diff --git a/Reading_MT/read_headers.py b/Reading_MT/read_headers.py
new file mode 100644
index 0000000000000000000000000000000000000000..5440592c3a03e6216dd47176610159bde5e97c8f
--- /dev/null
+++ b/Reading_MT/read_headers.py
@@ -0,0 +1,36 @@
+import codecs
+import struct
+
+file = open('Reading_MT/F000001.MT','rb')
+
+data = file.read()
+
+file.close()
+
+
+"""
+Dataset contains fixed blocks of blocklength 3840 bytes (characters in EBCDIC code)
+each block 30 records of 128 bytes
+first 24 bytes contain identification data
+
+Block nr    | Type
+
+1           | File Descriptor
+2,3         | Observation Header
+...
+
+
+
+Use codecs.decode(data, 'cp500') to read strings (C)
+Use struct.unpack('>l',data) to read 4byte long (F?)
+"""
+#Read first block of data
+block_1 = data[:3840:]
+
+record_size = 128
+records = 30
+id_data = block_1[:24]
+block_1 = [block_1[24+i:24+i+record_size] for i in range(0,records*record_size, record_size)]
+print((block_1[0:2]))
+#print(codecs.decode(block_1[0],'unicode-escape'))
+print(type(block_1[0]))
\ No newline at end of file