Smart search

Hi guys,

I couldn't find any answer in the FAQ section so I decided to take my chances here at the board. Since my music cullection has roughly 15,000 songs by now, I cannot start manually searching so I need some kinda smart feature.

What I'm looking for is a plug-in / script / etc. that has a smart search option to scan through the directories I point out in order to present me a list with all typos and (pseudo-) duplicates. For instance:

  • Different entries with artistname Tiësto, 3 entries saying DJ Tiësto, and 2 entries saying Tiesto. I'd like all these to be shown somehow so I can mass change them all.
  • A double entry where only one or two tag fields differ. For example the same artist/song but from a different album and with a different tracknumber.
  • Several entries of Bryan Adams where it says both Bryan with an 'y' as well as an 'i' and where it says both Adams with 'd' and with 'dd'.
  • Entries of André Rieu where it says both André with 'é' and with 'e'.
And so on... Is this possible without manually filtering every word I can come up with being a potential subject for typos / duplicates?

Mp3tag has no built in "duplicate" or "fuzzy" compare feature, you have to write your own tool, using the Mp3tag scripting language and the action feature.

These links can be of interest to step into:
Simplified character set ...
Unicode to ASCII
Duplicates ...
Wie vorgehen um doppelte per %_md5audio% finden


That's very unfortunate. Because that implies I have to program/script the 'smart' aspect myself. As I said in my post, that would mean I have to manually come up with every possible flaw there might be in my collection, and create individual solutions for every type of flaw.

I was actually hoping for some scriptkiddies around who'd find a rush in creating something like that, so that I do not have to make something myself.

Although, thanks for sharing those links!

...That's where another tool or tie-in comes in handy. Or to put it differently, isn't that the purpose of using more than a single app to handle audio files? I love Mp3Tag immensely & use regularly but use in conjunction with a few core tools that have some aspect that I prefer for certain tasks, sheer simplicity (e.g., id3ToFolder) or because of a certain UI (user interface) or look (e.g., I like TagScanner for surveying the entirety of our music library).

I scrobble all my music & got serious (aka anal) re tags as the source of the household music grew from a few online freebies to tons plus buys thru iTunes, Rhapsody AND Realplayer -- all before our now 13yo child even had his 1st iPod.

Doesn't the actual solution beg the layered complexity of what you want; doesn't the answer depend on a host of variables ... depend on a range of issues: Windows version, audio type (e.g., MP3, AAC,...), encoding (e.g., UTF-16, UTF-8) tag version (e.g., APE, ID3v2.3 or 2.4) dictated by software (e.g., foobar2000, iTunes vX, xmplay32...) and/or hardware (e.g., iPod 1st Gen, iRiver, iPod Nano), etc?

My formulaic inconsistencies (i.e., case conversion and string translation
problems) usually resolved w/regex examples from these forums or other taggers that have a feature set I prefer to handle transliteration; Tagscanner is developed by a Russian and comes setup w/character replacement examples that cover most of what I've needed. Also, Mp3 Tag Tools handles particular recurring capitalization cases well. Both help me errant "Dj" when I prefer all caps "DJ" unless it's dj BC.

Other software I've used with varying results to achieve the nirvana you seek:

Similarity: (an earlier v1.5.3 lacking newer features & fine-tuning)
Duplicate Music Files Finder
RenamerNG: good batch renamer that can handle id3; source gone so Google it
actions in MP3Ext: caveats if you actually care about strict ID3v2.x compliance

Similar to Mp3Tag & new find of mine is Kid3 which is truly cross-platform and relies on a bunch of standard Unix-y software libraries.

I find that much of dealing with the unknowns of mine or someone else's errors cascading into my finely tuned storehouse is to pick what are my "bibles" for data and then lean heavily on the wealth of Mp3tag features & tie-in to FreeDB + the sources & scripts by dano/pone/et al - especially those tied to Discogs and MusicBrainz (MB) with the latter's style guidelines setting how I tag but still with exceptions - such as for scads of video game music I have from

I use MB data via Mp3tag (or the latter's own tagger Picard) as the base (and final arbiter if there's conflict) for release info, spellings, etc but then fill in missing or different info from Discogs (and/or VGMDB) via add'l scripts that help eradicate those snarky user-induced problems. And, while I'm no great coder in the past I have even managed to tweak pone's tweaks of dano's Discogs scripts because of prefs I have for my household. (And, I still keep hoping that this tactic will help me hone my regex skillset!)

Here's how I generally handle non-artist/performer/band name inconsistencies since names are fixed mostly by Mp3tag scripts & varied web sources.

Before Musicbrainz's recent server upgradeI was fabulously happy w/the refinements that one MusicBrainz user-created plugin (lastfmplus) offered allowing me to enhance further how I use Mp3tag to restrict genres to my fixed user-generated list of 21 and put varied "style" or minor genres in the field CONTENTGROUP.

I love how this MB plugin gives me a means to standardize multiple fields for variations like R&B vs. R and B or Avant garde vs. Avantgarde vs. Avant-garde -- all of which I could accomplish differently w/i Mp3tag but I like crowdsourced solutions given that no API/database is perfect all the time. Plus, this multi-faceted attack prevents me from reinventing the wheel - on most days. (BTW, I suspect I will continue to be happy w/the MP3tag + this MB combo once the plugin dev can finish updating to a stable iteration that aligns to the new server & that plays nice with the current MB user release v0.15.)

Or ignore these ramblings and find thee a sorcerer.

Really nice! (No sarcasm here)
Yes, a lot of people look for a sorcerer although tagging is no witchcraft - just laborious. Thank God there are some tools like MP3tag that can do things that look almost like magic.
I appreciate the honest words.