About this video

Demo of the Voiceover In-Action


Here's a break down of the steps in the video:

Finding Dialogue Topic and NPC
1. Open .esp in TES5Edit
2. Browse for Topic to voice
3. Ctrl-Click to NPC
4. Copy down Editor ID (eg. EtC_Vampire1) and Voice Type (eg. FemaleYoungEager)

Find Dialogue and Voicetype
1. Open .esp in Creation Kit
2. Click Actor and open NPC with the editor ID you found in step 4
3. Click the Dialogue button
4. Click Export Dialogue to created exported dialogue .txt file in skyrim folder (eg. dialogueExportEtC_Vampire1.txt)

Filtering the Exported Dialogue
1. File is formated with tab seperated columns
cat dialogueExportEtC_Vampire1.txt
2. Filter to just columns 16 and 21
cut dialogueExportEtC_Vampire1.txt -f16,21
3. Filter to just lines with the .esp
cut dialogueExportEtC_Vampire1.txt -f16,21 | grep "Escape the Cell.esp"
4. Filter to just lines Relevant to the topics you wanted
cut dialogueExportEtC_Vampire1.txt -f16,21 | grep "Escape the Cell.esp" | grep Vampire
5. Remove the file path
cut dialogueExportEtC_Vampire1.txt -f16,21 | grep "Escape the Cell.esp" | grep Vampire | sed 's/Data.\+\\//'
6. Replace .wxm with .wav
cut dialogueExportEtC_Vampire1.txt -f16,21 | grep "Escape the Cell.esp" | grep Vampire | sed 's/Data.\+\\//' | sed 's/\.xwm/.wav/'
7. Save resulting output to file (eg. vampire.csv)
cut dialogueExportEtC_Vampire1.txt -f16,21 | grep "Escape the Cell.esp" | grep Vampire | sed 's/Data.\+\\//' | sed 's/\.xwm/.wav/' > vampire.csv

Creating Batch File for xVASynth
1. Install xVASynth if you don't have it already installed
2. Also install the model for the voicetype you need
3. Open xVASynth and click the table icon
4. Click Generate Sample
5. Open sample.csv
6. Delete all the rows except the first
7. Put in the basic info for the model you're using
8. Import the .csv you created earlier
9. Copy in text and output columns, save file

Generating .WAV files using xVASynth
1. Open xVASynth and Click the table icon
2. Drag and drop the .csv you created into xVASynth
3. Click Synthesize Batch (THIS WILL TAKE A WHILE)
4. The .wav files will be created in the base xVASynth folder

Generating .lip files
1. Close the Creation Kit if you have it open
2. Copy the .wav filers from the xVASynth folder to Data\Sound\Voice\[ESP NAME]\[VOICE TYPE]\ (eg. Data\Sound\Voice\Escape the Cell.esp\FemaleYoungEager\)
3. Run the Creation Kit with launch argument -GenerateLips:[ESP NAME] (eg. -GenerateLips:Escape the Cell.esp)
4. If it worked there should now be a .lip file for each of your .wav files

(Optional) Generating .fuz files
1. Install Yakitori Audio Converter if you don't have it already installed
2. Open Yakitori Audio Converter and open the folder with your .wav and .lip files
3. Select Output Format .fuz and click Convert
4. You can delete or move the .wav and .lip files now

Additional Notes:
Some mods specify the actor associated with dialogue topics via a faction condition. If the NPC has this faction in the Creation Kit, then no changes to the steps are needed. If the faction is added dynamically during the game (eg. CurrentFollowerFaction) you'll need to create a temporary patch that adds that faction to the NPC in order to get these dialogue topics to export and have .lip files generated. After the .lip files are generated you can safely disable this patch.

I've noticed a number of times the -GenerateLips automation process consistently skips certain .wav files. I'm not sure if this is a bug with the creation kit or some quirk of specific topic conditions. If you find yourself with some .wav files missing a .lip file after using the -GenerateLips, you'll need to go into the Creation Kit and manually generate the missing .lip files. If someone runs into this and finds a workaround this let me know!