Organize your mp3 collection – tutorial – in works
Fullmooninu Tutorials – Music (but mostly MP3) Organizing
Well, i was messing around after purchasing MusicIP and figured i should alter this massive wall of text to something less scary.
So i was thinkingi should make two versions.
So… here’s the thing, im gona actually produce a light tutorial on music organizing, all centered on MusicIP. MusicIP is like, ossum, and puts everything else the web has to offer out on a corner.
I’ll try to explain why. As in the previous version of this tutorial it had little exposure.
Alpha Quick and Dirty Version
Okay Guys, I was telling some of you folks i would write some tutorials for this and that.
So here’s what, im going to do one right now… watch it grow while i use this forum as an excuse for the hiatus
This tutorial is for anyone who has a large (messy) music collection and wants to learn the basics to keeping such a collection organized, and use all that organizing to your personal benefit.
It’s mainly directed to the organizing of MP3 files, mostly because thats what the gross of my collection is, and because MP3 files, still present a very good quality/size ratio – specially when you are speaking about a LameMP3 encode at 192kbps
(which is what i consider the current “sweet spot” for audio files, by the way)
First of all, you have to understand how i think: I like my music files to be properly organized, i make sure they aren’t corrupt, aren’t low quality and that i don’t have any duplicates in my collection.
I organize music using only two factors. Artist and Track title.
Some people like to work with Albums, others like to work with Genres.
Nonetheless, this guide can help you organize your music, whatever the way you prefer it to be organized.
Steps
The steps are basically these, in which i think is the most logical order:
1- Set file attributes
2- Check for duplicates
3- Check for corruption
4- Check for quality
5- Reversible Normalize
6- Tag and Organize
7- Generate Smart Playlists, using spectrum analysis
8- Check for duplicates and errors, using spectrum analysis.
1. Set attributes
Reasoning:
So, we’re going to work with your mp3 files.
We must prepare them first. Don’t want to get any errors further on because some of them are set as read-only files.
This will be a quick tutorial on how to set attributes in a windows environment.
Example of Method:
Lets say I keep all my organized music files inside directory d:\music\ and my unorganized (and fresh) music files inside d:\downloads\music
Open a command shell and write:
example:
attrib -R +A -S -H d:\music\*.mp3 /s

This will set the proper file attributes for all mp3 files inside the directory d:\music and respective subdirectories.
Do the same for the other directories:
attrib -R +A -S -H d:\downloads\music\*.mp3 /s
And for other file types you may have
attrib -R +A -S -H d:\downloads\music\*.ogg /s
Advantages:
By doing this you will be decreasing the errors you will have to deal with when you further organize your collection
Disadvantages:
None.
2. Check for file duplicates
This is probably the most basic thing you can do for any collection.
There are lots of pieces of software than can detect file duplicates. Just use anyone your familiar with.
Make sure you only delete exact duplicates, and always try to keep the one with the file name that is the most correct and delete the others.
I like to use CloneSpy http://www.clonespy.com/
Download and install the latest stable version, and open it.
Go to File -> Options -> Files
Enable Restrict to files or file patterns , and Add Mask. Type in *.mp3 , *.ogg , *.wma , *.flac , etc .
You should have something like this.

Leave everything else in default for now. Click Add Folder, and all all the folders that contain your audio collection. Subdirectories are included by default, no need to add those.
You should get something like this:

Just click on Start Scanning.
Voilá! Our first duplicate. Just select all but one, and click on Delete all marked files.

Keep deleting the duplicates until the scan is complete. Clonespy allows an automatic procedure, but i would recommend you try it the manual way first.
Advantages:
By doing this first, you will be cleaning your collection of duplicates, reducing the number of files you have to deal with, get more free space on your hard disk, and will avoid errors further on such as filename collisions.
Disadvantages:
- If you have a huge collection this should take you a while.
- Let’s say you have the exact mp3 file, but with 2 different ID tags: Clonespy will not see it as a duplicate.
Tip:
Spoiler:
3. Check for Corruption
You just got an audio file. How do you make sure its not corrupted ?
You probably have a few of these lingering around your collection. Files that when played, give out a *beep* or a *blop* here and there.
You’ll want to check your files for errors.
I’ve tested a few pieces of software created for this purpose. Right now i only trust Checkmate MP3 checker http://checkmate.linuxonly.nl/
I only care about doing this step for MP3 files. The reason being my collection dates back to Napster days and as such i built it from .mp3 files and rips off my CD collection, but mostly because i never found any good generic error checking software.
Download the latest stable version, install and let’s use it.
Go to a directory containing MP3 files. I recommend using this in your default download directory, where you probably have a lot of mp3 files lingering.

Select them all, by doing the default english Windows “select all” command Ctrl+A – this way the program only selects all .mp3 files
Press Enter

In this example, we found two corrupted mp3 files.
The problems that will be detected by checkmate
- Broken, resulting from errors in encoding or a broken file
- Not a mp3 file, this means your file isn’t a mp3. Some of these wont even be music files. Try and play these with your player and check their properties.
Checkmate will detect errors. But it wont fix them.
If you want to improve your collection, you should:
If broken: get a new clean version of the song from somewhere and delete your old one;
If invalid: find out the real extension for the file and change it.. or get a new clean version of it, and delete the old one.
Advantages:
- Cleaning your collection of corrupted files and files with an invalid extension, is a good way of improving the overall quality, and to avoid further errors when organizing your files.
Disadvantages:
- You should have noticed a thing: this softwarei recommend is still in development in terms of everything but mp3 error checking – in which it excels.
The worse part being: It’s not capable of working with subdirectories, so you will have to analyze all your music directories one by one.
For a big collection, this will take ages. So unless you’re really patient, you can skip this step.
In the end of this tutorial, there will be presented another way of checking for errors in all your files – but this other way is more a complement than a replacement, as the algorithm is not as good as this one.
Tip:
I use this software to check new mp3, before i move them to my collection. I definitely recommend this approach.
Ill try and add some other software for this function later. Recommendations are warmly accepted 
4. Check for quality
Still about neatness.
This is the part where we check the general quality of your collection.
We checked for file corruption, now we check for quality.
The software for this?
Encspot, by GuerillaSoft
They released this for years, and it had a Free and Professional, paid version.
But before they went under, they released the paid version as free ware.
They dont have an official site anymore, and a quick Google search couldn’t find the latest version, so i cant put a link here right now. Will appreciate some help finding the file (or hosting the file, private message me if you wana)
So, download enc spot pro and install it
It will ask for a serial, which was provided for free by guerillasoft. (this will be included in the download)

Now, this piece of software is on another level….
It will read the first frames of every mp3 you feed it, and tell you its properties, including: its relative quality.
So select a dir and check it out.
Green means good, red means bad, yellow is so so.
This scale is based on used encoder and file bitrate.
What you do is: browse through your collection (doesn’t support subdirectories), and check out the red files. You should really replace these with better quality versions.
You’ll realize, if you have no experience in this, that you had an actual tendency for liking the songs in red less, and liking the songs in green more.
Amazing, isn’t it ? One of the factors that makes you like a music file less, is its inferior quality.
Same could be said about the songs that have errors and the songs that don’t.
Proof: If you ask someone to make a compilation of their favorite mp3, they will give you a sample of their collection that contains a lot LESS broken files and bad quality files than the overall average in their collection 
You’ll want to remake your collection by getting better (error free versions of these songs) and then AFTER listening to the better version, you can delete the inferior scum -_-
Joy.
At this point you’ll realize you’re deleting some corrupt songs and inferior quality songs and not getting replacements… this means you never really liked them.
Unless you’re one of those guys that collects albums instead of songs, this will actually be a good thing: this is also organizing your collection.
Advantages:
- Getting better sounding versions of inferior quality songs.
Disadvantages:
- Since this is a very manual thing, you’ll probably want to be relaxed about it, and do it bit by bit.
Tip: when you have two versions of the same song, that dont check out as true duplicate, this is the software you should use to see which is the better version
5. Reversible Normalize
Problem: some music files are louder, some are quieter. When playing them, either you analyze them beforehand so they play at the same volume, or you’ll eventually have to reach for the volume.
This is where you make all your MP3 the same mean volume, so you wont have to keep changing the volume when you pass from a loud song to a more silent one.
The reversible means the process is completely lossless in the sense that it can be undone without any loss of information.
Note: This is NOT the way music is played in CDs. In CDs some tracks are louder and others are more silent. But this not practical for a large collection played on shuffle.
For this i recommend MP3Gain http://mp3gain.sourceforge.net/
Download the latest stable version, install it and open it.
Now, let’s set it right
The default value mean sound is 89dB
I find this too low for must uses, so i recommend you set it to 91dB

In the options make sure the “subdirectorios” and “dont clip when applying gain” options are set

Add a directory.

Keep adding all the directories you want. Add your whole drives (C: D: ) if you want.
Then press Track Analysis
When finished, the log report will certainly report a few error files. The files in the log report will not be normalized because MP3Gain will report them as containing errors or not being valid mp3 files. They are a minority but if you have the patience you can investigate those files with an error checker or a quality checker and see whats up. 
After the analysis is finished, and you had a quick look at the error report (if any), press the apply Track Gain and MP3Gain will now finally work on your, normalizing your entire collection. You can try this with different gains, if you find 91dB to be too low or too high.
6. Tag and Organize:
So you got a bunch of files, spread everywhere, some with incorrect names, some with incorrect artist and trackname, most of them have a filename format you don’t like, and you want to fix this ?
Sure, you could just add them to a media library, and use that to find the files whenever you wanted. But you’ll still have lots of incorrect information about.
Well, you could correct the information yourself.
But why should you, if you can use an online database to do it for you. 
Meet Music Brainz.
Music Brainz is a tagger that uses a checksum database and checks for file similarity. The checksum (and file profile) are submitted to the database (that withstands by holding info on most mp3 file in the world). The files are identified by album and artist and whatnot, and it gives the file the proper tag, filename and even a directory of its own. All of which can be configured to your taste.
Sweet non ?
I think so. *schlerp* Really sweet.
For the newfags: This technology was initially applied on a software called Moodlogic, which also generated smartplaylists. But thats another topic. I was there on the dead Moodlogic forums when it happened: the brain behind the moodlogic technology went to work for amazon.com (traitor =O) so they could increase their profits and a few starving guys joined together to create musicbrainz as an alternative.
I watched this baby grow, and have my way of using it. Ill provide as an example my personal way of doing it:
Go to the MusicBrainz webpage http://musicbrainz.org/
Make an account – you are going to need it later.
You could use mine, but i don’t want you to.
Get your own.
There’s several programs that access the MusicBrainz database. You can take your pick. I’m going to go for the free one: Picard Tagger
Download and install the latest stable version.
Open it and we’re going to configure it now.
Make sure you grant it access to the web, this is just a client for the database. Without the internets it’s worthless.
Go to the menu Options->Options
General:

Type in that login and password you got when registering at the site.
Also select auto
Metadata:
I like to “translate foreign artist names to english where possible”
leave the rest in default
Tags:
You can leave it in default: attention, some players, mostly portable ones, dont like the default text encoding, utf-8. But ill update this part later. For now just leave everything as default.
Cover art:
Fuck cover art, i disable this. But if you want jpeg files hidden inside your mp3 files, it’s up to you.
File naming:
The good stuff. Now dont get scared. It’s just bit of script with just a few possible variables.
Click the preview button to see how a mp3 file will be renamed, by default.
Looks good ?
I dont think so, i like my files to be like this Artist-Songname.mp3 .
This is because i personally just want the filename to tell me these two things. Who is playing and what they/he/she is playing.
Some people like to read more things from teh file name. Year, album, gender and all that.
It’s about personal taste. And for simplicity effect it’s what i recommend.
If you click the help button and then click on file naming you’ll get instruction on how to change the result to whatever you like.
Don’t forget to click the preview button to see if it works.
Plus, it is also in this field that you can work out the magic to organize a completely unorganied music collection, just by using the default values, and activating the Move Files option, musicbrainz can organize all your mp3 files, neatly by directory, give them a proper filename and proper tag.
The default is $if2(%albumartist%,%artist%)/%album%/$num(%tracknumber%,2) %title%
im going to translate this
It says, if there’s an recognizable artist for the album of the song, create a directory with the name of the artist, then create a subdirectory with the name of the album, then move the file there and change it to: track number(with two digits) title of the song. (like in the example).
If there isnt a recognizable artist for the album, just create a directory with the name of the artist of that track, then create a subdirectory with the name of the album and the rest is the same
I dont like this, so ill change it to the values i like.
If i was dealing with a completely unorganized collection i would select this.
%artist%/%artist-%title%
and select a folder to move your collection to.
This way only 1 directory will be created per artist, and its a lot easier to navigate IMO.
If the collection is already organized in a certain way. Just use the %artist-%title% thing i recommend, and dont select move files.
If you liked the default way of changing the filename but dont want it to be moving your songs about, change it to %album%/$num(%tracknumber%,2) %title%
I also select “replace non-ascii” characters, but thats because i dont like it when some players cant read this. You know these right? Stuff like é and õ and â and the like, are replaced by their more simple incorrect versions without acentuation.
And You’re done…
Close options.
Let’s try it out.
Select add folder.
Select a folder with a few songs, let’s essay the result before going for your full collection, mm kay ?
They should be added in the left column
Select a few and click scan.
They mostly should be identified, and moved to the right window.
Select all the albums and click “save”.
The files will be renamed and tagged properly. (unless they are not well identified, which can happen sometimes)
They will also be neatly organized by folders, if you asked for it.
So go and see for yourself. 
7. Generate Smart Playlists:
So now you got yourself a error free, true duplicate free, properly tagged and organized collection.
Im proud of you, Junior.
So what can you do with it ?
Good question.
How about we process your collection to a database based on song similarty and start generating playlists from that database based.
This is, we can select a few songs from the database that includes all the musics in your pc, and from those we can select a playlist that includes 20 or 30 songs that similar to that one?
Pretty sweet, i think *schlerp*
Now,
we’re gona use the latest version of this technology, MusicIP Mixer
Download and install the latest stable version of musicip mixer.
The free version is just fine, even with the 75 song per mix limit, i think. But the pro version has its advantages.
Open it
Since this is the first time you use it, it will ask you for the folder where you keep your music
It’s irrevelant if you give it one or not. You can do it later.
The place to add songs to the library is in
Library -> Add songs
Just add some songs, and it will start working. Attention, part of its database generating engine is online, so for adding songs to your database you need an internet connection.
Leave it working on your songs and let’s check out the options.
There is some stuff you can do in the Options, but nothing special, so let’s just use the defaults.
Just select a song or songs, and you can click the mix button to generate a mix from those songs.
Even the shuffle is intelligent instead of random. By default the program tries to minize the variance between songs in a shuffle.
Then press play and start enjoying your collection
8.
MusicIP Mixer is a great mix creator and what i use as default player.
But it does a lot more..
For instances, it can actually use the similarity database to find duplicate songs. The songs that arent real file duplicates but are 2 different encodes of the same song ?
No problem for MusicIP just select Library->List Duplicates
If you click List unalysable it will list all the songs with errors, that should be deleted and replaced.
(not all of them are error songs, Unalysable songs are sometimes songs that are identified as non music)
Finally, if you configure the options accordingly, you can use the Fix Tags button to tag your files properly. This is the best Tag fixer around. It will find without error the names, tracks, albums, year and genre of most of the songs everywhere.
It will tag UTF names properly, but the catch is. It doesn’t work that well with UTF, so it has a slight problem with Asian filenames =/.
On the other hand, picard tries to convert Asian chars to ASCII. So i guess thats a bit of a gray area.
Other than that, if there is one software of all these that i recommend, its MusicIP
It’s that good.
| Print article | This entry was posted by fullmooninu on 2009/07/23 at 16:42, and is filed under useful. Follow any responses to this post through RSS 2.0. You can leave a response or trackback from your own site. |

about 4 months ago
Gostei. Muito.
Agora vou testar uma ou duas brincadeiras e ver se funca
about 4 weeks ago
Please test at your will. I’m sorry it’s so fragmented. I basically wrote without looking back.