Creating a sound effects library with metadata

I’ve spent the last 3 weeks re-building my sound effects library and how I access my sound effects. I thought I’d share some of what I learned.

Where i started from:

I ripped all my CD’s many years ago, long before metadata and the like. Most of my files are simply named, for example BBC41 01.wav. However I like to travel light so I have an mp3 version of my sound library (encoded at 320bps) on my laptop hard drive for getting things fast when I’m in rehearsals and in the theatre.

I have a Filemaker database of most of my sound effects. I got this by downloading the free search engines from Sound Ideas and Hollywood Edge and merging them together. From there I used a combination of Butler, AppleScript and Snapper to audition the sounds. I’d click on a sound effect I was interested in hearing in the Filemaker database then press F12, which would launch an AppleScript from Butler. The AppleScript I wrote would look at what I’d clicked on in Filemaker then it would locate the corresponding folder, eg BBC41, then try to click on the right file within that so that Snapper would open and display the waveform and let me listen to it, convert it, etc.

This process worked reasonably well but wasn’t particularly fast or reliable.

Moving forwards

I’ve spent years looking at solutions like SoundMiner v4Net Mix ProBaseheadLibrary MonkeyAudio Finder. But they all seemed very expensive or didn’t quite have the functions I was after. And frankly some of the websites for these very expensive products contain virtually no info about what the product does, how it does it or how you might set it up – hardly conducive to spending many hundreds of dollars on them. A few of them are dongle based, and I hate dongles. I’m often having to work in less than ideal physical locations – balancing on a stool is not uncommon in the latter days of previews, so it’s a pain having these things plugged in, and a nightmare when they get lost. But enough was enough, and there were a few plugins I wanted that necessitated the reluctant purchase of an iLok. Then Soundminer HD was released which seemed to be perfect for me- enough functionality and not too expensive.

Time now for a simplified explanation of metadata. There are essentially 3 common forms in use: BWAV (aka BWF, it is essentially a wav file with an extra section containing info about the audio), Soundminer and mp3 ID3 tags.. The first two types are associated with wav files and the latter with mp3 files. The BWAV metadata is fairly simple, and it’s read and written by lots of different programs, and in different ways. The SoundMiner metadata has more ways of categorising the sounds and has been adopted by the big library producers as a standard, but it isn’t an “open” system so fewer other programs currently understand it. Some boutique sound effects come with just BWAV, or with BWAV and Soundminer. Soundminer handily can use both forms. There’s two nice articles about metadata here and here that go into more detail.

Creating an mp3 version of your files is tricky as most programs that can convert a batch of wav files to mp3 either don’t read the metadata to start with (such as iTunes), or they don’t translate the BWav or SoundMiner data to the mp3 data. Both Soundminer and Basehead (mp3converter and Injectore respectively) make programs that deal with making mp3 versions but they’re expensive ($199).  Soundminer HD has a handy feature to help with this though. It allows you to create a library using the wav files with their BWAV or SoundMiner metadata, then you can relink to mp3 files on another drive retaining the metadata.

So with this in mind I had the following to achieve:

• Get all files into .wav format (I’d ripped some as aif’s)
• Tidy up all the filenames and folders
• Embed metadata into all the .wavs
• Create mp3 versions of all the files
• Build the SoundMiner database with the wav files
• Relink the SoundMiner database to the mp3 files on my laptop

So to do this for yourself:

• Get all files into .wav format

This is a job for iTunes.
In iTunes preferences > Import settings, choose wav files
Create a new playlist and drop all files needing conversion in this playlist. Select all and right click, choose Create WAV version. All the wav files will appear in your main library – right click on one and select Reveal in Finder to get a window with all the files in.

• Tidy up all the filenames and folders

I used a combination of NameMangler and The Big Mean Folder Machine v2 to change the filenames and sort files into folders. After conversion in ITunes you end up with one folder full of files. The Big Folder Machine can be used to automatically create a bunch of folders based on the filenames and move the files into that folder. Name Mangler can then be used to clean up the filenames too, which is essential for embedding metadata.

• Embed metadata into the .wavs

This is one of the trickier aspects to describe so I made a little video tutorial. The Excel formulas mentioned are below.

Vodpod videos no longer available.
Click here to see it in full-screen:

The Excel formulas I used are:

For files that had index numbers, ie 6001 01 03.wav:

=TEXT(B2,0)&” “&TEXT(C2,”00″)&”.”&TEXT(D2,”00″)&”.wav”

For files without index numbers, ie 6001 01.wav

=TEXT(B2,0)&” “&TEXT(C2,”00″)&”.wav”

I also used this to tidy up tvarious duration fields, for example from “:53: to “0:53″:


I had some files that were extracted without index numbers, but with metadata relating to different indexs. Here I needed to combine the descriptions of the various files into one. I used some Excel formulas to speed this up, though Excel has a limitation of 255 characters to a cell so I had to truncate some of the descriptions.

• Create mp3 versions of all the files

Again I used iTunes for this. It took 2 mac mini’s running full time for about 8 days to convert my entire library. For files that were 192k and 96k I used Twisted Wave‘s batch convert feature to create the mp3’s. The Big Mean Folder Machine then sorted the 85,000+ mp3 files into folders.

• Build the SoundMiner database with the wav files
This was the simple bit – drag and drop my external drive’s wav files into SoundMiner. A very short while later I had a shiny database with all my sound effects complete with metadata.

• Relink the SoundMiner database to the mp3 files on my laptop

I quit SoundMiner and disconnected my external drive. I then restarted SoundMiner which came up with the error message: Cannot find files. Relink? With the option to ignore the file type. With the latter option checked I just had to select the folder of mp3’d sound effects on my internal hard drive and SoundMiner linked to them. Unfortunately it only seems to relink one folder at a time, and as i have one folder per CD that’s a lot of relinking.  But the end result is I now have a big shiny collection of sound effects, all with metadata embedded into them, which makes for faster searching auditioning and being able to seartch my entire collection.

The next step is for me to go though and add all the sounds I’ve created over the years…

PS I’m not affiliated or associated to any of the companies mentioned here.


10 thoughts on “Creating a sound effects library with metadata

  1. Hi, could You explain how is Soundminer acting on tracks with indexes? Does it merge metadata (specifically descriptions) of multiple indexes to one?

    Thanks in advance

    1. Hi Vladimir,

      SoundMiner doesn’t merge metadata for files that have multiple indexes.

      You have three choices:
      1) Rip the CD again using a tool like
      This wil rip the indexes as seperate files and embed the correct metadata for each index
      This is obviously very time consuming and the software is expensive.

      2) Manually split your audio files to match each index and embed as per the method above.
      Very very time consuming.

      3) Live with having one sound file containing multiple effects and adjust the metadata.
      You can use Excel to combine the metadata for multiple indexes to relate to the one file. Excel limits you to 255 characters by default.
      This is the fastest, cheapest approach.

      With option 3 you might end up with something like this in SoundMiner:

      File: CRT13 01 Description: (1) balloon rub (2) swannee whistle (3) slide whistle

      Hope that helps.

      1. Hi there,

        Excellent tutorial, thanks!
        But the thing about indexes is really giving me a headache. I ripped my files without indexes so I have one file that needs several descriptions. I exported the index column but I’m quite an excel noob: Hoy do you add the descriptions indexed greater than 1 to the description field indexed 1 and delete the rows afterwards. there has to be some formula for this – manually it is a lot of work…

  2. Hi Gustav,
    Yes, it’s hard work with files containing multiple indexes.
    There’s no easy way to do this I’m afraid.
    You can use the Excel formula string functions to add one cell to another, for example:
    = A1 & ” 2:” & B1 & “3: ” & C1
    will add three cells together. Having done this you can copy the cell, then Paste Special : Contents, to replace the formuia with the result of the formula.
    This can work well with short index descriptions, but it’s easy to exceed the 255 character limitation of a field in Excel, and then you need to manually truncate the descriptions.

    1. Thanks! Luckily I found out that one of my friends is a total excel expert and he came up with a good solution using your formula as well but also some more and I’m happily taggi g my library now 🙂

  3. Hi Gareth – Hollywood Edge has apparently taken down or moved their downloadable Filemaker db, and the web-based search doesn’t work in a way which will allow this process. Do you have any idea how I can get the HE database? Thanks!

    1. Hi Dave. Yes it looks like they’ve had a redesign of their website and removed the FileMaker database along with it. Might be worth emailing them to ask for it as I’m sure they’ve still got it. Or if you click on the individual libraries you can download the track listing which contains all the relevant info. Though obviously that’s rather dull if you own a lot of their libraries. G

      1. Could you please upload the hollywood edge filemaker database to a free hoster? Would be of great help.

      2. Unfortunately I trashed mine after completing this project. Might be worth contacting Hollywood Edge to obtain one – or a CSV list of the databases you require.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s