What’s the deal with 16 bit audio – CD’s still sound good in 2023
From a listening perspective there is nothing wrong with 16 bit audio. So why do we recommend recording and mixing in 24 bit audio formats
A fairly reasonable comparison in the Guardian was posted in 2021.
16 bit audio provides us with 96 dB of dynamic range and in reality you can’t hear all of that range all at once.
Louder sounds, mask out quieter sounds. If you are in a noisy club, how easy is it to hear the person right next to you?
In a good listening environment with high end equipment you may hear a difference. (although the difference in the CD And master versions listed in the article could also be related to a new master being produced for the hi res format)
So there’s nothing wrong with a final audio format at 16 bits.
But you shouldn’t be recording or exporting your final mixes at 16 bits.
There’s a good reason why..
24 bit audio – what’s it for then?
There’s a clue in your DAW
There’s a big clue in the way 24 bit audio meters look (usually)…
Take a look at your DAW and look at the channel strip meters
Or take a look at the ones from Harrison Consoles Mixbus 32c v9. (this is what I use for recording mixing and mastering – no affiliation though)
As opposed to 0 being the highest value on the scale it is 6.
In 16 bit audio, 0 represents unity gain – when the signal goes over 0 it clips.
It’s good practice when recording to aim for the signal to be around -12db to allow for some dynamic range, giving plenty of headroom to avoid clipping – and avoid pulling busses down just to avoid clipping when mixing.
The problem is that sometimes there’s a wild bit of audio that goes over that value. This could be the result of a harder attack on a drum, a singer who has just got much louder in a take than in sound check.
So 24 bit audio was developed to give a bit more headroom – above 0 to be able to capture that audio without clipping
But my mix down peaks at -6db — why should I export at 24 bit
Your Attractive Heading
Many people think that 44.1k digital audio looks like a bar graph sliced into 44100 bars per second – but that’s not really what’s going on .
The audio represents dots on a line graph. The DA converter then uses those to draw the curve -Sometimes the data implies that there is a peak in between 2 plot points those peaks get recreated by the DA converter.
These are called Inter-sample peaks or ISP’s. In extreme cases these can be bigger than 6db.
Often having some occur aren’t a big deal, but it’s best to let your mastering engineer work that out.
The best thing to do is record at 24 bit, so any “overs” aren’t clipped.
It’s just that simple – keep your Mastering Engineer Happy
There’s no real magic.
24 bit audio provides an important 6db over and above unity gain to keep the integrity of your recording.
When you are recording using live mics, this allows for unexpected peaks to be handled (6db doesn’t sound like much – but it is).
When mixing – the nature of digital audio means there can be an inter-sample peak that is above unity.
If you want to know more please contact me via the contact us page.
Make sure you get notified about our posts containing mixing techniques, mastering information, client success stories and occasionally receive a special offer.