public static Int32 GetInt32(byte[] buffer, int offset, EndianType byteOrder)
        {
            if (byteOrder == EndianType.LittleEndian)
            {
                return buffer[offset + 1] << 8 | buffer[offset];
            }
            else
            {
                return buffer[offset] << 8 | buffer[offset + 1];
            } 
        }

Found this in some code I've been maintaining/expanding.   Quite clever.  

Picture an array of bytes in binary 

00000111 11100000 00000000 00000000

Now picture an Int32 in Binary
00000000000000000000000000000000

To grab a Little Endian Integer (16 bits) out of the array.  

Take the second array element
11100000

Assign it to the Int32 variable
00000000000000000000000011100000

Bit shift it up by 8 
00000000000000001110000000000000

Now OR it with the first Array element - could also just add.
<pre>
        00000000000000001110000000000000
or     00000000000000000000000000000111
=      00000000000000001110000000000111
</pre>

Extracting an Int from a Little Endian or Big Endian Byte Array

math

C#