Hey all, My collection is almost exclusively ALAC (too many iDevices to go FLAC, sadly) and each time I upgrade to a new version of Traktor I have to reanalyse/beatgrid/cue point several files (up to 10% of my collection) because the NML database migration* loses data from some of my songs. This is most tiresome. This doesn't appear to happen with file formats where the grid/cues etc are stored directly in the file metadata (eg MP3), and I was just wondering why Traktor doesn't store its data in ALACs. Obviously the iTunes metadata standard is different to, say, ID3, but clearly it can be read and written by Traktor (e.g. for BPM) so I don't see why the core collection data isn't stored in a custom atom of the file data. Has anyone else had any experience with this? * it has occurred to me to manually (or with a Ruby script or similar) copy the entries from the old NML file, but I don't want to corrupt the new one and haven't found any documentation of the NML structure.