diff --git a/dhcp2024/render-chunks.sh b/dhcp2024/render-chunks.sh
index f5eb5d5..4a2898b 100755
--- a/dhcp2024/render-chunks.sh
+++ b/dhcp2024/render-chunks.sh
@@ -1,6 +1,19 @@
 #!/usr/bin/env bash
 set -euxo pipefail
 
+# This script is designed to help you automatically select the correct intro
+# and outro files (prerendered with built-in fades) and the correct chunks
+# of a chunked recording (e.g. OBS with automatic split every 5 minutes).
+# It will open the first and last chunk in mpv if you want, so that you
+# can find the start and end points (in seconds), you manually need to transfer
+# them into the script for now (no idea how to convince mpv to output it).
+# After that the script will render the introfile with the first chunk and the
+# last chunk with the outrofile, and then in the last step assemble everything
+# into the final recording with audio normalization. The final version should
+# try to use -c:v copy in the last step, so that no re-encoding of the chunks
+# in the middle will be done. This does not work yet, so please excuse the slow
+# rendering. Upcoming version hopefully fixes it.
+
 # This script requires you to have ffmpeg and fzf installed.
 which ffmpeg >/dev/null || (echo "Please install ffmpeg" ; exit 1)
 which fzf >/dev/null || (echo "Please install fzf" ; exit 1)