Today’s topic for discussion may well be elementary for any of you out there who are seasoned pros. But I thought it was worth talking about because hey…I’ve been doing this audio thing daily for over two years and when it came up…it gave me pause.
Let me set the scene. I recently had the pleasure of doing some telephone voice prompts for a wealth management firm. Now, I actually really enjoy doing people’s telephone prompts...I’m not 100% sure why but I do think it’s kind of nifty that there’s some person out there on hold for some company and it’s MY voice talking them through their menu options. Yes, I’m the “press 1 for accounts” lady.
But on to the technical stuff…
Usually when I submit audio files for telephone systems, I send through .mp3s, at 192kbps constant, 32 bit depth and mono. If any of that sounds a bit Greek, I’ll tackle it in more depth in another show. But that’s usually what I send: .mp3s at a higher quality than they’ll need, just in case.
BUT, in this case, my client’s system couldn’t take .mp3s. It needed .wav files. No problem, I thought, since all my project files are in .wav format anyway and I would just need to save with the right filenames. I made that change and sent the files again. A couple days later, I had another note in my inbox that the .wav files were too big, and that each of them needed to come in under 2.4MB. They were each around 6.
And here’s where I got stuck.
.wav’s are uncompressed audio. Usually, when I want to make a file smaller, I save it in a compressed format. My computer is very clever and uses one of several nifty algorithms to effectively cram the same amount of audio into a smaller box. My compressed format of choice is usually the ubiquitous .mp3 format. Other formats are available, but I find that most clients prefer working with .mp3. But here I wanted to shrink the file size, but still keep the audio “uncompressed”. I’ll confess I stared at my screen for a few minutes before heading to Google for a solution.
Of course, when I found the solution on Google, I felt kind of like an idiot, because like many things, it’s obvious, when you know the answer. All I needed to do was use Audition to change the sample rate and/or bit depth of my wav file, et voila! Smaller file size.
To understand why, let’s do some math:
When I record a piece of audio, my computer samples the incoming signal at a rate I specify (for me it’s usually 44.1kHz) so when I save it as a .wav, there are 44 thousand 100 samples for every second recorded. I also usually record in mono, at a bit depth of 32, so each second of audio contains 1 channel x 32 bits x 44 thousand 100 samples, which works out to 1 million 411 thousand 200 bits of data per second of audio. To put it in more manageable terms, recording at those settings results in about 10MB of filesize per minute of audio.
If I want to shrink a .wav file by half, I can do that by dropping either the sample rate, the bit depth, or both. It’s roughly linear, so if I take the same example from before, only at a bit depth of 16, that works out to be about 5MB of filesize per minute of audio. And if I use a bit depth of 32, like I did originally, but drop the sample rate to 22.05kHz, that too results in about 5MB of filesize per minute of audio.
Now, it’s important to note that while a .wav is uncompressed audio, doing it this way will impact the quality. Chances are you and your client will not hear the difference, but technically speaking, you’re using Audition (or the program of your choice) to throw out half the information in your waveform. Something to be aware of.
Have you stumbled across a conundrum in your work that you didn’t know how to solve? Or perhaps you’ve had a similar aha! Moment that you’d like to share? Drop me a line at firstname.lastname@example.org … I’d love to hear from you.