Skip to content

Instantly share code, notes, and snippets.

Embed
What would you like to do?
parse_and_convert_rws_file.py - parses a rws audio file (xbox 360) and converts it to pcm, then runs it through sox to produce a WAV
#!/usr/bin/env python3
#
# parse rws file format and extract it, then run it through sox to produce a wav
#
#
# written by Mark Grandi - May 9th 2014
#
import argparse, sys, traceback, os, os.path, subprocess, tempfile, shutil
def main(args):
'''
@param args - the namespace object given to us by argparse'''
# recurse and find .rws files
filesToProcess = list()
for dirpath, dirnames, filenames in os.walk(args.rwsFolder):
for iterFileName in filenames:
if os.path.splitext(iterFileName)[1].lower() == ".rws":
filesToProcess.append(os.path.join(dirpath, iterFileName))
# go through each rws file and convert it
tempdir = tempfile.TemporaryDirectory()
print("Temporary directory is {}".format(tempdir.name))
for iterFile in filesToProcess:
filename = os.path.split(iterFile)[1]
pcmFilePath = os.path.join(tempdir.name, os.path.splitext(filename)[0] + ".pcm")
testBytes = (0).to_bytes(100, "little")
with open(iterFile, "rb") as rwsFile:
with open(pcmFilePath, "wb") as pcmFile:
# read the header of the file, 2047 bytes
rwsFile.read(2047)
# now loop and read the audio data, and the null bytes
counter = 0
while True:
counter += 1
# test to see if we need to break
oldPos = rwsFile.tell()
testData = rwsFile.read(100)
# see if we are at the end of the file? cause at the end of the file we have more then
# 2020 NULL bytes so if we get 100 consecutive bytes of NULL then its pretty safe to say
# that we are at the end of the file
if testData == testBytes or len(testData) == 0:
# at the end of the file
break
# put back at old position
rwsFile.seek(oldPos)
# read audio data
iterAudioData = rwsFile.read(32796)
# null bytes
rwsFile.read(2020)
# write to pcm file
pcmFile.write(iterAudioData)
if args.justDumpRaw:
# we are not converting, just dumping the raw pcm file
outputPcm = os.path.join(args.wavFolder, os.path.splitext(filename)[0] + ".pcm")
shutil.copyfile(pcmFilePath, outputPcm)
print("finished {}: raw pcm copied to {}".format(filename, outputPcm))
else:
# convert as normal
outputWav = os.path.join(args.wavFolder, os.path.splitext(filename)[0] + ".wav")
# sox -t raw -r 44100 -e signed-integer -b 16 --endian little -c 2 <input file> <output file>
subprocess.call(["sox", "-t", "raw", "-r", "44100", "-e", "signed-integer", "-b", "16",
"--endian", "little", "-c", "2", pcmFilePath, outputWav])
print("finished {}: converted and saved to {}".format(filename, outputWav))
# delete temporary directory
tempdir.cleanup()
def printTraceback():
'''prints the traceback'''
# get variables for the method we are about to call
exc_type, exc_value, exc_traceback = sys.exc_info()
# print exception
traceback.print_exception(exc_type, exc_value, exc_traceback)
def isDirectoryType(stringArg):
''' helper method for argparse to see if the argument is a directory
@param stringArg - the argument we get from argparse
@return the path if it is indeed a directory, or raises ArgumentTypeError if its not.'''
path = os.path.realpath(os.path.normpath(stringArg))
if not os.path.isdir(path):
raise argparse.ArgumentTypeError("{} is not a directory!".format(path))
return path
if __name__ == "__main__":
# if we are being run as a real program
parser = argparse.ArgumentParser(description="parses a rws audio file (xbox 360) and converts it to pcm, then run it through sox to produce a WAV",
epilog="Copyright Mark Grandi, May 9, 2014")
parser.add_argument("rwsFolder", type=isDirectoryType, help="the folder containing .RWS files")
parser.add_argument("wavFolder", type=isDirectoryType, help="the folder where we output the .wav files")
parser.add_argument("--justDumpRaw", action="store_true", help="if set then we will just dump the raw .pcm files to "
"wavFolder and not run them through sox")
try:
main(parser.parse_args())
except Exception as e:
print("Something went wrong...error: {}".format(e))
print("##################")
printTraceback()
print("##################")
sys.exit(1)
@QueryMan1992

This comment has been minimized.

Copy link

@QueryMan1992 QueryMan1992 commented Jan 12, 2020

Hey, so I used this for 'The Legend Of Spyro - Dawn Of The Dragon' to convert to WAV and whilst I get the audio data I also get a load of hissing in the background.

I did find your topic here - https://forum.xentax.com/viewtopic.php?f=17&t=11528

But I'm trying to establish where you put "so I just added a 0x00 byte to the end of the raw PCM data i extracted, and now it works perfectly! Audio data is....picky >.>" that null byte in?

@mgrandi

This comment has been minimized.

Copy link
Owner Author

@mgrandi mgrandi commented Jan 13, 2020

Hey, so I used this for 'The Legend Of Spyro - Dawn Of The Dragon' to convert to WAV and whilst I get the audio data I also get a load of hissing in the background.

I did find your topic here - https://forum.xentax.com/viewtopic.php?f=17&t=11528

But I'm trying to establish where you put "so I just added a 0x00 byte to the end of the raw PCM data i extracted, and now it works perfectly! Audio data is....picky >.>" that null byte in?

Hi!

I'll have to find my notes to ensure, but it might have just been that I literally put a null byte at the end of the PCM audio that my script extracts, before it sends it to Sox , so maybe use the --justDumpRaw option, add a null byte at the end using a hex editor and then putting it through sox manually? I'll comment again once I find my notes

@QueryMan1992

This comment has been minimized.

Copy link

@QueryMan1992 QueryMan1992 commented Jan 13, 2020

Hey, just tried that and I still get hissing back, will await for your further response when you've confirmed from your notes.

@QueryMan1992

This comment has been minimized.

Copy link

@QueryMan1992 QueryMan1992 commented Jan 15, 2020

I think I've gotten somewhere.

This bit in your code:

rwsFile.read(2047)

I changed 2047 to 2048 and when importing the PCM file into Audacity it sounds fine with no hissing, but when using SOX to convert to wav it becomes garbled.

I will continue to play around with the conversion and see if I get the desired effect.


Nevermind I just exported it using Audacity and now it's all good with no hissing :)

Thank you very much for creating this!

@mgrandi

This comment has been minimized.

Copy link
Owner Author

@mgrandi mgrandi commented Jan 15, 2020

I'm noticing that the start of the audio file still has hissing problems, I think what happened is that this isn't my final code, i think i wrote it to be more intelligent about the file format, and I only created this gist to post my 'current at the time' code when i made the forum post

i'll see if i can find my final code for you

@mgrandi

This comment has been minimized.

Copy link
Owner Author

@mgrandi mgrandi commented Jan 22, 2020

@QueryMan1992 i actually found my code and made a github repo out of it, see https://github.com/mgrandi/parse_rws_format

specifically, this file: https://github.com/mgrandi/parse_rws_format/blob/master/parse_rws_ver2.py

it uses ffmpeg this time, but if you don't have that you can do py -3 parse_rws_ver2.py <rws folder> <output folder> --justDumpRaw and import it into audacity (signed 16 bit, big endian, 2 channel and export it to whatever you want

i checked myself and the start of the file is free of hissing :)

@QueryMan1992

This comment has been minimized.

Copy link

@QueryMan1992 QueryMan1992 commented Jan 26, 2020

Thanks for this, it's perfect! 👍

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment