diff --git a/examples/badger2040/image_converter/convert.py b/examples/badger2040/image_converter/convert.py
index 6e7931899..d8613cce5 100755
--- a/examples/badger2040/image_converter/convert.py
+++ b/examples/badger2040/image_converter/convert.py
@@ -1,15 +1,32 @@
 #!/usr/bin/env python3
+"""
+Converts images into a format suitable for display on Badger 2040.
 
-# converts images into a format suitable for display on badger2040. this
-# includes scaling the image fit the longest edge, cropping down to 296x128
-# and reducing to black and white with dither. the data is then output as an
-# array that can be embedded directly into your c++ code
+Optionally resizes images to 296x128 to fit the display.
+
+Crunches images down to dithered, 1bit colour depth.
+
+Outputs either in raw binary format or as a .py file for embedding into MicroPython.
+
+Output to py functionality is borrwed from data_to_py.py, Copyright (c) 2016 Peter Hinch
+"""
 
 import io
 import argparse
 from PIL import Image, ImageEnhance
 from pathlib import Path
-import data_to_py
+
+
+PY_HEADER = """# Code generated by convert.py.
+"""
+
+PY_FOOTER = """_mvdata = memoryview(_data)
+
+def data():
+    return _mvdata
+
+"""
+
 
 parser = argparse.ArgumentParser(description='Converts images into the format used by Badger2040.')
 parser.add_argument('file', nargs="+", help='input files to convert')
@@ -21,18 +38,67 @@
 options = parser.parse_args()
 
 
+class ByteWriter(object):
+    bytes_per_line = 16
+
+    def __init__(self, stream, varname):
+        self.stream = stream
+        self.stream.write('{} =\\\n'.format(varname))
+        self.bytecount = 0  # For line breaks
+
+    def _eol(self):
+        self.stream.write("'\\\n")
+
+    def _eot(self):
+        self.stream.write("'\n")
+
+    def _bol(self):
+        self.stream.write("b'")
+
+    # Output a single byte
+    def obyte(self, data):
+        if not self.bytecount:
+            self._bol()
+        self.stream.write('\\x{:02x}'.format(data))
+        self.bytecount += 1
+        self.bytecount %= self.bytes_per_line
+        if not self.bytecount:
+            self._eol()
+
+    # Output from a sequence
+    def odata(self, bytelist):
+        for byt in bytelist:
+            self.obyte(byt)
+
+    # ensure a correct final line
+    def eot(self):  # User force EOL if one hasn't occurred
+        if self.bytecount:
+            self._eot()
+        self.stream.write('\n')
+
+
 def convert_image(img):
     if options.resize:
-        img = img.resize((296, 128)) # resize
+        img = img.resize((296, 128))  # resize
     try:
-      enhancer = ImageEnhance.Contrast(img)
-      img = enhancer.enhance(2.0)
+        enhancer = ImageEnhance.Contrast(img)
+        img = enhancer.enhance(2.0)
     except ValueError:
-      pass
-    img = img.convert("1") # convert to black and white
+        pass
+    img = img.convert("1")  # convert to black and white
     return img
 
 
+def write_stream(header, footer, ip_stream, op_stream):
+    op_stream.write(header)
+    op_stream.write('\n')
+    data = ip_stream.read()
+    bw_data = ByteWriter(op_stream, '_data')
+    bw_data.odata(data)
+    bw_data.eot()
+    op_stream.write(footer)
+
+
 # create map of images based on input filenames
 for input_filename in options.file:
     with Image.open(input_filename) as img:
@@ -59,7 +125,7 @@ def convert_image(img):
                 output_filename = Path(input_filename).with_suffix(".py")
             print(f"Saving to {output_filename}, {w}x{h}")
             with open(output_filename, "w") as out:
-                data_to_py.write_stream(io.BytesIO(bytes(output_data)), out)
+                write_stream(PY_HEADER, PY_FOOTER, io.BytesIO(bytes(output_data)), out)
         else:
             image_code = '''\
 static const uint8_t {image_name}[{count}] = {{
@@ -68,5 +134,3 @@ def convert_image(img):
     '''.format(image_name=image_name, count=len(output_data), byte_data=", ".join(str(b) for b in output_data))
 
             print(image_code)
-
-