best

16 bit vs. 24 bit Audio, What Should You Record At? (FAQ Series)

greetings earthlings today we're going

to be discussing the differences between

16 and 24 bit recording as well as which

bit depth you should be recording at so

first things first we need to have an

understanding of what an audio interface

or an analog to digital converter does

and all it's doing is taking an analog

signal like me speaking into this

microphone and creating a digital

version of that now there are two really

important specifications when we're

talking about analog to digital

conversion and those would be sample

rate and bit depth sample rate is

essentially telling you how many times

per second your audio interface is

taking a snapshot of the analog signal

you're recording so if you're recording

if 48 kilohertz that means that your

audio interface or your analog to

digital converter is taking 48,000

snapshots of your analog signal per

second an imperfect analogy here would

be to think of sample rate as a frame

rate on a camera how many times per

second is your camera going to capture

an image it's the exact same thing here

just with audio but this video is not

about sample rate it's about bit depth

so let's get back to that bit depth on

the other hand tells us what resolution

our audio is being recorded at or how

accurate the sound being captured

actually is adding on top of the

previous imperfect analogy we could

think of bit depth as the resolution of

your camera so what you're seeing in

person with your eyes would be analogous

to the analog signal you're recording

and then on your camera you have the

option to record at 1080p or 4k 1080p

having less detail 4k having more detail

and being more accurate of a

representation and therein lies the

difference between lower bit recordings

and higher bit recordings higher bit

recordings are going to capture a lot

more information about the sound you're

recording and therefore create a much

more accurate representation of that but

now let's go ahead and look at a

visualization of a really extreme

example so we can get an even better

understanding of what bit depth is doing

so what your interface and analog the

digital converter are doing is saying we

have sixteen possible values that this

analog signal or waveform can be since

we don't have infinite possible values

we can't perfectly digitally recreate

this sine wave so we're going to need to

simplify this information and

can think of this like averaging

rounding or even sorting for instance

this point right here is negative four

point two and it's closest to negative

four so let's go ahead and throw it in a

pile with all the other negative fours

this amplitude up here is closest to six

so let's throw it in a pile with all the

other sixes

so during the analog-to-digital

conversion your interface is doing this

simplification and that simplification

is called quantizing or quantization now

if you're still kind of confused about

what quantization is let's go ahead and

use a really popular tool to demonstrate

that even further so now I've

initialized an auto-tuned effect and

logic pro and while the human voice is

capable of creating an infinite number

of pitches the auto-tune software is

saying this pitch right here ah is

closest to a so let's go ahead and smash

it down to a this pitch right here ah

it's closest to B so let's go ahead and

bump it up to B so basically it's just

taking whatever information you put into

it and making it fit within a set number

of parameters so now that you have a

basic understanding of bit depth let's

go a little bit further and talk about

what bit depth actually means and how it

directly relates to the sound quality

you're getting so first off bit depth is

just telling you how much information is

being captured every time your interface

or analog to digital converter takes a

sample of the analog signal and as I

already mentioned the higher the bit

depth the more information is being

captured therefore the more accurate the

representation the lower the bit depth

the less information and the less

accurate the representation now let's

hear how the decrease in information

actually affects the sound as we

decrease from 24 to 16 to 8 to 4 bits so

throughout this entire video I've been

recording at 24-bit 48 kilohertz and

this is how it's been sounding now I've

decreased the bit depth to 16 bit and

you may not hear a huge difference here

and that's probably because of YouTube's

compression the volume of my voice and

the limited frequency of the human voice

now I've decreased the bit depth down to

8 bit and as you can hear it really

starts to degrade the audio you're

starting to get digital artifacts and

it's also starting to introduce a lot of

extra noise and now on

[Music]

[Music]

now just for the hell of it let's go

ahead and dive further into the theory

and look at some equations that tell us

how much information is actually being

captured so the number of values that

your analog signal can be quantized to

is calculated by taking 2 to the nth

power where n is going to be the bit

depth you're using to record at so if we

look at 16 bit we take 2 to the 16th

power and that gives us 65 and a half

thousand possible values that your

analog signal can be quantized too but

once we jump up to 24 bit and take 2 to

the 24th power we get sixteen point

seven million possible values that your

analog signal can be quantized too so to

put this into perspective you can just

think of the accuracy of an image with

65,000 total pixels and then the

accuracy of an image was sixteen point

seven million pixels the latter is going

to be much more accurate so when we look

at these two equations you can easily

determine that the 24-bit audio is going

to be a much more accurate

representation of the analog signal but

there's even more equation fun so let's

keep going now bit depth doesn't just

tell us how much information is being

captured and how accurate that

information is it also has a direct

impact on a signal-to-noise ratio so

when your interface or a 2 D converter

is quantizing the analog signal to its

closest digital value it's not just

quantizing but it's also adding in noise

and the noise that we're adding in is

called signal to quantization noise the

equation you see on your screen is going

to provide us the signal to quantization

noise ratio or dynamic range in decibels

so if we use the latter half of the

equation we get an approximation of this

value we're gonna multiply 6.02 by q Q

being your recording bit depth and when

we throw 16 bit in here we get a dynamic

range or s Q and our of 96 decibels and

for 24-bit audio we're getting a dynamic

range or signal to quantization noise

ratio of 144 decibels and just a quick

side note here these are just

theoretical limits of the dynamic range

chances are in the real world you're not

going to be hitting the

so there you go that's a basic overview

of bit-depth but to sum up what is the

benefit of recording in 24-bit over 16

bit first off you're getting that

improved signal-to-noise ratio as well

as dynamic range and that directly

translates into the ability to record

quieter sound sources without those

sounds getting lost in the noise floor

and secondly you're just getting a more

accurate representation of the analog

sound source that you're recording but I

guess I got to throw in a downside here

and I guess in this case it would just

be that it's taking up more space on

your computer so now do you need to

record in 24-bit well if you have a

24-bit audio interface I see absolutely

no reason why you shouldn't now if

you're uploading to YouTube or another

website that compresses the heck out of

your audio you may not need to record in

24-bit and you may not be getting a huge

benefit but I've always thought that if

you have a higher quality product before

it gets compressed ultimately it's gonna

yield better results keep in mind that's

based on zero research so don't quote me

on that but I do gotta play the devil's

advocate for recording and 16-bit and

that would be that I heard fluff from

riffs beards in gear actually mentioned

that he as well as Joey Sturgis still

record in 16-bit 4401 so honestly just

do what you want just make sure that the

content you're making and the music

you're making is great and at least now

you know the difference between 16-bit

and 24-bit alright guys that's gonna do

it for today I hope you enjoy the video

and I hope you learned something I know

I learned a lot when I was researching

this stuff so if you want more videos

like this go ahead and click the logo

beneath me check out the discord server

link in the description and I will see

you all later bye

[Music]