[Barcode4j-users] PDF417 Incorrect encoding

[<Tyl...@cr...>]

I'm using Barcode4J version 2.1, trying to encode a string in a PDF417 
barcode.
Some sections of the message are either being encoded or read incorrectly.
The following is the test I'm running:

File outputFile = new File("out.png");
OutputStream out = null;
final int dpi = 150;
PDF417Bean bean = new PDF417Bean();
bean.setColumns(10);
bean.setErrorCorrectionLevel(5);
try {
  out = new FileOutputStream(outputFile);
  BitmapCanvasProvider canvas = new BitmapCanvasProvider(out, 
"image/x-png", dpi, BufferedImage.TYPE_BYTE_BINARY, false, 0);
  bean.generateBarcode(canvas, 
"[)>\u001E06\u001D9S1823\u001D20L401\u001E06\u001DP106158-021\u001DQ0\u001D14K267183+10\u001D11K123\u001E\u0004 
");
  canvas.finish();
} catch(Exception e) {
  e.printStackTrace();
} finally {
  if(out != null) {
    try {
      out.close();
    } 
    catch(Exception e) {
      e.printStackTrace();
    }
  }
}

I've uploaded the output image to here: http://i.imgur.com/DNsvT.png

Reading the barcode gives: 
"[)>#06#9S1823#CA#EAB#06#P106158-021#Q0#14K267183+10#BB&BCD## "
It should be something like: 
"[)>#06#9S1823#20L401#06#P106158-012#Q0#14K267183+10#11K123## "
Note: my scanner doesn't display control characters, only giving a box, so 
I've used '#' in place of that.
I've also tried a couple of online readers which also give the incorrect 
value.

I'll see if I can whittle this down to a simplified test message later.

Thank you,
Tyler Koenig