Fix SpecificCharacterSet when writing DICOM data

[ivmartel]

The current code (v0.30), when anonymising, tries to write data using the input character encoding. This is not supported by the web 'TextEncoder' that only writes data as UTF-8. The written data is corrupted since the SpecificCharacterSet could not be in sync. This must be fixed to always write with SpecificCharacterSet set to UTF-8 (ISO_IR 192).

[ivmartel]

The list of character sets in the 2022a standard: part03/sect_C.12.html#sect_C.12.1.1.2