I came across a challenge this week. Migrating Blob data from Oracle to SQL Server. I used MS Access for this since I could create an ODBC DSN to each and then have the best of ALL worlds. That is review the data, count the records and write very simple code. The link below is excellent and worked perfectly for what I needed for dealing with the Blobs.  

How To Read and Write BLOBs Using GetChunk and AppendChunk: http://support.microsoft.com/default.aspx?scid=kb;en-us;194975

The process I used for the Blobs was this:
1. Copy the data from Oracle to SQL exluding the Blobs Column.
2. Export the Blobs to File using the ID field as a key like (ID) & FileName. FileName was a column in my Oracle table so that was nice and pretty well needed so you can Identift the MIME (file type like XLS, DOC, PDF).
3. Finally Update the SQL Server table using File to Blob.

I had some other "things" in the code to deal with missing files from the Blob column but that was simple.
 

Wondering where the Oracle Drivers on on 64 bit Windows?

C:\Windows\SysWOW64 and open the odbcad32.exe executible.

Enjoy!