Here's the Converter we wrote to solve this exact problem.

    Public Class BinaryToImageConverter
        Implements IValueConverter
        Public Function Convert(ByVal value As Object, ByVal targetType As System.Type, ByVal parameter As Object, ByVal culture As System.Globalization.CultureInfo) As Object Implements System.Windows.Data.IValueConverter.Convert
            Dim binaryData As Binary = value
            If binaryData Is Nothing Then
                Return Nothing
            End If

            Dim buffer() As Byte = binaryData.ToArray()
            If buffer.Length = 0 Then
                Return Nothing
            End If

            Dim res As New BitmapImage()
            res.BeginInit()
            res.StreamSource = New System.IO.MemoryStream(buffer)
            res.EndInit()
            Return res
        End Function

        Public Function ConvertBack(ByVal value As Object, ByVal targetType As System.Type, ByVal parameter As Object, ByVal culture As System.Globalization.CultureInfo) As Object Implements System.Windows.Data.IValueConverter.ConvertBack
            Dim source As BitmapImage = value
            If source Is Nothing Then
                Return Nothing
            End If
            If Not source.StreamSource Is Nothing AndAlso source.StreamSource.Length > 0 Then
                Return GetBytesFromStream(source.StreamSource)
            End If

            Return Nothing
        End Function

        Private Function GetBytesFromStream(ByVal stream As System.IO.Stream) As Binary
            stream.Position = 0
            Dim res(stream.Length) As Byte
            Using reader As New System.IO.BinaryReader(stream)
                reader.Read(res, 0, CType(stream.Length, Integer))
            End Using
            Return New Binary(res)
        End Function

    End Class

Just set the converter when you define the field like this:

</igDP:Field>

Here's how we did it: 

PseduoCode:

Run through your returned dataset and read the byteArray of your image field to a MemoryStream, then push that to a BitmapImage object, then use Create a Windows.Controls.Image object and tie the source to the BitmapImage you just created.....then you should be good to go. 

I have this in the Page_Loaded event

//Begin code

      string connString = "Persist Security Info=False;Integrated Security=SSPI;database=YOURDATABASEHERE;server=.\\INSTANCENAMEHERE";
      string commString = "select SomeTableID, BinaryDataField from dbo.SomeTable";

      SqlCommand comm = new SqlCommand();
      comm.CommandText = commString;
      comm.CommandType = System.Data.CommandType.Text;
      comm.Connection = new SqlConnection(connString);
      SqlDataAdapter da = new SqlDataAdapter(comm);
      DataTable dt = new DataTable();
      try
      {
        da.Fill(dt);
        // we have to convert the byte array into an image.
        // first step is to create a column for the actual image
        dt.Columns.Add("Image", typeof(System.Windows.Controls.Image));

        //next step is to get the Photo column into a byte array.
        foreach (DataRow dr in dt.Rows)
        {
          Byte[] myByteArray = (Byte[])dr["BinaryDataField"];
          MemoryStream ms = new MemoryStream(myByteArray, 0, myByteArray.Length);
          ms.Write(myByteArray,0,myByteArray.Length);
         
          //Move to the head of the MemoryStream or it will fail
          ms.Seek(0, SeekOrigin.Begin);

          //Send the MemoryStream over to a BitmapImage
          BitmapImage bi = new BitmapImage();
          bi.BeginInit();
          bi.StreamSource = ms;
          bi.EndInit();


          System.Windows.Controls.Image image = new System.Windows.Controls.Image();         
          image.StretchDirection = StretchDirection.Both;
          image.Stretch = Stretch.UniformToFill;
          image.Margin = new Thickness(4);
          //Assign the Image.source to the BitmapImage's Data
          image.Source = bi;
          dr["Image"] = image;
        }
           
        // remove the byte[] column.  This is optional; you can simply hide it if you wish.
        dt.Columns.Remove(dt.Columns["BinaryDataField"]);
       
        //Assign the DataTable's DefaultView to the Datasource
        this.XamDataCarousel1.DataSource = dt.DefaultView;

//End code

I just used this, so I know it works.

I have been writing a test app for the carousel that had both a horizontal and vertical item path'd carousel and when I was using the actual image (stored on my C: drive) and referencing the path, the 1st carousel took it ok and the second one failed because it said they were already in use....so if this is a similar case, then I would say always load it into a MemoryStream and go from there. 

A note about storing it into sql or store the path.....when you store just the path, you run into issues of some end user switching the path and unknowingly totally wrecking your app.  However, if you store them in the Database...ex: sql 2005 express, it has a size limit of 4GB per database, so if you are putting large images/documents into it, it could fill up quite quickly. ***There's no size limit on a full version of sql 2005...just express.  We are using express for most of our deployment and when the size gets close to the limit, we branch off to another database and we have another 4GB to use. 

There's some pros and cons to both ways, but I would suggest for dummy proofing your code/design, store them in the database.  Some people say this is a performance hog, but I haven't run into that issue yet...just the size limit. 

-Jeremy

Thank you! - sixbitslocum & Jeremy's solutions both worked for me!

My next question is what is the ideal/preffered way of doing SQL images w/Infragistics? .....such as: should I be reading the images from a file instead of from sql?...or store the image path instead of the actual image?