What is the delimiter for json output? - json

Question

I'm querying a db that returns json output using odbc using isql. Here's the help for isql.
isql -h
**********************************************
* unixODBC - isql *
**********************************************
* Syntax *
* *
* isql DSN [UID [PWD]] [options] *
* *
* Options *
* *
* -b batch.(no prompting etc) *
* -dx delimit columns with x *
* -x0xXX delimit columns with XX, where *
* x is in hex, ie 0x09 is tab *
* -w wrap results in an HTML table *
* -c column names on first row. *
* (only used when -d) *
* -mn limit column display width to n *
* -v verbose. *
* -lx set locale to x *
* -q wrap char fields in dquotes *
* -3 Use ODBC 3 calls *
* -n Use new line processing *
* -e Use SQLExecDirect not Prepare *
* -k Use SQLDriverConnect *
* -L Length of col display (def:300) *
* --version version *
* *
* Commands *
* *
* help - list tables *
* help table - list columns in table *
* help help - list all help options *
* *
* Examples *
* *
* isql WebDB MyID MyPWD -w < My.sql *
* *
* Each line in My.sql must contain *
* exactly 1 SQL command except for the *
* last line which must be blank (unless *
* -n option specified). *
* *
**********************************************
I experimented with the -d flag, e.g: no -d flag in a dummy query:
cat bla.sql
select 1 as bla, 2 as ha
isql -v -c -b BQ < bla.sql
+---------------------+---------------------+
| bla | ha |
+---------------------+---------------------+
| 1 | 2 |
+---------------------+---------------------+
SQLRowCount returns -1
1 rows fetched
I can make csv using -d, like so:
isql -v -c -b -d, BQ < bla.sql
bla,ha
1,2
But my real query will contain json structure data. What is the appropriate deliminator to use with the -d flag here?