Edit 2: Well I’ll be damned. An extremely knowledgeable and kind stranger just reverse-engineered the whole thing and poured it into a python script. And it’s only Monday. See comments for the script.

Edit: Oh wow, this community is already on fire. Thanks for your advice everybody, I didn’t even think of intercepting the downloads in transit! Brilliant.

I will try to see how far I can get there, but that does sound much easier than rummaging around in iOS. Thank you all :)

Hello,

I’m trying to get the downloaded audio out of an iOS app, but I struggle because the information I can find is mostly rather old, needs some additional software I need to pay for, etc. The content is downloaded post installing the app, so simply accessing the IPA doesn’t help.

I have this app called naturespace (see naturespace.org), it’s an app that has really good recordings of rain, thunderstorms, etc. In my opinion those recordings are far better than anything I’ve heard so far.

Now, I did pay for the content, but the app hasn’t been updated for years now, and there’s also been no new content for years as well. I wrote to the owners but didn’t get any response. I guess you could consider it abandoned at this point.

Since I fear that anytime soon the app stops working, I’d like to save that content.

I’m a bit tech savvy, I can work with CLI and such, but I’m not a professional coder or hacker, any help is appreciated.

  • HiDefMusic@feddit.ukEnglish
    8·
    2 years ago

    If the content is being downloaded then you could do the following:

    • Setup a proxy like Fiddler2 on a PC on the same network your phone is connected to. You’ll need to configure Fiddler2 to decrypt HTTPS and then access the proxy certificate on your iPhone and add it as a trusted certificate. You’ll need to Google how to do this with Fiddler2 but it’s not hard

    • Configure your iPhone to connect through the Fiddler2 proxy by modifying your WiFi settings. If your PC is allowing connections to the Fiddler2 proxy port (e.g. 8888) and you’re trusted the cert on your iPhone, then connecting to websites in safari should be working

    • Open the naturespace app on your iPhone and delete the sounds you’re interested in. ONLY do this if you’re sure you can still re-download them. If not, this whole approach won’t work so just stop here.

    • Download the sounds again

    You should then be able to see the web requests that were made to download the sounds in Fiddler2 on your PC. As long as there’s no fancy encryption, you should be able to just save the content out of Fiddler2.

    • quandoquando@slrpnk.netOPEnglish
      2·
      2 years ago

      Ah, thank you! See, it didn’t even occur to me to just intercept the audio in transit. That’s really helpful, I will try this, thank you very much :)

      • HiDefMusic@feddit.ukEnglish
        4·
        2 years ago

        So I’ve reverse-engineered the naturespace Android APK and the files it downloads are definitely encrypted. They’re zip files (named as .nzp) that are XOR obfuscated with a rotating key every X amount of bytes. I haven’t quite worked out how the key rotates itself but I’m close. If I get it working I’ll put the details here and I can give you a Python script to grab whatever sounds you need.

        • quandoquando@slrpnk.netOPEnglish
          3·
          2 years ago

          Ha, I was just writing an update when your comment came.

          I followed your advice and installed mitmproxy (basically fiddler2 but open source), which was easy enough, and managed to find that the app just posts GET requests the homepage, which result in a 302 Temporarily Moved, which ends on a public S3 folder.

          The GET request includes some “ID”, which I’m not sure I should post publicly, maybe it might identify me? It’s like:

          GET http://www.naturespace.com/ns5ios/?command=download&path=%2Fmedia%2Fmodules%2Fcom.HolographicAudioTheater.Naturespace.Aegir&lang=en&id=REDACTED&bvrs=5.15&sysv=16.5&model=iPhone&bid=com.HolographicAudioTheater.Naturespace&sys=iOS&loc=en_DE HTTP/1.1

          But yes, it seems the files are encrypted. I couldn’t find anything to open them, and no file identifier knows what it is. If you manage to get somewhere, that’d be awesome, my tech knowledge definitely ends here lol.

          I guess it’s not actually illegal to post this, since it really is just a public folder, so if anyone els wants to look at it, here’s a file.

          • HiDefMusic@feddit.ukEnglish
            6·
            2 years ago

            Interesting, looks like they might be using a completely different file format for iOS versus Android. In any case, I’ve knocked up a script which will extract the track.ogg file from any pack of your choosing. Pasting directly here to see if it works (haven’t tried sharing code on Lemmy).

            You can browse available packs using the below URL. If you want to find out a pack name, just copy the banner image URL for it and you’ll see the “com.whatever” name in the URL itself.

            http://www.naturespace.com/android/v3/store/?live=true&udid=0

            Code:

            import sys
import requests
import hashlib
import io
import zipfile

ns_baseurl = "https://s3.amazonaws.com/naturespace/kindle_catalog/"

# Encryption key
key = b'DE2#We@(# sffsFSHfssfhsSFHSs_+&Gaa s,W.Z./lSFGSDF! NOWG!fjasdflasdkfjSADFKJASdflskgj fdkaG8HS42dncuFFSe=-56a'

def decryptNS(content):
	x = 1025
	y = 0
	dec = bytearray()
	for i in range(x,len(content)):
		if ((i+1024) % 1024) != 0:
			dec += bytes([content[i] ^ key[y % len(key)]])
			y = y+1

	return dec

if __name__ == '__main__':
	if len(sys.argv) < 2:
		print("Please provide a pack/module name (e.g. 'com.HolographicAudioTheater.Naturespace.TheImaginarium')")
		sys.exit(0)

	pack = sys.argv[1]
	json_url = ns_baseurl + pack + "/data.json"
	size = requests.get(json_url).json()["packageSize"]
	print(size)

	hashval = hashlib.sha1((pack + "8DvJx25sXwNwq2p" + size).encode()).hexdigest()
	dlurl = ns_baseurl + pack + "/" + hashval + "/" + pack + ".nzp"
	print(dlurl)

	content = decryptNS(requests.get(dlurl).content)
	"""
	with open(pack + ".zip", "wb") as f:
		f.write(content)
	"""
	zipf = io.BytesIO(content)
	zipfile = zipfile.ZipFile(zipf, 'r')
	track_nsp = zipfile.read('track.nsp')

	track_ogg = decryptNS(track_nsp)
	with open(pack + "_track.ogg", "wb") as f:
		f.write(track_ogg)
            • quandoquando@slrpnk.netOPEnglish
              4·
              2 years ago

              :O Wow.

              That is just simply amazing. Can confirm it works.

              The file names are always the same, prefix and the sounds name, e.g. if the sound is called “The Electric Forest” it’s com.HolographicAudioTheater.Naturespace.TheElectricForest.

              I guess I could write a scraper and parse the whole collection.

              Well, thank you kind stranger on the internet.

  • Clovis :bdx_town:@bdx.town
    6·
    2 years ago

    @quandoquando What about:
    * Connect your phone to a Wifi endpoint created from your computer (requires to two internet interfaces, 1 to emit the Wifi endpoint, 1 to still receive internet)
    * Start wireshark (https://www.wireshark.org/) to spy on network requests from your phone
    * Start the app
    * Watch for network requests from your phone to see if they point to something that looks like mp3

    @Piracy

    • quandoquando@slrpnk.netOPEnglish
      2·
      2 years ago

      Yes, I thought about that as well, but if possible I would prefer to retain the original data (also I just like to tinker with stuff :) ).

      However, like you said, they’re doing some “spatial” stuff they don’t disclose. I assume it’s just some EQ, but if they really have some algorithm to compute HRTF or even they’re own audio format, then re-recording might actually be the only solution.

  • TwilightKiddy@programming.devEnglish
    3·
    2 years ago

    The most reliable way is probably accessing app’s files directly, but you’ll need jailbreak for that. If it’s being downloaded, maybe you could get a packet trace? Found this article about it. Alternatively, you could setup a VPN in your network and just sniff the URL.