16 bit / 24 bit whats the difference?

T

TunaTheFish

New member
What is the difference between a 16 bit sound card and a 24 bit, i mean would there be a noticable sound difference when recording, playning music e.t.c For someone like me who only records his own stuff, nothing big would it be worth getting a 24 bit sound card?
 
8 bits.


Sorry couldn't resist. :D

24 bit is certainly easier to work with in terms of headroom and noise floor but it's far from being the most important part of the chain. I suspect whether you'll notice much of a difference or not will depend on the rest of your signal chain and the overall quality of your current recordings.

And it will all end up back at 16 bit anyway if you burn to CD.
 
There are those who will notice the difference. Most don't. I wouldn't worry too much about that, and concentrate on having good pre amps, cables, mics, and most importantly, a good solid performance that you can record.
 
The difference is in the resolution. With very bit of information, the resolution doubles. Imagine the difference between 16 bit and 24 bit to be the difference in a painting. if a painting looks great with one brush stroke across the canvas, imafine what it would lok like with the detail of two. Or four. Or eight. the difference between 16 and 24 bit is actually 256 times the resolution. The picture looks the same, but with more detail and depth. Same as audio.

Sorry, that was my "picturesque" rant for the day
:D
 
Yeah like what battleminnow said.

With digital audio, each sample is actually a number that represents the amplitude of the signal. When people are talking about bit depth, they are talking about the number of bits used to represent the amplitude.

With 16 bit audio, there is 2^16 possible values for the number. This works out to be 65536 possible numbers (0 - 65535). Now with 24 bit audio there are 2^24 numbers, which is a whopping 16777216 possibilities to store the amplitude. That simply means there are more numbers to store the amplitude of the signal. That means if your converters are accurate, your digital audio will be much more accurate. Then when you put it onto CD it gets converted back to 16 bit anyway.

It's the same with digital representation of colour. Open up the display dialog in Control Panel, click the settings tab and change the colour quality. Less bit depth = less possible colours. If it will let you go down to 8 bit, thats a tiny 256 colours, and everything will look ugly.
 
MichaelM said:
Now with 24 bit audio there are 2^24 numbers, which is a whopping 16777216 possibilities to store the amplitude. That simply means there are more numbers to store the amplitude of the signal. That means if your converters are accurate, your digital audio will be much more accurate.
Click to expand...

The way I understand it, is the additional numbers added with 24 bit resolution are added to the bottom of the db scale. So instead of getting better resolution at high levels (above 96db for 16 bit, or 144db I believe in the case of 24 bit) you actually get added clarity and resolution towards the lower end of the decibal scale. I think this is more recognizable in classical music, during soft passages of a song.
 
I hate to bust the bubble, but we've been over and over this in the mic forum and it makes absolutely NO difference. I've argued myself the point that 24-bit 96khz sampling was better because of higher resolution and I was proved WRONG.

This is marketing hype by the people who make the hardware to get you to buy something new. They take advantage of public ignorance of scientific principles and use it against you so that you'll buy the "latest and greatest" product, abandoning something that was just fine in the first place. (maybe even superior to the new product.)

Read this whitepaper by Dan Lavry. It's a PDF file.

http://www.lavryengineering.com/documents/Sampling_Theory.pdf

There's some heavy math, but it proves that 16-bit 44.1khz contains all the information neccessary in audio sampling. It's based on Dr. Nyquist's analasis of audio waves. 96khz may actually make it worse.

The most important thing in capturing audio is the quality of the A/D converters.
 
PhilGood said:
I hate to bust the bubble, but we've been over and over this in the mic forum and it makes absolutely NO difference. I've argued myself the point that 24-bit 96khz sampling was better because of higher resolution and I was proved WRONG.

This is marketing hype by the people who make the hardware to get you to buy something new. They take advantage of public ignorance of scientific principles and use it against you so that you'll buy the "latest and greatest" product, abandoning something that was just fine in the first place. (maybe even superior to the new product.)

Read this whitepaper by Dan Lavry. It's a PDF file.

http://www.lavryengineering.com/documents/Sampling_Theory.pdf

There's some heavy math, but it proves that 16-bit 44.1khz contains all the information neccessary in audio sampling. It's based on Dr. Nyquist's analasis of audio waves. 96khz may actually make it worse.

The most important thing in capturing audio is the quality of the A/D converters.
Click to expand...

Isn't that paper about sampling rate?

We're talking about bit depth.

Where does Mr Lavry state in that paper that 24 bit is no different to 16 bit?
 
Sorry, I was making an assumption. Most people who jump to 24 bit are uaually trying to get to 96khz.

My mistake.

Explain to me, though, how a 24 bit sample at 44.1khz is better than a 16-bit sample at 44.1khz?
 
Hmm that makes no sense to a noob like me :) But i have got the message that i wouldn't notice the difference and it isnt worth mr upgrading.
 
Suppose there's a pool, and a guy in it. You can say "there's a guy in the pool".

Now suppose you paint a line on the middle of the pool. It's the same guy in the same pool, but now you can say "there's a guy in the pool, and he is on the left side".

If you paint a line in the middle of the two halves, now you can say: "there's a guy in the pool, and he's on the left side. Although he's closer to the middle than to the border".

The more bits of info you use, the more detail you have, although it's the same guy on the same pool.

Don't know, that image works for me.
 
Thats a good link, also read the samples article. Dan Lavry's paper was talking about sample rate, not bit depth. Yes there's a lot of math, but from my understanding the 96Khz making things worse has more to do with the practicalities of designing circuits to handle high sample rates. I thought the theory proved that 44.1Khz was adequate for recording all the sounds that a human can hear, but doesn't state that higher sample rates are worse. That part comes from actual circuit design. Well, thats my measly undertanding of the thing ;)

rythmrmixed said:
The way I understand it, is the additional numbers added with 24 bit resolution are added to the bottom of the db scale. So instead of getting better resolution at high levels (above 96db for 16 bit, or 144db I believe in the case of 24 bit) you actually get added clarity and resolution towards the lower end of the decibal scale.
Click to expand...
That may be true, I wouldn't have a clue. The point is more possible values.

TunaTheFish said:
But i have got the message that i wouldn't notice the difference and it isnt worth mr upgrading.
Click to expand...
Ha yeah your right I think we blew this question out of proportion.
 
No, I get the bit rate part. It has more to do with dynamics accuracy. I just assumed to quickly it was about sampling rate.
 
PhilGood said:
No, I get the bit rate part. It has more to do with dynamics accuracy. I just assumed to quickly it was about sampling rate.
Click to expand...
Sorry to pick up on another of your posts but I don't believe bit 'rate' is the correct term, it's bit 'depth'.

We're not talking about the number of bits transferred per second (bit rate), bit depth (as I understand it) refers to the amount of bits available to 'describe' our audio signal.

As I said before I find that recording at 24 bit I get good levels without having to record too hot so the noise floor is lowered and the risk of clipping is reduced. It's by no means a significant difference but I need all the help I can get. :D
 
Last edited:
Digital audio is basically an attempt to accurately reproduce a true analog waveform. A given bit depth can be thought of as how close that attempt is to being identical to that waveform. 24 bits is much closer than 16 bits is, because the represention (resolution) of the waveform is more accurate. If you place a digital audio waveform next to an analog waveform, you would notice the digital waveform has small steps in it, where an analog waveform has no steps and is a smooth curve as it increases and and decreases in amplitude.
 
Sorry! Bit depth. Sheesh!

I won't type so fast next time. Did I spell everything right? Is my puctuation OK? Should I indent? I didn't end a sentence with a preposition, did I?
 
PhilGood said:
Sorry! Bit depth. Sheesh!

I won't type so fast next time. Did I spell everything right? Is my puctuation OK? Should I indent? I didn't end a sentence with a preposition, did I?
Click to expand...
Sorry to put your back up. If it's me (and it often is) I like to be corrected, that's how i learn.
 
PhilGood said:
Sorry! Bit depth. Sheesh!

I won't type so fast next time. Did I spell everything right? Is my puctuation OK? Should I indent? I didn't end a sentence with a preposition, did I?
Click to expand...

Actually, I think puctuation is spelled punctuation.

Sorry, I couldn't resist. :D
 
You must log in or register to reply here.
Back
Top