Got this note from a client today:
I’m working with a producer on an audiobook (outside of ACX) and he wants me to send him the raw audio in what he calls 48000. What does that mean and how do I do that?
My first urge is to tell K-E to find another producer. But I won’t. Here’s the answer.
First off, never change your sampling rate for audiobook work. There’s no upside to it, because that’s not what the rest of the world wants (they want the standard 44100 sampling rate, which is more than adequate for reproducing the entire range of human hearing). It’s also ACX’s and Audible’s standard.
Anything above that is for the high end video and audio producers who make really complex music, or who have to map audio to existing video.
That’s not you.
But what does that number even mean? Turns out, it’s all about the data.
It’s the number of individual chunks your sound software slices each second of sound into, when recording. Your software (Audacity) then assigns each slice a certain value that describes the sound of that sample 1/41000 of a second (or 1/48000 if you slice it up even finer) and stores all those values in order on your hard drive.
44,100 (you might sometimes see this as 44.1k Hz) is the standard for compact discs. 48,000 (48k Hz) is the standard for video production. 96,000 (96k Hz) is the high-definition standard for professional music producers.
Let’s look at how that plays out. Or, more accurately, plays back.
So a standard CD will contain 44,100 pieces of data on its surface for each second of sound. That means that a 3 minute and 30 second song will contain 3.5 x 60 (seconds in a minute) x 44100 (samples per second) or 9,261,000 individual values that will be reconstructed into the song as you listen.
The CD spins, the laser reads all the data (very quickly), the electronics of the CD player runs the data through a Digital-to-Analog converter, and out the speakers the sound comes.
When your sound software plays back the audio you’ve recorded, it lines up each slice in order and very quickly plays them all back so you can hear your continuous sound.
So then, how does sample rate matter when you’re recording? It’s the same process, but in reverse.
When you record your voice, your AT2020 USB Plus microphone’s Analog-to-Digital electronics (which are, amazingly, packed into the bottom half of the mic itself) splits your voice up into 44,100 slices per second as you speak, and continuously delivers that data, via the microphone’s USB cable, to Audacity.
Audacity not only cleanly and efficiently stores all that sound data for further processing later, it shows a visual wave form representation on the screen as you record.
Back to K-E’s question.
If you switch your sampling rate to 48,000, you’ll get finer slices of the data, which is useful for more demanding quality jobs.
But because you’re only recording your voice, even 44,100 is overkill for audiobooks.
So, I wanted to tell K-E to suggest to the producer she’s working with that even though he is used to 48000, the audiobook industry has standardized on 44,100. And there could be potential issues down the line if the final audio isn’t produced to Audible’s standards.
However, with all those caveats, it’s actually pretty easy to set Audacity to 48000:
Just open a New project in Audacity (File >> New) then use the drop down menu in the lower left hand corner labeled Project Rate (Hz) to choose 48000.
You cannot use this drop down menu to change the sampling rate of an existing project. If you’re trying to re-sample a track of audio, select the track, then go to Tracks >> Resample… and then choose 48,000:
Then export as your newly resampled 48,000 WAV. And play the exported WAV file to make sure it sounds right.
Or, as I suggest, push back on the producer, and don’t do this at all.
I hope this helps. But I also hope you are never asked to use it when recording audiobooks.