I have a few questions about the Empeg MkII and what happens during certain situations...

1 - If I am using Emplode and I delete some "tune" by selecting it and hitting the DEL key, is the actual MP3 file removed from the HD on the empeg? If that file is referenced by more than one playlist what happens? If its only referenced in that one playlist what happens? How can I make sure it gets cleanly deleted?

2 - Is there a way to determine if there are any orphan MP3's using up space on the Empeg HD? Is there a need to detect this? Does the Empeg perform any kind of cleanup procedures to locate this files and eliminate them?

3 - If I have a particular "tune", and its ID3 label says the song title is "ABC", by artist "XYZ"... And then I drag and drop another MP3 from my PC onto Emplode that has the same tune and artist names...what happens? Will the MP3 file be replaced by the new one? Will there be duplicates of the same song? What uniquely identifies songs? Is there any kind of "replace" feature with Emplode and sync...or is it always assumed that I'm either deleting by selecting tunes and hitting the delete key...or adding new ones...and if they happen to have the same names in the ID3 tags...then so be it?????

4 - when I update the "tune properties" using emplode...I'm editing what would normally be in the ID3 tag. At least that's where the Empeg gets the information to begin with, but it appears to get stored in the Empeg database as well (stored in both places). So when I edit these values in Emplode and do a sync, do the actual ID3 tags of the actual MP3 files on the Empeg get updated, or is just the Empeg database updated?

5 - Near the end of the sync when the emplode dialog says something like "Updating database", what is happening? is it rebuilding the entire database based upon MP3 file ID3 tags, or what exactly is happening?

Thanks. I know some of these questions may seem funny, but I really need to know. I've been uploading lots of MP3's to my Empeg and updating my ID3 tags in many cases after the fact in Emplode. I would really like to use FTP to download all my final MP3's back to my PC with the new ID3 tag information....

Also, anyone have any good suggestions for good ID3 tag editors that make it easier to cruise through directories where more than one MP3 file has a lot of common info (eg, artist, album, year the same for every song in a folder). ??

Thanks

-steve
sjs@bstage.com

Wow, that's a lot of questions. All good questions, somehow we need to work this into a FAQ. Anyway, here we go...

If I am using Emplode and I delete some "tune" by selecting it and hitting the DEL key, is the actual MP3 file removed from the HD on the empeg?

Yes, if it's the last instance of that tune in your playlists, then it also deletes the file.

If that file is referenced by more than one playlist what happens?

The file stays, until you delete the last instance of it. Only when you delete the last reference does the file itself get deleted.

How can I make sure it gets cleanly deleted?

Delete it from all playlists in which it appears. In version 1.1, the "search" feature will help you locate all references to a given tune.

Is there a way to determine if there are any orphan MP3's using up space on the Empeg HD?

Yes, they will appear in "Unattached Items". If your "Unattached Items" is empty, then you don't have any orphans. Usually orphans will only occur if something goes wrong during a synch.

If I have a particular "tune", and its ID3 label says the song title is "ABC", by artist "XYZ"... And then I drag and drop another MP3 from my PC onto Emplode that has the same tune and artist names...what happens?

Emplode dutifully adds a second copy of the song. This is for good reason- it could be a different recording of the song, a concert version, an acoustic version, a dance mix, anything. Emplode doesn't attempt to de-dupe based on the tags, because tags can be wrong or they might not have enough data to clearly indicate if the file is/isn't unique. It doesn't try to process the binary audio data for comparison because that's a big can of worms- what happens if you wanted the file ripped with two different encoders for comparison?

If you really want a second copy of the same song in two playlists, make a copy of the song within emplode and it'll make a second reference to that song without making a physical copy of the MP3 file. But you already knew that.

What uniquely identifies songs?

At the software level, the FIDs on the empeg hard disk uniquely identify each song. If you export the database to a CSV in emplode, you can see the FIDs and see how they work.

For your own personal use, you can deliberately name different versions of a song differently so that you can uniquely identify them when scrolling through the search results. For example, I have probably four different versions of Rush's "Closer to the Heart", and I identify each one with a unique song name so I know which one I'm selecting.

is it always assumed that I'm either deleting by selecting tunes and hitting the delete key...or adding new ones...and if they happen to have the same names in the ID3 tags...then so be it?

Correct. That's a good description of how it works.

So when I edit these values in Emplode and do a sync, do the actual ID3 tags of the actual MP3 files on the Empeg get updated, or is just the Empeg database updated?

The tags are only read by emplode at the moment you drop them onto emplode. After that, the tags are never touched or changed. Everything is in the empeg's database, and that's the only thing that's edited. The only use for the MP3 tags is to fill out the initial entries when you first add them. The player software and Emplode do all their work from the database and they ignore the tags.

Near the end of the sync when the emplode dialog says something like "Updating database", what is happening?

I don't know the technical details, but I think it's making sure all the records are updated and that the database is optimized. If you think about my answer to your previous question, it makes sense that the empeg would need to churn a bit when entries are added or deleted from the database, right?

I would really like to use FTP to download all my final MP3's back to my PC with the new ID3 tag information.

Unfortunately, that's not possible because the tags are untouched. Remember that the Empeg folks deliberately avoid giving us FTP capability by default and can't support that kind of activity.

The best procedure is to get your tags in order before you send them to the Empeg. This is something that everyone keeps stressing over and over again here on the BBS.

Also, anyone have any good suggestions for good ID3 tag editors...

I'll let some other folks answer that one, as I don't have much experience with tag editors. The only thing to watch out for is to make sure that the tag editor is V2-aware so you don't run into the "badV2-goodV1" bug.

___________
Tony Fabris

Emplode dutifully adds a second copy of the song..

This is no longer strictly true in v1.1. Emplode will now detect duplicate tracks, based on all of the following tags being identical:

title, artist, source, year, duration, length

Currently, if it finds a duplicate, it skips it. It brings up a dialog box after the import, summarising what went wrong. I could probably get it to ask the user, but that means writing a Yes|Yes All|No|No All dialog box.

It's also clever enough not to duplicate playlists from folder names, under any circumstances. This means that you can pick up your entire collection and drop it onto emplode, and only the changes will be spotted.

Near the end of the sync when the emplode dialog says something like "Updating database", what is happening?

The tags used by the player (as opposed to the ID3 tags) are kept in the *1 files in the music partition. When it says "updating database", the player is compiling the *1 files into /empeg/var/database,playlists and tags. These are optimised to fit into memory (the disk's generally spun down, remember).

The player is smart enough to compile the *1 files outside of the sync. If the sync went wrong, it takes longer to load the database from disk, however. Also, since the disk is read-only at this point, it can't write the compiled version back, and has to do it every time it starts up.




Roger - not necessarily speaking for empeg

I could probably get it to ask the user, but that means writing a Yes|Yes All|No|No All dialog box.

Lazy bum.

I once had to use an application (regularly), where the programmer was too lazy to write the proper dialog box. He brought up a standard yes/no dialog instead.

There was one problem: The question wasn't a yes/no question. So the text of the dialog was really funny. First it stated the question, then it explained that you had to press Yes for one answer and No for another answer.

So several times a day, I had to answer this strangely-worded question, and because of the odd wording I often chose the wrong one.

Anyway, you should go ahead and write that dialog anyway because you're also going to need it when you implement the V1/V2 tag preferences. Right? ("This file contains both V1 and V2 tags. Which would you like to use? V1 | V2 | V1 to All | V2 to All")

Also note that another way to do it (and I prefer this method) is to have the regular buttons for yes/no, and instead of two more "all" buttons, have a single "all" checkbox that you can check.

___________
Tony Fabris

Currently, if it finds a duplicate, it skips it. It brings up a dialog box after the import, summarising what went wrong. I could probably get it to ask the user, but that means writing a Yes|Yes All|No|No All dialog box.

Wouldn't it be nice if emplode/emptool could be configured not to skip duplicates if they are in different playlists, but to make a link (as opposed to uploading actuall MP3 duplicate)? That way those of us who would like their PC-based MP3 collection directory structure to mimic empeg playlist structure could achieve this using Unix hard links (as long as tune names are same).

Or, how about retaining original filenames and i-node numbers in *1 tags and optionally use them by emptool in order to duplicate the structure, links and all?



Dragi "Bonzi" Raos
Zagreb, Croatia
Q#5196, MkII#80000376, 18GB green

Wouldn't it be nice if emplode/emptool could be configured not to skip duplicates if they are in different playlists, but to make a link (as opposed to uploading actuall MP3 duplicate)? That way those of us who would like their PC-based MP3 collection directory structure to mimic empeg playlist structure could achieve this using Unix hard links (as long as tune names are same).

Emptool already does this, as long as the duplicates are both uploaded at the same time. It does this by checking all the files it's uploading for symbolic links -- and, if it finds two symbolic links to the same file, adding a reference to the original FID rather than a new FID.

For instance, I load up my test player by using emptool's "upload -p -r" on three directories: one a link forest under "Artist", one a link forest under "Decades" (then years), and the third a link forest under "Favourites". All three link forests are automatically generated -- the first two from running id3info on everything under /mp3audio, the third by examining mserv's database for "good" and "superb" tags.

Peter

PS. There's currently a limitation in emptool that it only spots this sort of thing if the leafname is a symlink. Symlinking to a directory will cause it to upload duplicates of all the files in that directory. This behaviour would be only slightly painful to fix if people wanted it though.

Wouldn't it be nice if emplode/emptool could be configured not to skip duplicates if they are in different playlists, but to make a link (as opposed to uploading actuall MP3 duplicate)? That way those of us who would like their PC-based MP3 collection directory structure to mimic empeg playlist structure could achieve this using Unix hard links (as long as tune names are same).

A few things on this:
I don´t know about emptool (haven´t used it yet, emplode is way more easy for me to use), but emplode will not upload duplicates (as of 1.1, see other posts on this board, it will even detect them across syncs) and it will certainly not skip them by default.
And symlinks would probably be better than hardlinks, since they are easier for a program to detect (no need to use low level FS functions like detecting the starting inode of the file) and work even across filesystem or via network (as long as you use relative filenames for them).
However, for emplode, there is no such thing as a symlink or hardlink available under Windows (I don´t count *.lnk files as actual links, and the only way to do hardlinks would render the FAT useless)

Or, how about retaining original filenames and i-node numbers in *1 tags and optionally use them by emptool in order to duplicate the structure, links and all?

While retaining original filenames might be usefull (but guys@empeg keep telling that this would destroy the idea of a centralized music database that is kept on the empeg), inode numbers certainly are not. Why? Simple: The are not restored when restoring from a backup. And I always keep backups on tape.

cu,
sven
(MkII 12GB blue, #080000113)

Hmm, that's a very cool feature...the symoblic link recognition by emptool. Any ideas how I might be able to do something like that with jemptool on WinTel?

1 - If I am using Emplode and I delete some "tune" by selecting it and hitting the DEL key, is the actual MP3 file removed from the HD on the empeg? If that file is referenced by more than one playlist what happens? If its only referenced in that one playlist what happens? How can I make sure it gets cleanly deleted?

2 - Is there a way to determine if there are any orphan MP3's using up space on the Empeg HD? Is there a need to detect this? Does the Empeg perform any kind of cleanup procedures to locate this files and eliminate them?


If a track is referenced from more than one playlist it is only physically deleted from the empeg when the track has been deleted in all the playlists it is in. The empeg software goes to great lengths to avoid getting orphan tracks. Occasionally you can end up with some, but they then appear in the "Unattached items" playlist in the root of the player (only visible from emplode).

3 - If I have a particular "tune", and its ID3 label says the song title is "ABC", by artist "XYZ"... And then I drag and drop another MP3 from my PC onto Emplode that has the same tune and artist names...what happens? Will the MP3 file be replaced by the new one? Will there be duplicates of the same song? What uniquely identifies songs? Is there any kind of "replace" feature with Emplode and sync...or is it always assumed that I'm either deleting by selecting tunes and hitting the delete key...or adding new ones...and if they happen to have the same names in the ID3 tags...then so be it?????

No it won't be replaced, you will have two different copies of the tracks. Emplode and empeg does not "know" that the two files (and hence tracks) are the same.

4 - when I update the "tune properties" using emplode...I'm editing what would normally be in the ID3 tag. At least that's where the Empeg gets the information to begin with, but it appears to get stored in the Empeg database as well (stored in both places). So when I edit these values in Emplode and do a sync, do the actual ID3 tags of the actual MP3 files on the Empeg get updated, or is just the Empeg database updated?

Once the files are on the empeg editing their properties in emplode will only change them in the empeg database. You need to make sure you complete the mp3 tag values to your satisfaction before you upload them.

5 - Near the end of the sync when the emplode dialog says something like "Updating database", what is happening? is it rebuilding the entire database based upon MP3 file ID3 tags, or what exactly is happening?

It is building some sort of database on the player that is based on the properties of all of the tracks and playlists as they are on the harddisk on the player. It is not based on the MP3 files on your PC.

__
Unit serial number 47 (was 330 in the queue)...

5 - Near the end of the sync when the emplode dialog says something like "Updating database", what is happening? is it rebuilding the entire database based upon MP3 file ID3 tags, or what exactly is happening?

It's compiling the contents of the *1 tags into a static database (in /empeg/var). This static database is loaded into memory at startup (the disk is usually spun down).




Roger - not necessarily speaking for empeg

>Also, anyone have any good suggestions for good ID3 tag editors that make it easier to cruise through directories where more than one MP3 file has a lot of common info (eg, artist, album, year the same for every song in a folder). ??

I use the shareware "Shuffle2", it does global changes and has a tag wizard which is very good for compilations ripped in the older version of audiograbber.

I use MPTagger http://surf.to/mptagger . I know there are several others. Try using the search part of the bbs, I know there's been a lot of posts on this topic.

-Zeke

just say you weren't paying much attention...