Hi, I recently analyzed a large batch of files (2145) with [3.5.2 284]. One thing that seemed odd: - analyzing itself took under an hour, roughly (with parallel processing, on a i7 7700HQ). Not super speedy, but all good there. - after each analyze step (Traktor seems to process these in 499 file batches [?] ) there was an extra step where Traktor seemingly was writing the metadata (as per the readout...) of each batch only after the analyze step of the same batch was finished. - what felt off is: the writing step, following the analyze process, took longer than the analyze itself... all in all the above batch took around 2.5 hrs. So writing the data adds another 1.5 hrs on top seems weird, especially when I can rewrite the metadata of the same batch of 2145 files via python (on the same external magnetic HDD / not SSD) in 34.6 seconds... [so it's not the drive either...] Observations/issues: - Traktor is taking over 90 minutes to write the metadata of 2145 files, whereas this can be done in less than a minute. - in the past I've occasionally had unexplained metadata loss on files that were recently analyzed, and I have a hunch, having analyze & tag writing as 2 processes might have been a large factor in this, that Traktor was shut down while still [silently ? in past versions maybe, as in [3.5.2 284] I can now see a readout for each file written...]. Just a thought that writing the analyze results should be more closely tied to the analyze step. Not a high prio [for me], just an oddity I noticed. Churs. c.