diff --git a/ToolsTOR.py b/ToolsTOR.py
index e7a0951..e79ec3c 100644
--- a/ToolsTOR.py
+++ b/ToolsTOR.py
@@ -74,12 +74,25 @@ class ToolsTOR(ToolsTales):
             self.extract_TheirSce_XML(scpk_file)
             self.id = 1
         
-    def get_theirsce_from_scpk(self, scpk, scpk_file_name, debug=False)->bytes:
-        header = scpk.read(4)
-    
-        if header != b"SCPK":
-            # sys.exit(f"{file} is not a .scpk file!")
-            raise ValueError("File is not a .scpk file!")
+    def get_theirsce_from_scpk(self, scpk_file_name, debug=False)->bytes:
+        
+        with open(scpk_file_name,"rb") as scpk:
+            header = scpk.read(4)
+        
+            if header != b"SCPK":
+                # sys.exit(f"{file} is not a .scpk file!")
+                raise ValueError("File is not a .scpk file!")
+        
+            scpk.read(4)
+            nbFiles = struct.unpack("<L", scpk.read(4))[0]
+            scpk.read(4)
+            filesSize = []
+            for i in range(nbFiles):
+                filesSize.append(struct.unpack("<L", scpk.read(4))[0])
+        
+            for i in range(nbFiles):
+                data = scpk.read(filesSize[i])
+                
     
         scpk.read(4)
         nbFiles = struct.unpack("<L", scpk.read(4))[0]
diff --git a/pak2.py b/pak2.py
new file mode 100644
index 0000000..a4b8b5c
--- /dev/null
+++ b/pak2.py
@@ -0,0 +1,162 @@
+import struct, sys, os, io
+from dataclasses import dataclass, field
+@dataclass
+class pak2_chunks:
+    theirsce: bytes = b""
+    lipsync: bytes = b""
+    unused: bytes = b""
+    image_unk1: bytes = b""
+    image_unk2: bytes = b""
+    image_blobs: list = field(default_factory=list)
+
+@dataclass
+class pak2_file:
+    #offsets: list
+    char_count: int = 0
+    slot_count: int = 0
+    image_count: int = 0
+    chunks: pak2_chunks = pak2_chunks()
+
+def get_file_name_noext(path):
+    return os.path.splitext(os.path.basename(path))[0]
+
+
+def get_parent_folder(path):
+    return os.path.normpath(os.path.join(path, os.pardir))
+
+
+def insert_padded_chunk(file: bytes, chunk: bytes, alignment: int = 4):
+    file.write(chunk)
+    pad = (alignment - (file.tell() % alignment)) % alignment
+    file.write(b"\x00" * pad)
+    return file.tell()
+
+
+def get_theirsce_from_pak2(file: bytes)->bytes:
+    offsets = struct.unpack("<3I", file[:12])
+
+    # Handle null 2nd offset because of course that's a thing
+    if offsets[1] == 0:
+        return file[offsets[0] : offsets[2]]
+    else:
+        return file[offsets[0] : offsets[1]]
+
+
+def get_data(file: bytes)->pak2_file:
+    offsets = struct.unpack("<6I", file[:24])
+    data = pak2_file()
+    data.char_count = struct.unpack("<H", file[0x18:0x1A])[0]
+    data.slot_count = struct.unpack("<H", file[0x1A:0x1C])[0]  # 0x20 always
+    data.image_count = struct.unpack("<H", file[0x1C:0x1E])[0]
+
+    # Handle null 2nd offset because of course that's a thing
+    if offsets[1] == 0:
+        data.chunks.theirsce = file[offsets[0] : offsets[2]]
+    else:
+        data.chunks.theirsce = file[offsets[0] : offsets[1]]
+        size = struct.unpack("<I", file[offsets[1] : offsets[1] + 4])[0] + 0x10
+        data.chunks.lipsync = file[offsets[1] : offsets[1] + size]
+
+    data.chunks.unused = file[offsets[2] : offsets[2] + (data.char_count * 4)]
+    data.chunks.image_unk1 = file[offsets[3] : offsets[3] + (data.slot_count * 4)]
+    data.chunks.image_unk2 = file[offsets[4] : offsets[4] + (data.image_count * 2)]
+    # image_data = bytearray(file[offsets[5]:len(file)])
+
+    blob_offsets = list(
+        struct.unpack(
+            "<%dI" % (data.image_count * 2), file[offsets[5] : offsets[5] + data.image_count * 8]
+        )
+    )
+    del blob_offsets[1::2] #Yeet all the odd offset, they be useless
+
+    for blob in blob_offsets:
+        blob_size = struct.unpack("<I", file[blob : blob + 4])[0]
+        if blob_size == 0:
+            blob_size = 0x400
+        data.chunks.image_blobs.append(file[blob : blob + blob_size])
+    
+    return data
+
+
+def create_pak2(data: pak2_file)->bytes:
+    output = io.BytesIO()
+    output.seek(0)
+    output.write(b"\x00" * 0x20)
+    offsets_new = []
+    offsets_new.append(output.tell())
+
+    # theirsce
+    offsets_new.append(insert_padded_chunk(output, data.chunks.theirsce))
+
+    # lipsync
+    offsets_new.append(insert_padded_chunk(output, data.chunks.lipsync))
+
+    # unused
+    offsets_new.append(insert_padded_chunk(output, data.chunks.unused))
+
+    # unk1
+    offsets_new.append(insert_padded_chunk(output, data.chunks.image_unk1))
+
+    # unk2
+    offsets_new.append(insert_padded_chunk(output, data.chunks.image_unk2))
+
+    # images
+    # Create image chunk
+    image_chunk = b"\x00" * (data.image_count * 8)  # minimum size
+    insert_padded_chunk(output, image_chunk, 128)
+    image_offsets = []
+
+    image_offsets.append(output.tell())
+
+    for blob in data.chunks.image_blobs:
+        image_offsets.append(insert_padded_chunk(output, blob, 128))
+
+    image_offsets = image_offsets[:-1]
+    image_offsets = [
+        val for val in image_offsets for _ in (0, 1)
+    ]  # image data offsets are duplicated
+
+    # Write image data offsets
+    output.seek(offsets_new[5])
+    output.write(struct.pack("<%dI" % len(image_offsets), *image_offsets))
+
+    # Write chunk offsets
+    output.seek(0)
+    output.write(struct.pack("<%dI" % len(offsets_new), *offsets_new))
+
+    # Write metadata
+    output.write(struct.pack("<H", data.char_count))
+    output.write(struct.pack("<H", data.slot_count))
+    output.write(struct.pack("<H", data.image_count))
+
+    end = output.getvalue()
+
+    return end
+
+
+if __name__ == "__main__":
+    if len(sys.argv) < 2:
+        print("Usage:")
+        print("pak2.py [pak2_path] [theirsce_path]")
+        sys.exit()
+
+    with open(sys.argv[1], "rb") as input:
+        file = input.read()
+
+    pak2 = get_data(file)
+
+    print("file size: %d" % len(file))
+    print("char_count: %d" % pak2.char_count)
+    print("slot_count: %d" % pak2.slot_count)
+    print("image_count: %d" % pak2.image_count)
+    print()
+
+    # Get new Theirsce, if there's no second arg just reinsert original
+    if len(sys.argv) > 2:
+        with open(sys.argv[2], "rb+") as f:
+            theirsce = f.read()
+
+    with open(sys.argv[2] + ".new", "wb+") as output:
+        output.write(create_pak2(pak2))
+
+    print("Done!")