(04-08-2022, 07:28 AM)iteachvader Wrote: Quick update: I tried using mediaextract by Panzi on Github. It was able to rip a considerable amount of .wavs, but it feels like a lot are still missing. I couldn't find the A-wing or X-wing engine samples that I've heard in-game, no AT-ATs, lots of blaster sounds are missing, etc. I may be overlooking them, but I was at least able to rip exactly 11,800 .wavs. It also can rip the voice clips, which are in .ogg format. If anyone has another low-level program that's specifically designed to extract .wav files, let me know so I can try it out!

Sadly, Mediaextract doesn't preserve file metadata, so none of the .wavs have their correct filenames. Not quite fit for upload to the Sounds Resource just yet. (Also, that .wem file I provided earlier seems to just be a .wav file that was misread by GameExtractor. I was able to drag it into Audacity and it worked perfectly!)

(06-06-2022, 12:05 PM)iteachvader Wrote: Better buckle up, this is gonna be a hefty post.

connorh315's DATManager extracts all of the audio from the game. Grab it on GitHub here: https://github.com/connorh315/DATManager
ikskoks over on the Xentax forums wrote a script for converting the .audio_data files with vgmstream:

Code:

codec = IMA
start_offset = 0x5B
channels = 1
interleave = 0x100
sample_rate = @0x0C
num_samples = data_size

Copy this into Notepad, save as all files, and name it ".audio_data.txth". Place it in the vgmstream folder, and now you can drag and drop the files onto the vgmstream exe and it will instantly convert them to wav. The result will be a separate .wav file in the same directory as the original. It even properly converts files with more than one audio channel, at least up to four as far as I've seen.

This also works with Foobar2000 with the vgmstream plugin installed, but you have to place the .txth file in the same directory as the .audio_data files you want to convert. Then you can simply drag and drop into the Foobar window. Please make sure you have the most recent version of Foobar2000 and vgmstream!

Only a very small handful of the .audio_data files don't decode properly, such as VEH_BWING_ENGINE_LP.AUDIO_DATA. They have an odd loop address that throws off the vgmstream script. You can fix these by removing the 44-byte loop section in the header with a hex editor. Lastly, the WPN_MAGNAGUARDSTAFF_SWIPE sounds are all in 32-bit float @44100Hz rather than ADPCM. You can import these into Audacity easily.
Oh and the METAL_CREAK_BIG_DISTANT files are all for some reason encoded in Ogg Vorbis instead and can be dropped into Audacity WHAT
These formats are a mess and I love it~

Big thanks to my friends cu2, Tatrabbit and robynsun on Discord for helping me out, and of course to connorh315 and ikskoks for the extractor and script!

(06-09-2022, 01:37 PM)dcsimpson Wrote:

(06-06-2022, 12:05 PM)iteachvader Wrote: Better buckle up, this is gonna be a hefty post.

connorh315's DATManager extracts all of the audio from the game. Grab it on GitHub here: https://github.com/connorh315/DATManager
ikskoks over on the Xentax forums wrote a script for converting the .audio_data files with vgmstream:

Code:

codec = IMA
start_offset = 0x5B
channels = 1
interleave = 0x100
sample_rate = @0x0C
num_samples = data_size

Copy this into Notepad, save as all files, and name it ".audio_data.txth". Place it in the vgmstream folder, and now you can drag and drop the files onto the vgmstream exe and it will instantly convert them to wav. The result will be a separate .wav file in the same directory as the original. It even properly converts files with more than one audio channel, at least up to four as far as I've seen.

This also works with Foobar2000 with the vgmstream plugin installed, but you have to place the .txth file in the same directory as the .audio_data files you want to convert. Then you can simply drag and drop into the Foobar window. Please make sure you have the most recent version of Foobar2000 and vgmstream!

Only a very small handful of the .audio_data files don't decode properly, such as VEH_BWING_ENGINE_LP.AUDIO_DATA. They have an odd loop address that throws off the vgmstream script. You can fix these by removing the 44-byte loop section in the header with a hex editor. Lastly, the WPN_MAGNAGUARDSTAFF_SWIPE sounds are all in 32-bit float @44100Hz rather than ADPCM. You can import these into Audacity easily.
Oh and the METAL_CREAK_BIG_DISTANT files are all for some reason encoded in Ogg Vorbis instead and can be dropped into Audacity WHAT
These formats are a mess and I love it~

Big thanks to my friends cu2, Tatrabbit and robynsun on Discord for helping me out, and of course to connorh315 and ikskoks for the extractor and script!

Do you know if there is anywhere to just grab the wav files?  This seems sort of complicated and I wasnt able to download anything from the link.

(06-12-2022, 12:53 AM)iteachvader Wrote:

(06-09-2022, 01:37 PM)dcsimpson Wrote:

(06-06-2022, 12:05 PM)iteachvader Wrote: Better buckle up, this is gonna be a hefty post.

connorh315's DATManager extracts all of the audio from the game. Grab it on GitHub here: https://github.com/connorh315/DATManager
ikskoks over on the Xentax forums wrote a script for converting the .audio_data files with vgmstream:

Code:

codec = IMA
start_offset = 0x5B
channels = 1
interleave = 0x100
sample_rate = @0x0C
num_samples = data_size

Copy this into Notepad, save as all files, and name it ".audio_data.txth". Place it in the vgmstream folder, and now you can drag and drop the files onto the vgmstream exe and it will instantly convert them to wav. The result will be a separate .wav file in the same directory as the original. It even properly converts files with more than one audio channel, at least up to four as far as I've seen.

This also works with Foobar2000 with the vgmstream plugin installed, but you have to place the .txth file in the same directory as the .audio_data files you want to convert. Then you can simply drag and drop into the Foobar window. Please make sure you have the most recent version of Foobar2000 and vgmstream!

Only a very small handful of the .audio_data files don't decode properly, such as VEH_BWING_ENGINE_LP.AUDIO_DATA. They have an odd loop address that throws off the vgmstream script. You can fix these by removing the 44-byte loop section in the header with a hex editor. Lastly, the WPN_MAGNAGUARDSTAFF_SWIPE sounds are all in 32-bit float @44100Hz rather than ADPCM. You can import these into Audacity easily.
Oh and the METAL_CREAK_BIG_DISTANT files are all for some reason encoded in Ogg Vorbis instead and can be dropped into Audacity WHAT
These formats are a mess and I love it~

Big thanks to my friends cu2, Tatrabbit and robynsun on Discord for helping me out, and of course to connorh315 and ikskoks for the extractor and script!

Do you know if there is anywhere to just grab the wav files?  This seems sort of complicated and I wasnt able to download anything from the link.

It's not gonna be small. Here is every last bit of audio from Lego Star Wars: The Skywalker Saga converted to wav: https://www.mediafire.com/file/2yy9kffa5vhxoam/LSWTSS_SFX_WAV.zip/file
I've submitted the majority of these (aside from ambience and cutscenes) to the Sounds Resource.