Serialization using Base 64 and Hex Encoding... Which is Better

Question

I have serialized an object and stored in one of the Database column's. I had 2 options to choose from;
Option 1: Serialization into the database using Base 64 Encoder
public static void write(Object obj, PreparedStatement ps)
                throws SQLException, IOException {
                ByteArrayOutputStream baos = new ByteArrayOutputStream();
                ObjectOutputStream oout = new ObjectOutputStream(baos);
                oout.writeObject(obj);
                oout.close();
                byte[] buf = baos.toByteArray();
                s = new sun.misc.BASE64Encoder().encode(buf);
                ps.setString(1, s);
//                ps.setBytes(1, Base64.byteArrayToBase64(baos.toByteArray()));
              ps.setBytes(1, baos.toByteArray());
              ps.setInt(2, ++count);
      public static Object read(ResultSet rs, String column)
               throws SQLException, IOException, ClassNotFoundException {
           byte[] buf = new sun.misc.BASE64Decoder().decodeBuffer(s);
//           byte[] buf = Base64.base64ToByteArray(new String(rs.getBytes(column)));
           if (buf != null) {
           ObjectInputStream objectIn = new ObjectInputStream(
                          new ByteArrayInputStream(buf));
                          Object obj = objectIn.readObject();         //Contains the object
                          PersonDetails p = (PersonDetails)obj;
                          System.out.println(p.getName()+"\t"+p.getAge()+"\t"+p.getSex());
             return null;
Option 2: Serialization into the database using Hex Encoding
static char[]      encodeHex(byte[] data)
          Converts an array of bytes into an array of characters representing the hexidecimal values of each byte in order.
static byte[]      decodeHex(char[] data)
          Converts an array of characters representing hexidecimal values into an array of bytes of those same values.Using the above api functions encode and decode for serialization by writing similar Java Serialization read and write functions as shown above..
I want to know which is recommended out of the 2 encodings chosen based on performance and space used by the field in the database

Answer 1

The one which is best is the one which you determine to be best based on measuring your usage patterns in your environment on hardware as close as possible to what you will be using.
There is also a third option... why are you converting your byte[] into characters? You could use a binary field on the database - e.g. BLOB/BINARY/OBJECT instead of CLOB/MEMO/TEXT whatever