diff --git a/dfm/README.md b/dfm/README.md
new file mode 100644
index 0000000..85be751
--- /dev/null
+++ b/dfm/README.md
@@ -0,0 +1,44 @@
+
+## Radiosonde DFM-06/DFM-09
+
+#### Files
+
+  * `dfm06ptu.c`
+
+#### Compile
+  `gcc dfm06ptu.c -lm -o dfm06ptu`
+
+#### Usage
+  `./dfm06ptu [options] <audio.wav>` <br />
+  * `<audio.wav>`: FM-demodulated signal, recorded as wav audio file
+  * `options`: <br />
+      `-i`: invert signal/polarity (DFM-09)<br />
+      `-b`, `-b2`: integrate rawbit-/symbol-samples<br />
+      `-r`: output raw data <br />
+     `-v`: additional info <br />
+     `--ecc`: Hamming code error correction <br />
+     `--ptu`: temperature <br />
+
+
+#### Examples
+  * `./dfm06ptu --ecc --ptu -v dfm-audio.wav`
+
+  FSK-demodulation is kept very simple. If the signal quality is low and (default) zero-crossing-demod is used,
+  a lowpass filter is recommended:
+  * `sox dfm-audio.wav -t wav - lowpass 2000 2>/dev/null | ./dfm06ptu --ecc --ptu -v`
+
+  If timing/sync is not an issue, integrating the symbol-samples (option `-b2`) is better for error correction:
+  * `./dfm06ptu -b2 --ecc --ptu -v dfm-audio.wav`
+
+  For DFM-09 or if the signal is inverted
+  (depends on sdr-software and/or audio-card/settings), try option `-i`.
+
+
+####  Error correction
+  Use error correction `--ecc` together with option `-b2`.
+  The codewords are 8-bit long and the \[8,4\] extended Hamming code can correct one-bit errors
+  and detect two-bit errors. If there are more than two errors, it is possible that the codeword is
+  corrected to a different codeword closer to the received word. The minimum distance is 4.
+  Thus if the signal level is low, the decoding can go wrong.
+  A high number of codewords with bit-errors in a frame could indicate unreliable decoding.
+