Written by Javier Zúmer
I once heard that musicians need to learn the notes first to then “forget them” so they can focus on the performance. I see sound design in a similar way. When creativity is flowing, you don´t want to be slowed down by hard to find or low quality FXs. Your sounds are your tools and you need them sharp, ready to go and easy to access so you can focus on the story, the characters and the information and emotions you need to convey.
This is especially important as you become more experienced, since your library gets bigger and your deadlines tighter. So this is a brief account of everything I wish I’ve known when I started building my library.
Backups & Folder Structure • Editing and Mastering • Metadata • Rich, Meaningful Information • Consistency & Naming Conventions • Keywords • Working with your own libraries • Working with third party libraries • Library Management Software
Backups & Folder Structure
First things first: Unless you are quite old-school, your library is probably just a collection of digital information stored in a drive. We need to make sure this is accessible and secure. I don’t have a dedicated drive for my libraries, mainly because hard drive capacity is huge these days (or maybe my library is not that big) but that is definitely a good option. You can really cram in a lot of sounds in 5 TB.
Make sure you have at least a copy of your library, two is ideal, one of them sleeping in a different building or in an outside server.
In my case, I just use a folder on my “Work” drive that I back up using Apple’s Time Machine. You can also use other backup methods, but Time Machine is easy, efficient (incremental copies) and safe. An online back up service would also be nice. In any case, make sure you have at least a copy of your library, two is ideal, one of them sleeping in a different building or in an outside server.
Once things are safe, let’s see folder structure. Inside my main library folder I have subfolders indicating the origin of the libraries inside. My folders are organized like this:
- My own libraries. Libraries created by me, both for personal use or for selling.
- Free libraries. Libraries that are free, usually in both price and right to use.
- Purchased libraries. Commercially released libraries.
- Impulse responses.
- Sample based stuff like Kontakt libraries.
I like to keep all of these separated in order to always have a clear picture of the origin of each library. I have collected many free libraries and sounds throughout the years but not all of them have a CC0 license. Some of them require attribution which complicate things when working on certain projects. So for those cases, I just don’t include search results from those folders.
Having impulses and samples separated also helps when you want to get more meaningful results since these are only needed in very specific situations.
Within the folders, I always use the same naming convention. In my case this is: “Library Creator Name – Library Name”. This keeps things tidy and clear. Is also handy to have the creator name clearly referenced so you can always check their website for new libraries in case you like their sounds.
Inside of each of the library folders, I always keep the license, pictures and extra information if available. This is very good to have, especially for free libraries. I also like to make sure all the audio files are in same format, preferably uncompressed interleaved wav.
Editing and Mastering
You should have your library files prepared in the most comfortable way possible to work with them. This would apply to your own libraries and also commercial ones. When you buy a library, it is good to have a look and see that everything is ready to go. Although most purchased libraries are going to be in a very good state is always good to have a look at:
- Editing: Check that instances of sounds are well separated in either different files or within the same file with clean fades.
- Noise Reduction: This should be applied beforehand whenever needed so you don’t have to do it every time you need to use the library. Removing broad and narrow band noise, clicks, pops or wind is usually a must although be sure to not overdo it.
- Levels: There are two schools of thought here: Either leave the audio at a low broadcast level so it sits more naturally in the mix or boost levels so they are easier to monitor by themselves. I tend to prefer the first option, especially if you are monitoring SFX through your own DAW, but both can be a good choice. Just make sure you don’t have any libraries with levels that are too loud or too low.
As you may know, metadata is all that extra information that can be saved within a file. It uses many different fields like:
I believe 90% of the potential and power of metadata relies on just the first two fields: the name and the description.
I think you can see how this can be very helpful information. Having said that, don’t forget the goal is to find sounds faster so you can keep your head on the actual work. Is great to have some of those advanced fields, please go ahead and use them but I believe 90% of the potential and power of metadata relies on just the first two fields: the name and the description.
The name of the file is the only field you are always going to be able to read, no matter which software or operative system your are using. That’s why is important to make sure to have the best information possible in there. A name like “recording saturday other take” won’t help you that much, especially in the future when you don’t remember what you were doing that day.
The description, on the other hand, can be used to add additional information and details, even going beyond what the sound is, and describing what the sound could potentially be. More on this later.
Rich, Meaningful Information
The first thing we need to think about when naming a sound, is to imagine ourselves trying to find that sound in the future. Which keywords are likely going to be used to find it? Be descriptive and imaginative but use terms that are not too broad, in order to make your searches more effective. For example, avoid using terms like “noise” since this could apply to almost anything.
I usually use the name field to describe the sound in terms of what it really is. A vacuum cleaner, a dog growl, a metal chime. On the other hand, I use the description field to talk about how the sound may be used, even if this is very different to the original source. So the vacuum cleaner could be an engine element for a sci-fi vehicle, the dog growl could potentially be a monster and the chime could be used to create a spell SFX.
This is just my method, you may use the name and description fields in a different way, but I like to do it like this because I get a very clear picture of what the sound is (name) and what it could become (description).
For both names and descriptions, adjectives like intense, slow, multiple, distant, etc, are great for adding more information about the sound in question and narrow your searches down.
So, if you are looking for a big, impactful and loud metal sound and you search for “impact metal”, your search could yield hundreds of results since it is too broad. But if you used adjectives when naming files, you can search instead for “impact metal big transient resonance” and that will probably get you much closer to what you want.
Here are some examples from my own recordings for different projects and libraries. The first sound is just an ashtray spinning but the descriptions shows you how much more you could do with it. Notice how all the terms used in the names and descriptions try to be as rich as possible so the sounds are easily found in the future.
Consistency & Naming Conventions
As soon as you are dealing with multiple libraries, consistency becomes the only way to have a robust system. Ideally, you should keep a glossary with the terms you use and how you use them. For example, do you use “close” for the action of closing (like a door or a window) or do you use it to specify that something was recorded close to the microphone? As you may imagine, if you use a word in two different ways, this is going to lead to confusion and you will get loads of useless results when searching. In my case, I use “close” for the action and “near” to specify mic placement. There are many other little examples like this that you will need to discern. Again, being consistent is key, especially if you share your library with more people. Keep track of how you use your keywords in a glossary.
That last “close” example is a case of ambiguity, a word may have two or more different meanings and that can lead to irrelevant results while searching. A different case is redundancy, which happens when two words can be used to express the same thing. An example of this would be “motorcycle” and “motorbike or “beep” and “bleep”. Which one to use? Pick the one that makes more sense to you and stick with it.
Sometimes, you can take two words that mean kind of the same thing, like “engine” and “motor” and deliberately use them in different ways so they are both individually useful. For example, in my case, I use “engine” for moving vehicles like cars, boats and planes while I reserve “motor” for stationary things like a blender or a drill. Again this may feel kind of arbitrary but as long as you keep track of how you use each word and are consistent, defining these cases are going to allow you to work much faster.
The key here is to strike a balance between too little information … and too much information.
The key here is to strike a balance between too little information (searches yield a few results only) and too much information (searches yield so much options that it takes forever to browse through them). Achieving that balance will requiere time, experience and a lot of tweaking but it is absolutely worth it.
Popular on A Sound Effect right now - article continues below:
We can take certain words and give them additional meaning in oder to connect them to useful concepts. An example could be “isolated”. I use this term to imply that the sound is “clean” and can be used in almost any context. Let me elaborate. If, for example, you need a bird phrase for a establishing shot, you may search for that and find many bird sounds in different contexts like a jungle, a forest, with wind, with many other birds, etc. Unless you find something that perfectly matches your scene, chances are you are better off using a clean isolated bird phrase that is usable anywhere. That’s why I use the term “isolated” to express the idea of a sound separated from its context and usable in almost any situation.
Another example would be the pair “single” and “multiple”. I use these to specify how many sources of a sound are in a particular recording. Read that again. How many sources, not how many sounds, the difference is important. So for example, if you have formula 1 cars going by, you can use “single” if there is only one car passing by at a time, even if there are several passes along the same recording. In contrast, “multiple” would be used if there are more than one car passing by at the same time. As you can see, this would be very useful when building a racing scene.
We shouldn’t use keywords every time they technically apply, only when is relevant to point out the information.
I think defining the meaning of these keywords is very helpful but they can also be easily overused. Many sounds would be “single” and almost any foley recording is going to be “isolated” but that doesn’t mean you need to use those terms in all cases. We shouldn’t use them every time they technically apply, only when is relevant to point out the information. Striking this balance can be tricky and there are edge cases for sure but, again, it is worth it in the long run as it will speed up your searches.
Working with your own libraries
It is a great habit to try to think about your recordings as if they were going to be sold in a commercial library. Even if you are not planning to do so, this will force you to be more mindful of each phase of the process.
Every time that you go out and record, you will be tempted to just get what you need and move on to the next thing on the project. This can result in loads of recordings with not very helpful names that were maybe suited for that specific need but not for general use. It is a great habit to try to think about your recordings as if they were going to be sold in a commercial library. Even if you are not planning to do so, this will force you to be more mindful of each phase of the process.
When doing the actual recording, you should get different intensities and perspectives. Play around a bit, try different things. Of course, sometimes you don’t have the time to do this and you just need to get something done but nevertheless do it as much as possible. The extra coverage may not be needed today but believe me, it will be handy at some point in the future.
Moving on to the editing process, try to edit your effects so the result is comfortable to use as I was saying in the editing section above. I personally prefer to have a file with 10 variations of a sound and not 10 different files with one variation each. But that’s just me. I also tend to use RX to clean all sorts of noises but only when I know for sure the noise would not be good to have in any case.
Finally, when writing metadata, try to go beyond the nature of the sounds themselves and the use that you are going to give them in this particular instance. As i said earlier, ask yourself: If I needed this sound in the future, what would I tend to type to find it?
If you apply this long-term mindset whenever possible, you will see how your personal library grows not only in quantity but also in quality. Anybody can buy libraries but your personal collection is unique to you. Make it a valuable asset.
Working with third party libraries
What good is to have amazing sounds if they are named “CD_3_Track_043”? You are not going to remember that that track has a great wind for the scene your are working on but if it was named something like “wind ambience spooky howl gust trees”, that seems like a much easier file to find in the future, right?
Sometimes old libraries come with generic names like the former and/or a PDF with all the metadata. This is less than ideal since we need to eliminate as much friction as possible when working. In these cases, what I do is to use macros to rename the files with all the information from the PDF. I use Keyboard Maestro myself but any similar software would do.
Each library may have its own way of naming and organizing their sounds. It is worth going through them and changing any metadata that may get in the way of your search precision. Apply your naming conventions and rules to them.
Luckily, nowadays vendors and sound designers sell libraries that are well-labeled, but each of them may have its own way of naming and organizing their sounds. It doesn’t seem very practical to learn each of these styles, at least when you library gets larger and larger. My advice then is go through each new library you get and change any metadata that may get in the way of your search precision. Apply your naming conventions and rules. As mentioned before, you can think about this in terms of avoiding redundancies (multiple terms that have the same meaning) and ambiguities (the same term has several meanings).
I usually use search and replace functionalities to make all the metadata fit my naming conventions. For example, I use “pass by” for vehicles going by and not “drive by”. I use “speech” for voice recordings that are not musical and I use “positive” or “negative” for the emotions implied in user interface sounds. All of these little rules help me find what I want faster so it is important to go through third party libraries and make sure their metadata fits my conventions.
Library Management Software
All the advice above is applicable even if you don’t use an specific piece of software to manage your library but certainly when your collection gets big enough, it’s a huge advantage to use one of these.
They allow you to edit metadata in bulk, do more advanced searches, monitor sound effects with FX applied on the fly and instantly transfer sounds to your DAW.
Soundly (pictured above) is the one I use but you may find other options are best suited to your needs.
Which one to use? There are many options out there but here the most popular ones. It all depends on your needs:
- Soundminer is pretty much the industry standard and it offers the more advanced features, specially in metadata editing.
- Basehead is also a popular option, offering tons of professional features at a more affordable price point.
- Soundly offers a free version, which can be good if you just need something to edit metadata with but has a limit on the amount of local files you can manage. They also have a monthly paid version which allows to have unlimited databases plus gives you access to loads of free cloud based libraries. I have written a guide on how to use Soundly, if you are interested.
- Your own DAW is also an option, as some of them give you access to metadata and can perform searches.
- Finally, finder on a mac and file explorer on a PC can be useful if you make sure you use rich and comprehensive names. They would be enough to make simple searches.
All the time you put into your library today will be re-paid to you in the years to come, I assure you. Keep it healthy, growing and sharp so you can focus on the work we actually enjoy doing: delivering amazing sound design.
Please share this:
+ free sounds with every issue: