FSB Extractor (many of them) will attempt to extract the unmodified .ogg file, which lacks a header due to using FMOD's library system to define things internally.rumpo wrote:Anyone had success converting sounds? I'm getting no headers with FSB extractor.
Luckily there's a way around that which essentially "records" the .ogg file into a PCM/WAV through sound emulation. The result is a playable .wav file with accuracy and multiple channel support.
You'll need both of these tools: viewtopic.php?f=17&t=13615
! NOTE: In order to use the tools, you will also need the library from http://zenhax.com/viewtopic.php?t=1317&start=40#p7400 and the libraries from viewtopic.php?p=112968#p112968
Why two tools? Because fmod_extr.exe will handle the majority of the audio detection and extraction but seems to fail on multi-channel sounds. fsb_aud_extr.exe will handle those specific failure cases. Luckily for you (and others who find this post) I've done the hard work of identifying which files require which tool.
Batch file for the audio within Dark Souls III\Game\sound\ , which will also list the tool needed: https://gist.github.com/NodusCursorius/ ... b51340aefd
But wait, there's more!
Batch file for the audio within Dark Souls III\Data5.bdt\ , after you use Atvaark's BinderTool on \Data5.bdt: https://gist.github.com/NodusCursorius/ ... c8dadf2ee0
Batch file for the audio within Dark Souls III\Data1.dbt\ , after you use Atvaark's BinderTool on \Data1.bdt: https://gist.github.com/NodusCursorius/ ... cfdc207270
----
As a side note, to future Googlers, many of the .dds files that you extract with BinderTool (from the .tpf files) may appear as unreadable or undecodeable, depending upon your preferred image viewer. This seems to be due to how some of the images are not just in format DX1, 2, or 5, but some are using DX10 and DX11 format which is a completely different beast. Microsoft has a way to easily input these kinds of images and export them to a different format. Using their DirectXTex (DirectXTex texture processing library) converstion tool found here: https://github.com/Microsoft/DirectXTex ... exconv.exe
It's a command line executable, but iterating through everything is easy enough with this line, which will process all .dds in all sub-directories and create .png files alongside each .dds :
Code: Select all
for /r . %T IN (*.DDS) do texconv.exe -ft PNG %T -o %~dpT! NOTE: If you're batch scripting the above, don't forget to change %T to %%T in all three places Also, texconv does not work if your directory structure has spaces, even when quoted.
----
5/27/16 - Batch files and gist links updated to reflect improved ds3_Data5_sound_extract.bat (for newer names due to Atvaark's tool) and ds3_Data1_sound_extract.bat added for .. well, obvious reasons. Cheers, Sentuh, for leading me to check discrepancies with newly discovered names.
