference to the whole drum mix and maybe using an omnidirectional overhead mic would be better for capturing the room and atmosphere. Anyway alot can be said for recreating the scene as opposed to recording it all in ORTF, better immersion can be created by bringing out the subtle nuances of an actors voice which you’ll need better distance or overhead boom micing to ensure you don’t interfere with the sound as it travels.But if you stick a mic infront of someones face the sound will hit the mic and muffle up and you’ll get alot of base as oppsed to what they actaully sound like if you were to talk to them - thats why movies use boom mics and rappers stand back and sing over the studio condensor and popstars without a voice lip the mic basically lol But sorry afgain didn’t know you were the sound guy here lol I just thought an apporach like this would be worth considering for a game focusing on realism
@Ndustry, so far in just the alpha this game sounds better then any other Ive heard. Its like ear sex.
I would add actual input but I really don’t know anything about sound recording.
@Ndustry I’m not the sound guy, but @NLY is and there are two posts 
However I’m interested in getting into game audio. I might not be a professional yet as I lack of knowledge and experience. But I have the passion and interest for this area.
Regarding the bass in the voice: It is true what you’re saying about the mics’ behavior when you’re very close. My Sennheiser E845 has a similar behavior and since I don’t have such a deep voice it can have a quite positive effect because in radio broadcasting some of the bass frequencies get cut. Still all depends on your microphone and voice in general. Not every microphone works well with every voice and not every distance to the microphone works with every voice either like you said yourself. But I understand your quality standards and @NLY gets some feedback from me on the sound already, what I like and don’t like about the current implementation.
I’m going to visit Warhorse Studios in May and I’ll see what kind of experience I get from all those guys. I’m interested in their mocap studio, since it’s also used to create scenes for the game and I’m wondering if sound recording will be involved in this process as well. Then I’ll see what else the sound guys have to offer. And I hope I don’t distract them too much from their actual work. 
@McWonderBeastEarl Thanks. It means a lot to me.
@YuusouAmazingKing It’s not always true that shotgun microphones are better for voice capturing. Generally the narrower mic directivity you have the bassier will the voice sound at close range. It also picks portion on sound from behind of the mic. Therefore it’s common practice to use hypercardioid (something inbetween shotguns and cardioids) microphones indoor and cardioids in the studio where you know you won’t get much reflections. I always use shotgun mics when recording exterior voice.
I like the video where the guy records fire sounds, I’m also planning recording session with a torch. There’s good reason he’s mainly using shotguns here: Resulting sound is bassy and it’s a good thing here - no one likes thin fire sound :] He knows fire sound is quiet enough to be recorded in atelier like this one. Notice strong reverb when he talks, he wouldn’t record decent dry voice there I think.
I don’t think I’ll record anything in our mocap studio, it’s far from being an ideal accoustic space. But I’m looking forward to your visit here, it can be positive experience for both of us!
@Ndustry I wouldn’t ever use ORTF to record single voice, especially in studio. I wasn’t clear on this one, sorry. I’m using ORTF to record outside atmosphere like forrest, meadow etc.
I actually kinda like proximity effect as we hear it on all those voiceovers nowadays. It’s not very ‘natural’ but we’re used to it. And if you don’t like it you can still cut or suppress low frequencies in the post.
I can’t agree with you on reason why they use boom mics for movies. Shotgun mic will generally bring you even more bass at close range than any other microphone. Shotgun microphones are great for dealing with exterior (and some interior) unwanted noises but they’re far from sounding natural. Look at their frequency responses at different angles, it’s a rollercoaster. Most ‘natural’ microphone is generally an omni microphone. That’s why they mainly use omni microphones when they need to capture sound as it is at particular point in space - typically when you’re recording orchestral music.
1 Like
That’s why I use my vocal microphone, which is supercardioid. It doesn’t record that much from the back and since I’m more or less speaking to an empty wall, I’ll benefit from this directional characteristic. For recording outside or in noisy environments I’m using the shotgun. As said I don’t have much experience in different recording environments, but I’ll read up on that. Also I don’t have that many different mics either