diff --git a/src/core/CanMessage.cpp b/src/core/CanMessage.cpp
index 386cfbfc..948f1647 100644
--- a/src/core/CanMessage.cpp
+++ b/src/core/CanMessage.cpp
@@ -184,33 +184,31 @@ void CanMessage::setByte(const uint8_t index, const uint8_t value) {
 
 uint64_t CanMessage::extractRawSignal(uint8_t start_bit, const uint8_t length, const bool isBigEndian) const
 {
-//    if ((start_bit+length) > (getLength()*8)) {
-//        return 0;
-//    }
+    if ((length == 0) || (length > 64)) {
+        return 0;
+    }
 
     // FIXME: This only gives access to data bytes 0-8. Need to rework for CANFD.
-    uint64_t data = le64toh(_u64[0]);
-
-    data >>= start_bit;
-
-    uint64_t mask =  0xFFFFFFFFFFFFFFFF;
-    mask <<= length;
-    mask = ~mask;
-
-    data &= mask;
-
-    // If the length is greater than 8, we need to byteswap to preserve endianness
-    if (isBigEndian && (length > 8))
-    {
-
-        // Swap bytes
-        data = __builtin_bswap64(data);
+    const uint8_t avail_bits = 8 * ((getLength() > 8) ? 8 : getLength());
+    if (start_bit + length > avail_bits) {
+        return 0;
+    }
 
-        // Shift out unused bits
-        data >>= 64 - length;
+    const uint64_t mask = (length == 64) ? ~(uint64_t)0 : (((uint64_t)1 << length) - 1);
+
+    if (isBigEndian) {
+        // DbcParser has already normalized start_bit into an MSB-first bit index
+        // counted from the first transmitted bit of the payload. The signal
+        // therefore occupies bits [start_bit, start_bit+length) of the payload
+        // read as one big-endian 64 bit value.
+        uint64_t be = 0;
+        for (int i = 0; i < 8; i++) {
+            be = (be << 8) | _u8[i];
+        }
+        return (be >> (64 - start_bit - length)) & mask;
     }
 
-    return data;
+    return (le64toh(_u64[0]) >> start_bit) & mask;
 }
 
 void CanMessage::setDataAt(uint8_t position, uint8_t data)