I came across DICOM objects that contain sequences in private tags; each item in that sequence contains its own Specific Character Set (0008,0005). All Specific Character Set elements have the same value. I also found that these extra Specific Character Sets create problems in some applications. Is it allowed by DICOM standard to have multiple Specific Character Sets in a single object? If it is, what are the scope rules?
I could not find the answer in Nema documents. I also check several DICOM Conformance Statements, with the same result.
Everything about sequence encoding is written in PS3.5, chapter 7.5 and in PS3.3 for each individual sequence. In neither of these documents, you will find references to including the attribute Specific Character Set.
Long story short: This is plainly wrong.