Bit rate.

[MichaelAlan]

can someone explain bit depth theory to me in a way I can understand. I am not a scientist and everyone else seems to be. I understand sampling rates...but not bit depth. How much of a difference will I hear recording in 24 bits over 16?



mike

[trevord]

in 2 words "dynamic range" or headroom

to oversimplify in recording terms
if you are recording someone who screams then whispers
at 16 bits -
if you adjust the pre-amp so the screams dont distort, the whisper will be too low (and may not even register)
if you turn up the pre-amp to get the whispers, the screams will distort.

at 24 bits -
you will have enough room to turn up the pre-amp to get the whispers and the screams will NOT be distorted

in real terms the dynamic range provided by a "pro" 16 bit machine is equal to a whisper compared to a jet enigine at a reasonable distance
24 bit range is a whisper to something that will blow your head off

the main problems with 16 bits come from really really soft things like reverb tails
you would notice it by getting more "room" by turning up your pre-amp without distortion (from your a/d anyway)



that is it - in recording terms
the net has many scientific explanations

[Zeppelin4Life]

bits can be very simple. they are just a measure of dynamics on recrord. In theory the 'silence' on 144dB is softer.
simply take the bitrate and multiply by 6 and thats maximum dynamic range in dB

16bits(6)=96dB
24bits(6)=144dB



sure, you can get in depth.but thats about all you need to know

[bap]

More bit depth basically gives you more dynamic range. It also allows you to record at somewhat lower levels because the 'noise floor' will be farther below the recorded sound.

Go to project studio handbook http://www.theprojectstudiohandbook.com/articles4.htm and look under digital audio.

I am also not a scientist but still always record at 24 bits.

[dynomike]

Man, I just typed the longest response and the net crashed! Grrr

Okay well in terms of recording:

24 bits will get you 256x the dynamic range of 16 bit. Thats a big deal. You can set the levels so you have ample headroom, but still don't have to worry about low-level noise (or distortion). The lower the level gets in digital, the more "quantization errors" occur - at very low levels, the signal cannot be accurately represented by the available number of bits.

For low level stuff, like trevor mentioned, reverb tails, it is really important to have these extra bits at your disposal.

In mixing:

More bits=less truncation errors. When you pull a fader up or down or basically change anything in any way (plugins.. faders.. pan) a complex calculation is being done. The resultant number (for each sample) is going to be longer than the original 16 or 24 bits.. so the end numbers get truncated. This results in small errors which over the course of a few plugins, several faders, etc can be significant. Having a higher internal bit depth (PTHD has 48 bits, some digital mixers have 56 bit) means that there is better accuracy in the calculations, and you won't hear those nasty little digital artifacts that come from truncation and quantization errors.

Hope that makes sense. You asked how much you can hear the difference... that is totally up to you. Try it out. Of course, all these technical reasons kinda make you feel guilty if you don't hear it...

Mike

[Zeppelin4Life]

I should add, if your burning a 24-bit to a CD (16) you will add dither, and the effect of 24-bits will be brought to a minimum. so, all CDs (right now) are 16-bit. something to think about

[Zeppelin4Life]

24 bits will get you 256x the dynamic range of 16 bit. Thats a big deal.
Mike

. 16 is 256 X bigger than 8. I think this is true too.

[bap]

Recorded bit depth, of course, is not entirely responsible for good or poor digital sound. Truncation/dithering/wordlength problems are by and large the results of poor quality hardware and software. If your gear can't do the math then different sample rate/bit depth will not perform many wonders when you decide to twist a lot of knobs!

[psychicoctopus]

bits can be very simple. they are just a measure of dynamics on recrord. In theory the 'silence' on 144dB is softer.
simply take the bitrate and multiply by 6 and thats maximum dynamic range in dB

16bits(6)=96dB
24bits(6)=144dB



sure, you can get in depth.but thats about all you need to know

144 dB dynamic range would be awesome, too bad no analog circuitry exists that can reproduce it!

[Disasteradio]

the difference between 24 and 16 bit is crazy huge. case in point: increasing any digital numbering system by one bit doubles its precision. ie:

3-bit binary:
000
001
010
011
100
101
110
111
(8 degrees of precision, or the maximum decimal number represented is 8 )

0000
0001
0010
0011
0100
0101
0110
0111
1000
1001
1010
1011
1100
1101
1110
1111

(16 degrees from four bits.. so you can imagine the size of the numbers we're dealing with with 24 bit recordings!)
(and ain't it fun to note that as you go from right to left, the alternations between sets of zeros and one double in length!)
Getting your head around binary sometime would be totally helpful. it's a cinch. http://en.wikipedia.org/wiki/Binary_numeral_system

[bap]

144 dB dynamic range would be awesome, too bad no analog circuitry exists that can reproduce it!

Remind me to get the earplugs if I'm ever subjected to that kind of loudness!

[Zeppelin4Life]

yeah..so 16 bit is

0000000000000000
0000000000000001
0000000000000011
0000000000000111

oh man..

24 is
000000000000000000000001
000000000000000000000011
000000000000000000000111

damn..

so do the math..

lets see.

16 bit would be..
2^16...= 65,536 possible values (this is range)
2^24= 16,777,216

yes, a huge increase...you are right..256X

PS..144dB is insanely loud..with that said I think the loudest launch ever was like 185dB from Cape Canaveral..that might be a record..im not sure...ripped the roof off a field warehouse some 3 miles away

[Phiz]

144dB is not a volume level. In this case it defines a range of intensities. 144 dB SPL is a volume level, because it is referenced against some base pressure intensity -- the threshold of human hearing.

[dynomike]

I should add, if your burning a 24-bit to a CD (16) you will add dither, and the effect of 24-bits will be brought to a minimum. so, all CDs (right now) are 16-bit. something to think about

Dave,

Dither is a good thing. It actually increases the effective dynamic range of 16 bit to beyond 96 db, and ADDS detail to low level signals by modulating noise at the lowest bit, which reduces quantization distortion..

Ok I'm getting too scientific. You should look into this for yourself though. Dither is necessary, and improves drastically the quality of 16 bit, and is especially important when coming from 24 bit. You can't just truncate that.. you'll get nasty quantization errors. Good dither makes a big difference. Check out www.digido.com or get Ken Pohlmann's "Principles of Digital Audio".

Mike

[Zeppelin4Life]

I should add, if your burning a 24-bit to a CD (16) you will add dither, and the effect of 24-bits will be brought to a minimum. so, all CDs (right now) are 16-bit. something to think about

Dave,

Dither is a good thing. It actually increases the effective dynamic range of 16 bit to beyond 96 db, and ADDS detail to low level signals by modulating noise at the lowest bit, which reduces quantization distortion..

I know. correct me if im wrong though, but dither is only needed while going from 24-bit to 16-bit (or any other times reducing bits). dither is NOISE added to the 24-bit signal to fill in thoes extra 48dB with noise. you can hardly hear it, but its there to make thoes bits that are 'below' the dynamic range of 16-bit. the noise makes them better 'heard' to the 16-bit decoders. I don't think it 'increases' the dyanmic range of 16-bit per say, it more or less makes 16-bit 'more like 24-bit'..wew. So, if you are recording and mixing in 16-bit, dither is not needed. Without dither from a 24-bit source to a 16(like CD) will result in some quantization distortion, yeah, which is a light distortion during quiet times in the music. simply saving a 24-bit file to a 16-bit CD would be 'truncating'...literally chopping off bits.

I can bet tho, that most people wouldnt even notice the difference between all choices

its cool to mention a lot of mastering uses 32-bit and even high floating point