diff --git a/pypdf/filters.py b/pypdf/filters.py
index 82d2e0c9b..b2bffa385 100644
--- a/pypdf/filters.py
+++ b/pypdf/filters.py
@@ -721,6 +721,8 @@ def _get_imagemode(
         Image mode not taking into account mask(transparency)
         ColorInversion is required (like for some DeviceCMYK)
     """
+    if isinstance(color_space, NullObject):
+        return "", False
     if isinstance(color_space, str):
         pass
     elif not isinstance(color_space, list):
@@ -931,6 +933,9 @@ def _handle_jpx(
         extension = ".jp2"  # mime_type = "image/x-jp2"
         img1 = Image.open(BytesIO(data), formats=("JPEG2000",))
         mode, invert_color = _get_imagemode(color_space, colors, mode)
+        if mode == "":
+            mode = cast(mode_str_type, img1.mode)
+            invert_color = mode in ("CMYK",)
         if img1.mode == "RGBA" and mode == "RGB":
             mode = "RGBA"
         # we need to convert to the good mode
@@ -1028,6 +1033,8 @@ def _handle_jpx(
             False,
         )
     else:
+        if mode == "":
+            raise PdfReadError(f"ColorSpace field not found in {x_object_obj}")
         img, image_format, extension, invert_color = (
             Image.frombytes(mode, size, data),
             "PNG",
diff --git a/tests/test_filters.py b/tests/test_filters.py
index 54afcfc0e..3bd9422d1 100644
--- a/tests/test_filters.py
+++ b/tests/test_filters.py
@@ -626,3 +626,18 @@ def test_singleton_device():
     name = "pypdf_with_arr_deviceRGB.pdf"
     reader = PdfReader(BytesIO(get_pdf_from_url(url, name=name)))
     reader.pages[0].images[0]
+
+
+@pytest.mark.enable_socket()
+def test_jpx_no_spacecode():
+    """From #2061"""
+    url = "https://github.com/py-pdf/pypdf/files/12253581/tt2.pdf"
+    name = "jpx_no_spacecode.pdf"
+    reader = PdfReader(BytesIO(get_pdf_from_url(url, name=name)))
+    im = reader.pages[0].images[0]
+    # create an object without filter and without colorspace
+    # just for coverage
+    del im.indirect_reference.get_object()["/Filter"]
+    with pytest.raises(PdfReadError) as exc:
+        reader.pages[0].images[0]
+    assert exc.value.args[0].startswith("ColorSpace field not found")