16 BIT VERSUS 24 BIT

[mrinshredible]

so ultimately when we hear down graded music in itunes format or mp3 formats, its usually 16 bit. at 44.1khz

some people mixdown and dither to 16bit.

i do this.

so my question for the pros is... why do i need to be recording in 24 bit (or higher) if everything i do ends up being 16 bit when i mixdown.

i get bit depth and sample rates. but why rely on software even further when recording/converting when 16 bit is the "playback standard." meaning why record in 24 bit, only leaving the audio to be processed YET again... at mixdown AND, to be processed AGAIN when someone imports it into their itunes, w/e they have.

this makes 3 steps of digital conversion with any bit depth higher than 16bit, instead of 1 step of conversion at initial recording of 16 bit.

catch my drift? plus wont ALL of my plugs run better if i set up my daw to record 16 bit. i just want to know the real benefits of recording 24 bit besides thearing the "better quality."

 

[Jarkko Mattheiszen]

The advantage of recording in 24 bits is the amount of dynamic range you have versus 16 bits. You have more headroom and a relatively lower noise floor.

 

[mrinshredible]

yeah true.

i guess its more of a preference than anything as theoretically it is superior.

i want to know preferences and why, not the facts. i dont record in 16 bit but im also not convinced i need to record in 24 bit.

how many people on here record in 16 bit. how many 24 bit. and do you dither from 24 to 16?

 

[rfahey86]

The advantage of recording in 24 bits is the amount of dynamic range you have versus 16 bits. You have more headroom and a relatively lower noise floor.

Thats the same reason I record in 24 bit. Also, DVD disks play at 24 bit. I'm sure in the near future 24 bit will be the standard.

 

[Mesa4x12er]

yeah true.

i guess its more of a preference than anything as theoretically it is superior.

i want to know preferences and why, not the facts. i dont record in 16 bit but im also not convinced i need to record in 24 bit.

how many people on here record in 16 bit. how many 24 bit. and do you dither from 24 to 16?

No it's really not a preference thing at all.

 

[mrinshredible]

lol yeah "HIGH DEFINITION" Audio is actually 24 bit.... yet we see higher bit rates. i think when people can afford high def cards, 24 bit will be standard. for the time being software and hardware for computers right now is so minimal for average users that the quality is actually way ahead of the technology as well as the prices and the never growing market with new technologies! besides the new intel chips with 8 core processing, theres really nothing else an enthuist can pruchase besides a "high definition" (again 24 bit) audio card. but there in we are limiting to software and processing power... but maybe they buy a 16 bit cd, theres a waste of money.

i know from my standpoint, i like to record in 24 bit for a higher sampling rate, but there really isnt much audible difference in bit depth unless you have dirty power from your pc or a dirty signal from pre=amps.

 

[mrinshredible]

how isnt it. theres pros who record with 16 bit.

 

[jval]

i think this was debated and decided in the industry a long time ago. i doubt any pros are recording at 16 bit. there's no reason to.
read the bob katz book.

 

[mrinshredible]

no thanks

 

[Pedro Teixeira]

LOL wtf?

 

[egan.]

lol yeah "HIGH DEFINITION" Audio is actually 24 bit.... yet we see higher bit rates.

Where? What A/D and D/A hardware supports PCM bit depths greater than 24? Don't confuse 32bit float in software with true 32bit recording.

I use 24bit b/c it sounds better to me and is worth easily worth the extra HD space. I print mixes 24 bit and let my mastering engineers truncate them for the CD format but ultimately clients receive unmastered 44.1/24bit mixes from me as well.

http://www.tweakheadz.com/16_vs_24_bit_audio.htm
http://www.digido.com/more-bits-please.html

i know from my standpoint, i like to record in 24 bit for a higher sampling rate, but there really isnt much audible difference in bit depth unless you have dirty power from your pc or a dirty signal from pre=amps.

"Bit rate" and "sample rate" are two different things. Of course your chain is only as strong as the weakest link, but that's a side step of the actual issue.

 

[Jarkko Mattheiszen]

It sounds like you don't really have a grasp of the meaning of bit rate nor sample rate. And "no thanks" to reading Bob Katz's Mastering Audio? Really?

 

[Jarkko Mattheiszen]

i just want to know the real benefits of recording 24 bit besides thearing the "better quality."

i want to know preferences and why, not the facts.

lol yeah "HIGH DEFINITION" Audio is actually 24 bit.... yet we see higher bit rates. i think when people can afford high def cards, 24 bit will be standard.

there really isnt much audible difference in bit depth unless you have dirty power from your pc or a dirty signal from pre=amps.

Sorry dude, but after reviewing the thread, you really don't make much sense to me at all.

 

[arv_foh]

why wouldn't you use 24 bit if you have the option to?

 

[abt]

Headroom.

 

[PaoloJM]

 

[TheWinterSnow]

OP is a noob but I have to chime in on the 16 vs 24 debate.

16 bit has a resolution of 65,535 and 24 bit has a resolution of 16,777,215. Now at first that may seem like a lot but we are comparing the logarithmic difference between 65K and 16.7M. So if you use the equation to calculate dynamic range 20Log(x), 16 bit has a DR of 96db and 24 bit has a DR of 144db. The difference in DR of the two resolutions is 48db, just as 32-bit resolution has a dynamic range of 48db more than 24-bit (192db).

To put how much 48db is in perspective to you, the difference in volume from a quiet room to a soft whisper is about 40-50 db. That means that noise that was 40db louder than the noise floor recorded on 16 bits would not be audible on 24 bits, but typically speaking the total noise from preamp and other mechanical noises will never equate to anything audible. If you track with correct levels and keep the noise lower than 12db (generally low enough for the human ear to not be able to detect) then that noise will not be detectable on both 16 and 24 bit sampling. In terms of DR itself in terms of realism, you loose all those dynamics when you convert down to 16, the only thing you ensured was a low noise floor, a noise floor precaution that isn't needed if your total noise is quieter than a whisper.

I have tested this out for myself. I have tweaked around with recording noise, just nothing into a mic, both with 16 bit and 24 bit and while 24 bit may be a little quieter, its not huge. 16 bit is about twice as loud, but if you record a few seonds of silence follwed by a guitar lick doing this for both 16 and 24 bit, matching the level of your guitars so that they are coming out of your monitors at about 83db, you won't hear the noise from the few seconds of silence on any track, both 16 or 24, so really does double the noise really matter when you can hear neither.

In reality yes 24 bit is quieter, but is it actually worth 256 times more space for the audio files for a very small difference in noise, probably not. In the end, it really doesn't matter, if you tracked correctly if the final result is in 16 bits, it doesn't matter if you tracked with 16 or 24, no one, including Sneap himself could tell if a song was recorded in 16 or 24.

 

[Jarkko Mattheiszen]

OP is a noob but I have to chime in on the 16 vs 24 debate.

16 bit has a resolution of 65,535 and 24 bit has a resolution of 16,777,215. Now at first that may seem like a lot but we are comparing the logarithmic difference between 65K and 16.7M. So if you use the equation to calculate dynamic range 20Log(x), 16 bit has a DR of 96db and 24 bit has a DR of 144db. The difference in DR of the two resolutions is 48db, just as 32-bit resolution has a dynamic range of 48db more than 24-bit (192db).

To put how much 48db is in perspective to you, the difference in volume from a quiet room to a soft whisper is about 40-50 db. That means that noise that was 40db louder than the noise floor recorded on 16 bits would not be audible on 24 bits, but typically speaking the total noise from preamp and other mechanical noises will never equate to anything audible. If you track with correct levels and keep the noise lower than 12db (generally low enough for the human ear to not be able to detect) then that noise will not be detectable on both 16 and 24 bit sampling. In terms of DR itself in terms of realism, you loose all those dynamics when you convert down to 16, the only thing you ensured was a low noise floor, a noise floor precaution that isn't needed if your total noise is quieter than a whisper.

I have tested this out for myself. I have tweaked around with recording noise, just nothing into a mic, both with 16 bit and 24 bit and while 24 bit may be a little quieter, its not huge. 16 bit is about twice as loud, but if you record a few seonds of silence follwed by a guitar lick doing this for both 16 and 24 bit, matching the level of your guitars so that they are coming out of your monitors at about 83db, you won't hear the noise from the few seconds of silence on any track, both 16 or 24, so really does double the noise really matter when you can hear neither.

In reality yes 24 bit is quieter, but is it actually worth 256 times more space for the audio files for a very small difference in noise, probably not. In the end, it really doesn't matter, if you tracked correctly if the final result is in 16 bits, it doesn't matter if you tracked with 16 or 24, no one, including Sneap himself could tell if a song was recorded in 16 or 24.

I do understand your points and they are very valid indeed, but I think you're seeing only one aspect of the debate in a way. The difference might not be crucial at all when recording guitar licks, but tracking a classical orchestra, for example, is a whole different story. Not that I'm an expert in classical music recording by any means, but I was taught the fundamentals of digital audio in practice mostly by people who are. And at least to me, the 48dB difference in the available dynamic range is reason enough to track in 24 bits, especially since there really is no downsides to that considering the processing power and price of disk space today.

 

[crazykarl]

I do understand your points and they are very valid indeed, but I think you're seeing only one aspect of the debate in a way. The difference might not be crucial at all when recording guitar licks, but tracking a classical orchestra, for example, is a whole different story. Not that I'm an expert in classical music recording by any means, but I was taught the fundamentals of digital audio in practice mostly by people who are. And at least to me, the 48dB difference in the available dynamic range is reason enough to track in 24 bits, especially since there really is no downsides to that considering the processing power and price of disk space today.

You and WinterSnow pretty much summed it up. Rock isn't going to necessarily need the dynamic range of 24-bit, but it doesn't hurt. For those that want to see what bit depth does, take a classical song and bounce it down to 8 bit without dither, during the quiet parts it might just go completely silent.

Now let's end this before someone starts debating sample rates and denies the Nyquist theory's existance.

 

[Jarkko Mattheiszen]

Now let's end this before someone starts debating sample rates and denies the Nyquist theory's existance.

Quoted for truth.