True - we can always streamline as much as possible.

Search criteria can be set assuming that there will be false positives to be discarded and the user will have to spend time correcting and organizing his library. But isn't that what any Blisss user is willing to do?

Thanks for explaining, that makes it clearer. Definitely something that cannot be automated I think...

Issue a warning and allow to correct it or discard it permanently. But there are more case studies with this. When you have a large library (over 50,000 albums in FLAC in my case) I have repeated albums with similar artist names. My library is organized in iTunes mode, /artist name/album/ If the artist name changes (it can be written the name of only one of the musicians who have recorded the album or the name of all of them in the folder name) the album is repeated in two different folders. To fix this it is necessary to look for similarity in the name of the album folder and in the name and number of the elements in the folder, being able to exclude txt, pdf, nfo, etc, from the count. I have tried to achieve this with an app called Umbrella http://www.hyperbolicsoftware.com/Umbrella.html but it doesn't work. All the apps seek to identify equal files, same name and size or similar or equal tags. Music players like JRiver only identify matching songs. Roon improves that because it has access to a large database for which it pays royalties. And yet it can't identify more than a portion of the duplicate albums. That's why I think a better solution may be to identify the files on the outside and in a less rigid way. And then let the user decide if they are equal albums and he wants to delete one or keep it.

Interesting idea.

To what extent should an album be considered duplicate?

For example, let's say you had an original release, and an "anniversary edition" release with three extra tracks. What should be the behaviour?