Please tell me what this sentence means?

Guys I’m so new at this!!

I’m trying to record a voice over a the customer has asked for a sample which is 48kHz 32bits with audio content up to 24kHz.

Mate what does that even mean and how do I provide it to them?

If they were just asking for 48 kHz or 24 kHz then I think I could work it out but they’ve asked for both. How do I do that?

The maximum possible audio frequency in “PCM” digital audio is half the sample rate (see here for a detailed explanation:

I assume that they mean that they want the sample rate to be 48000 Hz (48 kHz), which gives an audio bandwidth of 24 kHz (half the sample rate).

Note that “32 bit” could be “signed 32-bit integer”, “unsigned 32-bit integer” or “32-bit floating point (IEEE)”.
I’ve never come across “unsigned 32-bit integer” audio, so I’d say that was very unlikely to be what they mean.
“32-bit float” (floating point) is what Audacity uses internally. They could mean this.
“Signed 32-bit integer” is used by some other apps. They could mean this.

Audacity supports exporting in either 32-bit float or signed 32-bit int (also 24-bit, 16-bit, and others). See the “Encoding” options in the WAV export dialog:

Aren’t you happy you asked?

In general, there are three common sampling rates: 44100 is the one for Audio CD and while it’s not a gift from the angels. it does OK and everybody knows what it is. 48000 is the sample rate for video. They got a later start and didn’t have to put up with the restrictions of the early CD disk.

Studios generally use 96000. As people point out constantly, you can’t hear that improvement, but it guarantees perfect multi-pass editing and you can tell the client their work is the same standard as Warner Brothers and Glen-Glenn Sound.

Either 48000 or 96000 sampling rate will give you audio transmission up to 24000Hz.

The odd duck is 32 bit format.

CD and Video use 16 bit and Studios use 24 bit. Audacity uses 32-bit Floating internally. Nobody breathing uses plain 32-bit.

But it’s not that hard to do. Make sure your system is at 48000 and it says that in the Audacity lower-left window.

Export > WAV > and 32-bit is one of the export options.

Screen Shot 2021-04-29 at 5.55.35 PM.png
I told the Mac to give me INFO on the sound file and it did.

Note in the middle bunch: 1 channel (mono), 48000 - 32 bit.

Screen Shot 2021-04-29 at 5.57.11 PM.png
The exact specification they wanted.

You still have to produce actual work of that quality. Are you doing this on the kitchen table? Good luck.

Stop me anywhere if I lose you.


If you play your cards right, they’ll accept the sound file standards and won’t ask any questions. Don’t volunteer any information.

There’s almost zero chance your microphone and system will pass high-pitch tones up to 24000 Hz. Some dogs would have trouble with that.


“Signed 32-bit” is occasionally used. For example, SoX supports signed 32-bit. It is used to achieve an extremely low digital noise floor. Example:

Most likely they want 32-bit float unless it is for some sort of scientific analysis.

SoX supports signed 32-bit.

That’s a process or tool, not a company or job. The forum post rings bells of someone whose demands are either bogus or doesn’t understand their goals.

I’m trying to record a voice over

…so the client can subject it to rigorous scientific or medical analysis? Maybe, but not likely.


Whatever polar pattern you’re using, the Yeti records it all at an impressive 16-bit/48kHz audio resolution.

That’s from the Yeti product description.

Many popular microphones will record at 48000 (video) sample rates.

Record the work, export it at 48000 32-bit, Examine the file with Mac Right-Click > Get Info, pack it up, and send it. Don’t explain anything. Make it appear like you do this every day.

You didn’t address the actual vocal quality, delivered volume or other production specifications. We assume you do do this every day and the only real question was the odd deliverable requirement.