That is even easier 
It depends on what you call incorrect.
I loaded to Studio a lot of classical music albums. And this is a big issue.
Those 70% MusicBrainz IDs that you consider as being correct are still garbage metadata.
I will have for instance as a composer: “Beethoven”, “Ludvig Van Beethoven”, “L.V. Beethoven”, “Beethoven, L.V.”, “Beethoven (1770-1827)” and so on… All are correct indeed, but what a mess in the library.
As an actual case of why this should not happen: I use the date and timestamp of the music files on my Mac to see which new ones I need to copy onto my audio devices. I keep a record of the last copied file’s date and time. Then after I’ve bought many more CDs and digitally downloaded files, I copy the latest batch of new music that I’ve purchased onto my other audio devices. If AS alters the date and timestamp of the FLAC files in my music library folder then this would cause me a lot of problems. I wouldn’t know which new music files I would need to copy.
2 Likes
First of all, people did not know that AS was altering files, because they where not informed.
Damien even said (incorrectly) in this thread somewhere that the meta data in user tracks where not affected. Of course I believed that, because I (still) think Damien is a thrustworthy person. In hindsight he meant that only a MUSICBRAINZ ID was added to the file, but no other metadata was altered (or added). He wrote this very vaguely so most people assumed that no files where changed.
That files where changed I found out yesterday by reading this forum.
Lots of people (including myself) have a backup program running every day. The time I found out that my files where changed I was simply too late and the altered files where already backed up. The damage was already done.
As I mentioned somewhere else in this thread: I am listening to music via computers for years already with different programs. This is the first time in my experience that my music files where changed by a computer program without my permission or even without letting me know.
Correcting the wrong ones track by track is an almost undoable tedious task, because a MusicBrainz ID is only a number pointing to some entry in the online MusicBrainz database. I have 4 TB of data in my music library. This would mean opening every individual track or album with Picard and let Picard searching for the ID in the database and checking it all over manually again.
Basically than I am doing all the work I did for the past few years. Doing this for 4 TB of data is almost impossible.
I installed AS on the 17th of may. In this case it is much more practical to load a bunch of folders in MP3Tag, sort everything on date modified and delete the MusicBrainz Ids added on the 17th of May or later. After that rince and repeat until ready.
I didn’t mean to actively go searching for wrong MusicBrainz tags. Just fix the wrong ones when you spot it’s wrong.
Those are in the thousands. It is much easier to do it in batch as I described in a post before. There is the risk of deleting a couple of Id’s that should not be deleted, but than it is in a handfull of files instead of thousands. Works better for me
1 Like
I know what you mean. Classical music is always a mess with tagging. Even Roon (which is considered better in tagging classical music than other programs) a lot of it is a mess.
In case of classical music I prefer to remove those MusicBrainz ID tags al together. Also in this case AS added LOADS of tags to my classical music files. In this case I think I will follow your lead and delete them all too Another reason why AS should not do this automatically (or at all).
You keep using the word “just,” and I have a rule that when someone says “you just have to do this,” I multiply the anticipated effort by an order of magnitude. 
Regarding backup: What if both your local backups have the wrong tags, and only the cloud backup is reliably correct? Download terabytes of data? Many backup companies call this a “restore “ and charge a lot for it.
Regarding “just fix the files as you see them” - the incorrect metadata messes up your searches for music, and who wants to spend 10-15 minutes of every listening session fixing data anyway?
Regarding local backup: OK, I want to back up my most recent albums onto SD cards. Which files, of the 60% of my files recently updated, are actually new?
There are actual problems here. You might think about helping people to fix them rather than saying they aren’t real.
4 Likes
Roon does not write to the files and does not offer you an option to order it to write to the files.
In addition to A+ 2.5, I’m a user of JRiver.
JRiver allows you in its settings an option to allow it to write to your tags or to deny it. If you choose to deny it to write to your tags, it creates in the folders of your albums an XML file for each track. In this XML file, it writes the data related to the track, such as lyrics, etc…
When Damien will fix this issue, he should check how the other players are functioning.
I wouldn’t backup audio files directly to the cloud directly on every update. When I work on the metadata, I might update the same file file multiple times. I wouldn’t want every update to trigger upload.
One should always have a local backup that is schedule based. I use rsync for that.
I know that Roon does not alter the original files at all, but writes everything to its internal database. Which is preferable in my opinion, because if anything goes wrong it only affects Roon and not my music files. So Roon does not have to ask me for permission because it does not alter my files.
Also Roon has a ‘backup’ option in its settings. There you can schedule a backup for Roons database. If something goes wrong an older database can be restored from within Roon.
Foobar also puts everything in its internal database files and never changes your music files either (unless you explicitely change their metadata within Foobar of course).
I do not use JRiver, but I also use KODI, which offers a lot of options how to store the metadata info. But also KODI does not change the music or video files without permission.
1 Like
I don’t back up to the cloud, I have a local backup and still both my files and the backup were affected, because my local backup is on enterprise grade RAID-1 drives.
That’s why I backup to the NAS, but also periodically to an external hard drive.
1 Like
I backup from the NAS (where my music library is) to another external drive once a day (scheduled). And from that external drive to the cloud (Open Drive) once a day (scheduled) also. Both my backups where affected So now I am correcting my tags on the NAS. After that the backups will take care of the rest.
I’ve never seen a player, and I tested many, that changes the tags of the files by itself. I don’t believe it exists.
As someone said on this thread: AS is a premier!
Damien is very talented, maybe even a genius, regarding sound. But he did an amateurish work by allowing AS to write automatically to the files.
2 Likes
I have many instances where the data is just plain wrong!
You should backup to the external manually, on demand, and than to the cloud automatically. It would be safer in case if issues.
1 Like
Have been playing music for the past 4 hours.
Surprise surprise, AS has crashed yet again.
To be honest, sounded absolutely flawless prior to a restart of my NUC server.
After a restart the analysation of my files has started yet again.
The analysation hasn’t resumed from the last shut-down as I/we were assured it would.
It’s begun from the start. AGAIN.
Pathetic.
Any newbies wanting to ‘Beta test’ this software - don’t bother. Leave it a couple of months.
It may actually work by then (although I wouldn’t hold my breath).
Such a shame.
I read on another thread a frustrated message of a member of the forum who fixed manually wrong metadata in the library. But AS ignored his inputs and continued to display the wrong metadata. And he could not understand why.
This could be because of the wrong MusicBrainz IDs tags that remain in the tracks and AS continues to read.
With respect, Doudou, a Raid mirror is not a backup, as is usually and widely meant.
It is by its nature a functionality to make one set of data more robust by providing redundancy.
I perfectly understand and empathise with you, but your predicament is self inflicted.
Maybe it is your turn to ponder on what lesson you can learn from this experience; we all do, far more often than we would wish.
1 Like