Tracks listed twice

[ccomley]

Most odd.

My library has been created by MMJB. But the library itself is just a tree full of MP3 files.

When I load up iTunes (mac or PC version) and say “go scan R: for music” it lists every file twice.

Well as I prefer not to use iTune I don’t really care. Just thought it odd.

But

Now Iv’e sent Firefly to list all the MP3s in my library. And it *too* has every file twice- when I select an album, it builds a queue which has each track in it twice!

Any idea what’s going on? I’m *sure* that each track exists only once within the directory tree that the server’s been pointed at. There is a second copy of the entire music tree (it’s called a backup!) but Firefly can’t see it. (It’s on a NAS, when I *tried* to get Firefly to see it I failed!)

[rpedde]

@ccomley wrote:

Most odd.

My library has been created by MMJB. But the library itself is just a tree full of MP3 files.

When I load up iTunes (mac or PC version) and say “go scan R: for music” it lists every file twice.

Well as I prefer not to use iTune I don’t really care. Just thought it odd.

But

Now Iv’e sent Firefly to list all the MP3s in my library. And it *too* has every file twice- when I select an album, it builds a queue which has each track in it twice!

Any idea what’s going on? I’m *sure* that each track exists only once within the directory tree that the server’s been pointed at. There is a second copy of the entire music tree (it’s called a backup!) but Firefly can’t see it. (It’s on a NAS, when I *tried* to get Firefly to see it I failed!)

Is it possible that there is a symlink in there somewhere pointing back to the tree?

You can look at the database itself in firefly and see where they are… I’m not sure if you are using windows or mac or some other unix for the firefly server, but assuming the server is unix, you might do something like this on the firefly server:

foo@server:~$ sqlite /path/to/firefly/db/songs.db

sqlite> select path from songs where title like '%some song%'

Obviously replace the “some song” with a unique fragement of a song title, and it should tell you what paths it found it at. That will give you an idea of what’s up.

— Ron

[ccomley]

This is Windows firefly running on an XP box. Is there a similar way to peep at the data tables? Even if I have to do an “export” to a text file or summat? I do have Linux boxes but nt in the house.

The “invisble” data is on a NAS box which I’ve yet to work out how to get Firefly to access when running as a service, so unless I’ve fixed that without realising it, I’m pretty sure Firefly can only see the copy of the NAS data subdir that I copied to the XP box local drive.

[rpedde]

@ccomley wrote:

This is Windows firefly running on an XP box. Is there a similar way to peep at the data tables? Even if I have to do an “export” to a text file or summat? I do have Linux boxes but nt in the house.

The “invisble” data is on a NAS box which I’ve yet to work out how to get Firefly to access when running as a service, so unless I’ve fixed that without realising it, I’m pretty sure Firefly can only see the copy of the NAS data subdir that I copied to the XP box local drive.

Yeah, you can use the sqlite binary from here:

http://www.sqlite.org/sqlite-2_8_17.zip

It’s a command line program, docs are here:

http://www.sqlite.org/sqlite.html

But it will work just as above, only from a dos prompt:

c:>cd program filesfirefly media server

c:program filesfirefly media server>sqlite songs.db

sqlite> ... etc ...

— Ron

[ccomley]

Wow – nice compact piece of code. Reminds me of the good old days.

Anyway it reveals taht as well as e:musicthat the system has found and catalogued a whole bunch of MP3 files from e:recycled.

The reason there are a bunch of MP3 files in there is probably down to a factor of the software I use to repllicate the content of the MP3 store on the NAS (SecondCopy) and teh fact that the NAS is linux based, which has caused problems with timestamps not *quite* lining up. SecondCopy authors have a fix for this, which I’ll implement and then test again.

meantime – does a “full scan” in firefly *purge* from Songs.db stuff which it can no longer find on disk? Or would I be better songs.db by hand (either see if I can remember enough SQL to create a suitable delete statement or just delete the lot and let a Scan re-build the list)?

Cheers…

[rpedde]

@ccomley wrote:

meantime – does a “full scan” in firefly *purge* from Songs.db stuff which it can no longer find on disk? Or would I be better songs.db by hand (either see if I can remember enough SQL to create a suitable delete statement or just delete the lot and let a Scan re-build the list)?

A full scan does indeed delete everything from the database and start from scratch. Although, a regular scan would recognize that the songs were gone and delete them from the database, so either would work. Really, the difference is whether you want to force it to update the song metadata.

A regular scan will just update any changed files, and remove any deleted files. But a full scan will update ALL files.

— Ron

[ccomley]

Hah.

It was a combo-fault.

The NAS box is Linux based and the timestamp on a file can be a second or two out when viewed from Windows.

The prog which copies the files *to* the NAS box from the machine I used ot use to RIP didn’t know this so instead of copying new files once a day, it would re-copy ALL files once a day.

The NAS box has a built-in “wastebin” function and kept copies of some of the old versions of the above re-re-recopy process.

The prog I used to create the “local” library copy for Firefly to use copied taht too, and Firefly indexed it.

Having tracked it down, “spare” copies deleted in no time flat, and it didn’t take firefly more than about 45 seconds to re-scan the whole lot and dump out the dupes from the d/b, confirmed by repeating the SQlite query.

Thanks for the input.

I’m assuming it’s why iTunes had dupe entries too, I’ll test that out if I can be botehred to fireup iTunes again…

I *assume*, by the way, that when the advertising for Firefly claims it provides the link between iTunes and ROKU SoundBridge, it’s *actually* providing a link from the stored MP3s and SoundBridge, with iTunes being just one way to create, store and manage the MP3s? I don’t *need* iTunes if I hate it?

[grommet]

Firefly doesn’t need iTunes… but it can use the database XML for iTunes Playlists and to augment any tags/metadata. (It’s also mandatory to get metadata for AIFF and WAV files, which are tagless from Firefly’s point of view.)

[paulwa]

I’m running Firefly on WinXP, with my music tracks sat on a Netgear SC101 NAS drive. I’ve copied my iTunes Music Library.XML file into the same folder as the music files. Firefly seems to be scanning every song twice. The song count on the admin web page doubles and I can see each track twice on my Soundbridge.

When I delete the XML file from my NAS drive I will then see only one copy of the music files, but I also lose my iTunes playlists. If I build a smart playlist with Firefly, to replace an iTunes playlist, it ends up being empty.

What setup should I use so that I end up with only one visible copy of all songs, plus use of playlists (iTunes or Firefly) that aren’t empty?

Thanks

[ccomley]

Why are you copying the XML file to the NAS box?

Firefly doesn’t need it, it just loads the MP3 files it finds into its index. If you *have* to add the XML file can you put it somewhere NOT included in Firefly’s scan area?

[Author]

Posts