I finally figured out how to fix the mod! Disclaimer: I'm not a modder, so it took a lot of googling and trial and error to get it to work. There might be an easier way to do it, but this is the only method I found that worked for me. And I tried A LOT of things. It's a bit of a long process and you have to download a few different things to convert the files to the format you need. The main issue is that the old audio files are wem opus, but the game now uses wem vorbis files.
First, you need to unpack the data from the mod using bg3 modders multitool (you can just download this and extract to a folder). To do this, run the bg3-modders-multitool.exe and drop the mod .pak where it says drop "mod .pak here." The unpacking should happen automatically, but it may take a while (it usually stalls at 99% for a while when I do it). Make sure that the program says it's complete before you try to grab the unpacked files, or you might not get all of them.
Once the mod's unpacked, got to the shortcut dropdown and click "unpacked mods" to get to the directory where your unpacked mod files are. Next, you'll want to move or copy the unpacked audio files to a folder where you will do your conversion work (I just called mine "convert" and then added folders for the mod names inside so that I could tell them apart). You only need the sound files, not the lsx file, so navigate into (voicemodfoldername)/Mods/ Gustav/Localization/English/Soundbanks to find them (you also won't need the gitignore file in there). If you're working on the point-and-click files, it'll be the files inside of (voicemodfoldername)/Public/Shared/Assets/Sound.
Next, you will need to download and install wwise. I'm using the most recent stable version of wwise rather than the beta one. You can choose which to install in the wwise launcher. If you choose to use the beta one, I can't promise the walkthrough of my steps will be the same.
Now that you have the modded audio files, the goal is to convert them from wem opus to wem vorbis. Unfortunately, wwise doesn't allow for importing .wem files (that I could find through extensive googling, anyway), so first you have to convert them to .wav or .ogg.
To do this, the easiest way I found was to use vgmstream (download and extract it to a folder of its own). You will want the vgmstream folder open in one tab and the folder containing the audio files in another tab. This is because you need to select the files and drop them over the executable for the conversion to run. This takes a bit of patience because there's a limit to how many it can run at a time. For companion audio files, I found it was about 210 for the regular files and about 300 for the point-and-click. Narrator files limit was 199. For this process, select a file, hold shift and scroll down a few times and select another file some way down to multiselect everything inbetween. File explorer should tell you in the bottom left how many you have selected all together, so once you have your desired number, drag them into the vgmstream tab and drop them over the executable. The great thing is that the program converts the files directly into the folder they're already in. Once the conversion batch has run, your old .wem files will still be selected and you can just press delete to get rid of them, since you only need the .wav files going forward.
This is the most time-consuming part of the process (I did it while bingeing Superstore lol). Everything after this is much faster. Once all of your files are converted to .wav, you might notice that unfortunately, the converter program included ".wem" in the file names (so it will say .wem.wav at the end of the file). It's not the extension, but in the actual name. You have to get rid of every single ".wem" in each file's name or the game won't be able to read it. To do this, you have to do a batch rename. Right click inside the the folder (not on a file, but in the blank space) and select "open in terminal." You need to paste the following code exactly: get-childitem *.wav | foreach { rename-item $_ $_.Name.Replace(".wem", "") } Unfortunately, there are too many files to run all at once and I noticed about half of them were getting missed, so I made a new folder that I called "half" within my folder, split half of the files into there, and then did this process in both folders to rename them all before putting my halved files back into the main folder. For the point-and-click files, you don't need to do this and can just rename them all in one go. After all of this, you can finally import the files to wwise to convert them.
This is where the steps get pretty confusing, and I can't figure out how to embed pictures, so hopefully my explanation makes sense.
After you have wwise installed, open a new project and name it whatever you want. Under the lefthand corner dropdown labelled "Project", select "import audio files" then select "add folders" in the prompt box that pops up. Navigate to the folder where your audio files are installed and select the folder (the files won't show up because it's only looking for a folder). You can do more than one folder at once, but I highly suggest having your folders named after the characters or mod you're converting so that you don't get mixed up (for example, "gale voice mod" and "astarion voice mod"). If you're converting the point-and-click files too, it's best to keep them in a separate folder as well (so "gale voice mod" and "gale point-and-click") because they have to be put into a different game directory at the end to load into the game.
Once you select your folder, the files inside of it should show up in the import pop-up menu. Click the import button and wait for them to be added to the project. You will now find your named folders in the "Audio" tab in the left side-bar. You will have to expand the Actor-Mixer Hierarchy and then the Default Work Unit*" folder within that to see your added folders. If you expand your folders, you should see your imported audio files if you've done the .wav conversion properly.
Double click on a file inside the folders you just uploaded and this should bring up a bunch of settings for the file in the main window of wwise. You'll have to navigate to the Conversion tab, right beside the General Settings tab that has opened. You'll see that the conversion settings are attached to a preset. All of the files have the "default conversion settings" set, so it's easiest to just edit this one rather than create a new one and have to apply it one-by-one to each file. Click the little arrow box next to the default settings preset and it will open a pop-up menu. You will see some dropdowns and options you can change. The only one you want to change is the dropdown under the Format heading that says "PCM." Change it to "vorbis" and then click the save icon in the top right of the pop-up window before exiting. Now the settings should be applied for all files.
Go to the Project drop-down in the top left corner of wwise and select "convert all audio files." You may have to wait a while for the conversion, but for me, it was generally less time than it took to import the .wav files.
Once the conversion is done, navigate to your user documents folder, then to WwiseProjects. From here, you want to select the folder with the name of the project you created. Then open .cache, then Windows, then SFX. You should see your named folders (i.e. gale voice mod, astarion voice mod) with the converted files inside. The file extension should now be back to .wem, but it should be the correct vorbis wem for the game to read (as opposed to the wem opus it was before). But you might have noticed a pesky batch conversion identifier has been added to EVERY SINGLE FILE by wwise during the conversion. I couldn't find a way to turn this off in the settings, so unless you do, you're just going to have to do batch rename again. Same as before, you want to split half the files into a folder inside the folder you're working in and right click in each to open in terminal. This time, you want to input the code: get-childitem *.wem | foreach { rename-item $_ $_.Name.Replace("batchidentifiergoeshere", "") } and you need to type the batch identifier into the spot that says batchidentifiergoeshere. This should look something like _8E324C96.
Once all of the files are renamed, you are finally done. All you have to do is go back to the mod folder you originally unpacked with bg3 modders multitool and delete the wem opus files and put the wem vorbis ones in their place. You can delete the folders containing the mod settings files as well. You don't need those. Now you can drag the Mods folder (or Public folder for point-and-click lines) into your BG3 Data directory and the voice mod files should copy into the right directory and work in game. If you know how to properly repack mods into a .pak using bg3 modders multitool, go ahead and do that, but when I did tried, even though I had the .pak activated, the modded files wouldn't play in my game. So I'm just using the loose files.
I've tested this method with this mod and other replacers I was using that haven't updated since patch 6. So far I've done three companions, the narrator, and a menu music replacer and they have all worked for me, including point-and-click lines.
I'm seriously not a modder though, so if you try to follow these steps and it doesn't work for you, I'm sorry but I don't think I'll be able to provide support.
I tried following your tutorial, but I can't seem to get it to work, If you have a functioning Astarion Voice 1 file, I'd really appreciate it if you could share the file with me.
Your files worked for me, at least the Gale ones. Saved me a huge headache, thank you. Maybe you should try see if the mod author will let you release them on here?
All companions with changed voices got muted in my game from patch 6. Rn only Halsin doesnt have changed voice so all of them speak without sound except him.
Yeah it is and I could do it, tho almost all my companions got new voice or/and new look(gender change) on this playtrought so I will have to revert it and it's not immersive for them to just change like that(Laezel, Shadowheart and Karlach was male, Astarion was changed around his new voice, same with Gale and Wyll) so I might just abandon this playtrought and start new with vanilla voices instead, I'm in mid of act 3 but oh well.
Right now I'm just doing break from BG3 in general, playing other games in it's place, because patch 6 did bring a lot of bugs, some are breaking game so waiting is better then starting new playtrought with possibility of being it broken later ihmo.
I am probably what most people on this site would consider "old," and I have to say I agree with what others have said... this complaint comes off as very childish and whiney. I love mods and the modding community, and one of the things I love about it is the ability to have choices and customize things to play they way you want. Especially in a game that relies heavily on RP and immersion, making the characters more relatable to the individual or customizing a character to better fit an idea you have in your head is super cool! Also, as others have said, mods are completely optional and are made by the community. To have the devs create a game where the official "canon" characters are older, and then come on the comments section for community created, unsupported mods and complain that people made a mod that doesn't fit your ideal head canon is super freakin weird and comes off as rude at best, and (as previously mentioned) childish and whiney at worst...
Speaking of being old, I hate that I can't find any mature looking heads. Even with the in game slider, it makes people look like they have turkey necks, and it doesn't match the age added to the face. It's like all or nothing. I am not skilled enough to mod, so I rely on other's talents. I have been modding a long time as well and feel that mods add so much to games. That being said, I have never seen a game that had every patch break so many mods. I think a mod creator should at least say that they are too busy/have no intent to update a mod for whatever reason so people can choose whether they want the mod or not. I won't berate or whine, but people are allowed to be disappointed. Who knows what is going on in this creator's life right now. I really liked the Gale version offered in this mod because I find his actual voice in game to be a bit too overdramatic (yes, I know it suits his personality - doesn't mean I have to like it).
I know how you feel about not being able to find a mature looking head that isn't Withers level LoL. The best and closest I can get to making a mature looking guy is by simply adding facial hair. There are lots of mods on here, and one for giving facial hair to elves (like Halsin <3) which is awesome. I made a Kratos tattoo for his face and body, added the closest looking beard to his as I could and BAM! got a hot mature looking Tav X'D Lmfao Larian really needs better customization. Was it too much to ask for a setup like Dragon's Dogma or something? Dang... As for updating, it depends on whether they're modding for free or being paid, and also how difficult it is and if they also have other things in life to deal with like work and family, etc. so as for updates, I'm trying to look into just learning how to make my own. Anyone know any good guides or sources on how to make ai voice mods for the companions? Also how to change the Narrators voice and the Emperors. I'm having fun getting into making ai voice models and been laughing my butt off at Raphael singing "Baby Said" by Maneskin X'D so yeah, I'd appreciate any advice/help.
I think the author made this mod for themselfs when played game and shared it to community, but stopped playing game before patch 6 so basically abandoned this mod also.
My recommendation (or two cents) is that you use AI voices that already have British accents if you do any other versions. The mix of American accent from the AI voice and British accent from the original actors is jarring to me personally and maybe others might feel the same.
Donwloaded Astarion v1 and hoping it will be fun to play with. Thank you for putting all this work into this mod!
Great voices, I was looking for Wyll voice replacement to be more race-neutral and less robotic. This will do, tho I wish it could be a tad deeper. But if not that will do..
70 comments
First, you need to unpack the data from the mod using bg3 modders multitool (you can just download this and extract to a folder). To do this, run the bg3-modders-multitool.exe and drop the mod .pak where it says drop "mod .pak here." The unpacking should happen automatically, but it may take a while (it usually stalls at 99% for a while when I do it). Make sure that the program says it's complete before you try to grab the unpacked files, or you might not get all of them.
Once the mod's unpacked, got to the shortcut dropdown and click "unpacked mods" to get to the directory where your unpacked mod files are. Next, you'll want to move or copy the unpacked audio files to a folder where you will do your conversion work (I just called mine "convert" and then added folders for the mod names inside so that I could tell them apart). You only need the sound files, not the lsx file, so navigate into (voicemodfoldername)/Mods/ Gustav/Localization/English/Soundbanks to find them (you also won't need the gitignore file in there). If you're working on the point-and-click files, it'll be the files inside of (voicemodfoldername)/Public/Shared/Assets/Sound.
Next, you will need to download and install wwise. I'm using the most recent stable version of wwise rather than the beta one. You can choose which to install in the wwise launcher. If you choose to use the beta one, I can't promise the walkthrough of my steps will be the same.
Now that you have the modded audio files, the goal is to convert them from wem opus to wem vorbis. Unfortunately, wwise doesn't allow for importing .wem files (that I could find through extensive googling, anyway), so first you have to convert them to .wav or .ogg.
To do this, the easiest way I found was to use vgmstream (download and extract it to a folder of its own). You will want the vgmstream folder open in one tab and the folder containing the audio files in another tab. This is because you need to select the files and drop them over the executable for the conversion to run. This takes a bit of patience because there's a limit to how many it can run at a time. For companion audio files, I found it was about 210 for the regular files and about 300 for the point-and-click. Narrator files limit was 199. For this process, select a file, hold shift and scroll down a few times and select another file some way down to multiselect everything inbetween. File explorer should tell you in the bottom left how many you have selected all together, so once you have your desired number, drag them into the vgmstream tab and drop them over the executable. The great thing is that the program converts the files directly into the folder they're already in. Once the conversion batch has run, your old .wem files will still be selected and you can just press delete to get rid of them, since you only need the .wav files going forward.
This is the most time-consuming part of the process (I did it while bingeing Superstore lol). Everything after this is much faster. Once all of your files are converted to .wav, you might notice that unfortunately, the converter program included ".wem" in the file names (so it will say .wem.wav at the end of the file). It's not the extension, but in the actual name. You have to get rid of every single ".wem" in each file's name or the game won't be able to read it. To do this, you have to do a batch rename. Right click inside the the folder (not on a file, but in the blank space) and select "open in terminal." You need to paste the following code exactly: get-childitem *.wav | foreach { rename-item $_ $_.Name.Replace(".wem", "") }
Unfortunately, there are too many files to run all at once and I noticed about half of them were getting missed, so I made a new folder that I called "half" within my folder, split half of the files into there, and then did this process in both folders to rename them all before putting my halved files back into the main folder. For the point-and-click files, you don't need to do this and can just rename them all in one go. After all of this, you can finally import the files to wwise to convert them.
This is where the steps get pretty confusing, and I can't figure out how to embed pictures, so hopefully my explanation makes sense.
After you have wwise installed, open a new project and name it whatever you want. Under the lefthand corner dropdown labelled "Project", select "import audio files" then select "add folders" in the prompt box that pops up. Navigate to the folder where your audio files are installed and select the folder (the files won't show up because it's only looking for a folder). You can do more than one folder at once, but I highly suggest having your folders named after the characters or mod you're converting so that you don't get mixed up (for example, "gale voice mod" and "astarion voice mod"). If you're converting the point-and-click files too, it's best to keep them in a separate folder as well (so "gale voice mod" and "gale point-and-click") because they have to be put into a different game directory at the end to load into the game.
Once you select your folder, the files inside of it should show up in the import pop-up menu. Click the import button and wait for them to be added to the project. You will now find your named folders in the "Audio" tab in the left side-bar. You will have to expand the Actor-Mixer Hierarchy and then the Default Work Unit*" folder within that to see your added folders. If you expand your folders, you should see your imported audio files if you've done the .wav conversion properly.
Double click on a file inside the folders you just uploaded and this should bring up a bunch of settings for the file in the main window of wwise. You'll have to navigate to the Conversion tab, right beside the General Settings tab that has opened. You'll see that the conversion settings are attached to a preset. All of the files have the "default conversion settings" set, so it's easiest to just edit this one rather than create a new one and have to apply it one-by-one to each file. Click the little arrow box next to the default settings preset and it will open a pop-up menu. You will see some dropdowns and options you can change. The only one you want to change is the dropdown under the Format heading that says "PCM." Change it to "vorbis" and then click the save icon in the top right of the pop-up window before exiting. Now the settings should be applied for all files.
Go to the Project drop-down in the top left corner of wwise and select "convert all audio files." You may have to wait a while for the conversion, but for me, it was generally less time than it took to import the .wav files.
Once the conversion is done, navigate to your user documents folder, then to WwiseProjects. From here, you want to select the folder with the name of the project you created. Then open .cache, then Windows, then SFX. You should see your named folders (i.e. gale voice mod, astarion voice mod) with the converted files inside. The file extension should now be back to .wem, but it should be the correct vorbis wem for the game to read (as opposed to the wem opus it was before). But you might have noticed a pesky batch conversion identifier has been added to EVERY SINGLE FILE by wwise during the conversion. I couldn't find a way to turn this off in the settings, so unless you do, you're just going to have to do batch rename again. Same as before, you want to split half the files into a folder inside the folder you're working in and right click in each to open in terminal. This time, you want to input the code: get-childitem *.wem | foreach { rename-item $_ $_.Name.Replace("batchidentifiergoeshere", "") } and you need to type the batch identifier into the spot that says batchidentifiergoeshere. This should look something like _8E324C96.
Once all of the files are renamed, you are finally done. All you have to do is go back to the mod folder you originally unpacked with bg3 modders multitool and delete the wem opus files and put the wem vorbis ones in their place. You can delete the folders containing the mod settings files as well. You don't need those. Now you can drag the Mods folder (or Public folder for point-and-click lines) into your BG3 Data directory and the voice mod files should copy into the right directory and work in game. If you know how to properly repack mods into a .pak using bg3 modders multitool, go ahead and do that, but when I did tried, even though I had the .pak activated, the modded files wouldn't play in my game. So I'm just using the loose files.
I've tested this method with this mod and other replacers I was using that haven't updated since patch 6. So far I've done three companions, the narrator, and a menu music replacer and they have all worked for me, including point-and-click lines.
I'm seriously not a modder though, so if you try to follow these steps and it doesn't work for you, I'm sorry but I don't think I'll be able to provide support.
Best of luck!
If you have a functioning Astarion Voice 1 file, I'd really appreciate it if you could share the file with me.
I've only done the Astarion Nero and Gale Revolver Ocelot voices. I haven't tested it with patch 7, but it works for patch 6.
Are you planning on adding male voices for female companions? Just asking.
A well done mod.
Really hope it will be. I'm so in love with the voices, would be sad to play with the original ones and not be able to hear them anymore.
Right now I'm just doing break from BG3 in general, playing other games in it's place, because patch 6 did bring a lot of bugs, some are breaking game so waiting is better then starting new playtrought with possibility of being it broken later ihmo.
Oh mod author, wherefore art thou? :)
Donwloaded Astarion v1 and hoping it will be fun to play with. Thank you for putting all this work into this mod!
Lemme guess. you use a white wyll mod too.