In MS SQL Server, you use SSMS (SQL Server Management Studio) to query character data from the SQL Server database which uses a non-Unicode data type. For instance, the SQL Server uses the char, varchar or the text data type.

The code page of client computer differs form code page of database.

In such circumstances, the character data may be represented incorrectly. For instance, you may face any of the following issues:

The characters are represented as the question mark. You might come across this issue if you have inserted or updated character data as non-Unicode data type prior to querying character data.

The characters are represented as corrupted. The characters of code page A are stored in non-Unicode column of code page B. Furthermore, the characters are not translated.

In such circumstances, the MDF file becomes inaccessible due to corruption and might lead to the data loss.

This crisis occurs as the characters of code page are stored in non-Unicode column of page B. In addition, the characters are not translated properly.

In SQL Server, when you use the string literal of non-Unicode data type, the string literal is changed by using the default code page of the database which is derived from collation of the database. Storing the character data of code page in a column of B code page might result into database corruption and data loss.

To stop the database corruption and work around this issue, you may attempt the following:

Try a Unicode data type in place of non-Unicode data type

Try appropriate collation for SQL database

Try binary or varbinary data type

Try a diverse tool for storing and accessing data and disable data Translate parameter

These fixes will stop the database corruption but won’t be able to repair the corrupted database. To do so, you need to use third party SQL recovery software.