Hello, everyone. Chances are, you're here from YouTube and are trying to learn how to use Face Poser, the tool in the Source SDK. I created this site because I end up answering a lot of the same questions, and since my video tutorials apparently don't cover every subject perfectly, I thought I'd address some of the more common problems here. While I know basic HTML, I'm no whizz, so leave me alone unless you want to design the website for me (and your own website looks absolutely nothing like MySpace) I'll add questions as they come in. If you have a Face Poser question for me, E-mail me at katananokage (at) Google mail.

Q: My extraction always gives a red/yellow error! What can I do?

A: If your error is red, it means the extraction failed outright. This may mean that you don't have the Microsoft Speech SDK installed, or you're running Windows Vista. To install the Speech SDK, visit this link , scroll down, and click the download for SpeechSDK51.exe. This is a Winzip self-extractor, so extract it anywhere, then double-click the setup file inside. Windows Vista, however, uses a newer Speech SDK and is incompatible with the old one. Unfortunately, there is no word on when Valve intends to update Face Poser to use this new SDK. However, you can still synchronize your phonemes manually, and I'd like to note that Ross Scott, creator of the Civil Protection series, uses this method for all his lipsynching.

However, if your error is yellow, it either means that the audio is too low quality, or has too much background noise (see Tutorial 3 if you're getting it from a source other than a voice actor, such as a song). You can also try using more "phoneme-literal" spelling of the speech text you enter. Remember things like PH, and the silent E aren't read how they sound. (but don't overuse these unless you're getting problems)

If you're running Windows Vista, or you just can't get auto-extraction working (or it doesn't work well for your sound in particular) you can still manually synch the audio, and I'd like to note that for reasons of precision, some professionals choose this method. Here's a very good tutorial on manual extraction at ModDB.

Q: My audio is 0 seconds long when I put it into choreography!

A: This almost always means that it can't find your audio file at all; this problem is not related to bitrate, lipsynching issues, etc. Face Poser will always attempt to find your audio in the "sound" directory of the game you're working in. If you selected Half-Life 2 from the SDK menu, then your sound should be in yourusername/Half-Life 2/hl2/sound. If you selected Half-Life 2: Episode Two, then your sound should be in yourusername/Half-Life 2 Episode Two/ep2/sound. (etc) 

Additionally, be sure that you entered your sound name correctly. As long as your sound is in the root "sound" directory, you do not need to specify any folder names. Also, be sure you are entering your sound as a file underneath the big list shown in the Wav File dialog; your sound will not appear in this list unless you have entered it into the big sound manifest TXT (not necessary for basic Face Poser or machinima, and not covered in my tutorials)

Q: How do I get my scenes into Garry's Mod?

A: Face Poser ingrains itself into the input/output of a map, so this is difficult to answer. The short answer is that you look at an NPC, and enter the command: 

ent_fire !picker setexpressionoverride "scenename.vcd"

This is a debug command used to see how scenes look by developers. However, I explicitly leave this out of my tutorials because I want to entirely discourage its use. It breaks many of Face Poser's features such as expressions and Look At commands, and causes them to glance around as they speak (something normally eliminated with "Look At !self")

Q: So, what? You want me to make a map for every scene that I make?

A: There are numerous possibilities on how to run your scene. I'll try to list them out here.

1. You can make a custom map for your scene, yes. If you're a good mapper, this isn't a problem. Just be sure you have some good lighting.

2. You can record directly from Face Poser, using Fraps. Just expand the 3D view to take up almost all of the window before you record. In the Options menu, you can change the background color, lighting color, and even ground color, which could be very useful if you want a certain effect, or want to try greenscreening it into your machinima. The audio is a bit bad, though, so you'll probably want to edit the original audio over the recorded.

3. You can use a "greenscreen map" to record your faceposer stuff, then put it into a screenshot of a HL2 scene. For that matter, you can greenscreen it anywhere, though it probably won't be as convincing. I'll try to remember to put up my own greenscreen VMF here for everyone to use; it's a simple box with a button on the side.

4. If you really want to get your characters into a HL2 area, and the camera needs to be moving, the lighting must be good, and greenscreen isn't an option, you aren't out of luck. Download a program called VMEX and decompile the map, then open it and put a logic_choroegraphed_scene entity in, a button to trigger it, and the necessary NPCs. This takes a little more effort and requires you to tweak with Valve's existing maps, but can definitely be accomplished.

Q: My TF2 scene worked ok, but now I can't hear any taunts or voice commands in other servers!

A: This one's my fault for not mentioning this in the tutorial. When you recreate scenes.image, it only includes scene files that exist in your tf/scenes folder. Since the game's own scenes aren't there, it'll miss them and only take yours. To remedy this, you'll need to put the game's scenes in your scenes folder. Use GCFscape, open team fortress content.gcf and extract the "scenes" folder to yourusername/team fortress2/tf.

Q: Where do I get Face Poser?

A: Click the Tools tab in Steam, then install Source SDK. Open it, and click Face Poser.

Q: I don't have Steam, but I have Half-Life 2. Where do I-

A: Fuck off, pirate.

 Minor/Random questions

Q: What did you type at the end of the second tutorial?

A: weapon_knife (not an actual weapon, just a joke)

Q: What map is that in your TF2 tutorial/other videos?

A: ctf_bridgetoofar. It's a Mine-Dog.net exclusive map, so visit their server if you want to play it.

Q: What's that at 0:48 in the PC vs Mac video?

A: If you looked at the description, comments, or any piece of technology outside of an XBox, you would know.

Q: Okay, but isn't that a bit random?

A: No it isn't. Pay a bit more attention to the dialog.