Hi. My current player only recognises multiple genres in a file, if they are in a single tag-field (not sure if I'm using correct terminology here!) and separated by ";". Fortunately most of my music library is tagged this way; but some albums, and I don't know where they all are, contain files having multiple genres as separate tag-fields (which my previous player supported), which Mp3tag indicates in the format (e.g.) "Genre1\\Genre2\\Genre3". How can I find all of these files and convert them to a single tag-field, e.g. "Genre1;Genre2;Genre3"?
I tried filtering a selection of files known to contain multiple genre tag-fields, on "%genre% HAS \\", but I don't have enough working knowledge of Mp3tag to know why this didn't work. Thanks for any suggestions.
A filter could look like this: "$if($eql($meta(genre,1),%DUMMY%),0,1)" IS 1
(you would not need to create that extra column then. The rest for merging the fields has been explained by @ryerman
Nice
I forgot to use double quotes in my initial attempts at a filter expression.
This also seems to work: "$meta(genre,1)" PRESENT
and also makes creating a temporary field unnecessary.
Do I understand this filter correctly:
The ,1 means "at least 2 two genres"? So ,2 would mean "at least 3 genres"?
(Null based index, ,0 would mean "at least 1 genre")?
That is the effect, if not the meaning.
From the documentation:
"$meta(x,n) returns the nth value of field x (where n starts at 0, e.g. $meta(artist,1))."
Given that my library contains >13000 tracks across >1200 albums, with each album in its own folder within (generally) an artist (or, for classical, recording label) folder, are there any recommendations on how such a ("global") filter-and-merge should be run? 13000 sounds like a huge amount of tracks to process at once.
I would load all the files
then apply the filter
and do the merge only for those filtered files.
If you fear that memory may become a problem, switch on the library function in Tools>Options>Library.
Excellent! I completed the whole process within 15 mins. 193 tracks were identified and "corrected". I'm still amazed that it only required two discrete actions: filter, merge.
Memory was one of my initial concerns, but, even with browser windows up at the same time, memory (8GB fitted) usage never exceeded 46%.