Bit and Sample rate question...

Flight 16

New member
I have thought about it and think it is cool for me to "Record" in 24 bit 441Khz

What I wanted to know is with the bit rate, As I understand it 24 bit is effective when recording as it provides more sound bits to be picked up so to speak, what I was thinking is, If it is recorded in 24bit and then converted to 16bit, the sound should be no different? as the sound has already been picked up, or will changing it to 16 bit take some of the Audio away?

I understand with the Sample rate it is more of a player configuration...so you can record at a low sample rate but convert it to a high sample rate for mixing?....also will this make the timing of the tracks out of sink in Cubase?

am I talking rubbish or am I on the right track :)?
 
If you record 24 bits and you convert them to 16 bits you will be loosing data.
The bits have to do with volume, each bit is about 6db. The difference between 16 bit and 24 bit is how quiet a signal can be recorded. You can think of it as pushing the noise floor down. (all the extra bits are added to the bottom)

The only way changing the sample rate will scew up the timing of your tracks is if you only change some of them and not others.
 
Flight 16-
If you are using Cubase why not go for 24bit 48khz?
Cubase offers a plugin for Apogee [that works some kind of magic] that I put on my master bus. It takes the extra bits and adds them somewhere to 16bit. Maybe someone else can explain it better, but Apogee rocks.
And make sure you select 32bit floating for all projects in Cubase, unless you know all tracks are 16bit.
J
 
I see no real point in doing up-sampling after the fact. Meaning ... if you want to record at 44.1, 48, 96 or whatever kHz .... just do it and leave it at that till you master for CD. It won't help any to increase it afterward, because what you captured is what you captured .... nothing more will be added by up-sampling a recorded track. The only reason I can see sample rate changes is if you are trying to add other tracks that are of a different sample rate than what your project is at.
It can be beneficial, however, to convert the bit depth (word length) to a higher number for DSP to maintain accuracy in the processing (ie 16 bit to 24 or 32 bit float). 32 bit float would be the logical choice if you use very many plugins (DSP) in that most plugins and multitracking software, for that matter, do their internal DSP in 32 bit float. This saves from needing to use dither till the end (Mastering for CD).
Recording with 24 bit converters gives you a greater dynamic range than 16 bit ones do.
Here is a little info on bit depth and sample rate.
Some good info on dithering .

HTH

-Ken
 
Flight 16 said:
What I wanted to know is with the bit rate, As I understand it 24 bit...
It's not "Bit Rate", it's "word length", "word size", or "bit depth"........

"Bit rate" has NOTHING to do with word size of digital recording but instead, applies to compressed media formats such as MP3s and MPEG and refers to the number of bits transferred per second.
 
Farview said:
If you record 24 bits and you convert them to 16 bits you will be loosing data.
The bits have to do with volume, each bit is about 6db. The difference between 16 bit and 24 bit is how quiet a signal can be recorded. You can think of it as pushing the noise floor down. (all the extra bits are added to the bottom)

The only way changing the sample rate will scew up the timing of your tracks is if you only change some of them and not others.


ahh , i see, cheers, so if i wanted to do some rocking guitars, and drums, 16 bit would be fine? and then do some softer vocals on 24bit?, can you do a project with some 16bit tracks and some 24 bit tracks??
 
Bass Jas said:
Flight 16-
If you are using Cubase why not go for 24bit 48khz?
Cubase offers a plugin for Apogee [that works some kind of magic] that I put on my master bus. It takes the extra bits and adds them somewhere to 16bit. Maybe someone else can explain it better, but Apogee rocks.
And make sure you select 32bit floating for all projects in Cubase, unless you know all tracks are 16bit.
J


Hi, I am useing Cubase, but I am recording most of the Audio into a Digital Tascam recorder and then installing the audio files into Cubase,

How high sample rate does Cubase allow you to record?
 
crankz1 said:
I see no real point in doing up-sampling after the fact. Meaning ... if you want to record at 44.1, 48, 96 or whatever kHz .... just do it and leave it at that till you master for CD. It won't help any to increase it afterward, because what you captured is what you captured .... nothing more will be added by up-sampling a recorded track. The only reason I can see sample rate changes is if you are trying to add other tracks that are of a different sample rate than what your project is at.
It can be beneficial, however, to convert the bit depth (word length) to a higher number for DSP to maintain accuracy in the processing (ie 16 bit to 24 or 32 bit float). 32 bit float would be the logical choice if you use very many plugins (DSP) in that most plugins and multitracking software, for that matter, do their internal DSP in 32 bit float. This saves from needing to use dither till the end (Mastering for CD).
Recording with 24 bit converters gives you a greater dynamic range than 16 bit ones do.
Here is a little info on bit depth and sample rate.
Some good info on dithering .

HTH

-Ken

Hi, The Tascam recorder is 24 bit, 441 samp, like "Bass Jas" said, i think in Cubase i will recrd 24bit 448 samp.

i was just wondering if changing what i have recorded on 24 bit to 16 bit would i still have a better quality then recording in 16 bit, then once i change the 16bit to 24 bit it would be less work on my PC when mixing and so on...

do you think there are records being put out on 16 bit 441 samp rate?

I think i understand now :)
 
Blue Bear Sound said:
It's not "Bit Rate", it's "word length", "word size", or "bit depth"........

"Bit rate" has NOTHING to do with word size of digital recording but instead, applies to compressed media formats such as MP3s and MPEG and refers to the number of bits transferred per second.


I will get there :)
 
STOP!
The internal processing in Cubase is always 32 bit float no matter what you record it at.

There is no reason to up-sample after you record, It won't give you any more information than you recorded in the first place. If you record at 44.1k 16 bit and change it to a 88.2 24bit, you will end up with a much larger file that has the same information in it. You can never have more than you recorded.
If your recordings will end up on CD just record at 44.1k. If you can record at 24 bit and don't mind the file size do it. If you record at 16 bit, you will be fine as well.

There is no compelling reason to go with 48k as the downsampling to 44.1 might do more harm (debatable) than just recording at 44.1k.

You are worrying about something that really doesn't matter as much as you think.
 
Farview said:
There is no compelling reason to go with 48k as the downsampling to 44.1 might do more harm (debatable) than just recording at 44.1k.

Explain (debatable) .
 
Back when the earth was still forming, ( the early 90's) converting from 48k to 44.1k was one of the more time consuming ways to make everything sound like crap. So it was a big no-no. Everyone just recorded at 44.1k if it was going to CD and 48k if it was for video. Now there are better sample rate conversion programs that don't sound as bad as they used to. Depending on which one you use, you might be all right.

The problem we all run into is that there are a lot of misconceptions about all of this, and rules of thumb that no longer apply because technology has moved on. Not to mention the companies trying to sell us more gear that's better, stronger, faster.

One of the points I was trying to get across was that all these things that we are wringing our hands about only make the smallest difference. Some of the first digital recordings were 12 bit 50k and the albums sound great. I don't see the point of getting your underwear in a bunch about the difference between 44.1k and 48k when you are just plugging in a Rode mic through a behringer pre.
 
It can be debatable as to whether or not you will hear the effect of down-sampling. One generally does not want to use sample rates below 44.1k, because of the "Nyquist Frequency" a formula that indicates that the audio bandwidth of a sampled signal is restricted to half the sampling frequency. So in order to cover the approximately 20khz range of human hearing, the equipment must sample at more than 40kHz. So if you are able to hear a 24kHz signal, then you would be able to hear the loss of the highest frequencies due to down-sampling from 48kHz to 44.1kHz.

-Edit to add- And yes, earlier conversion algorithms where poor at best.
 
I C, So recording at 24bit 441 is perfectly ok for producing an album quality recording?

I was only going to up the samples to 448 when putting them in Cubase for when I do vocals....as all the tracks would have to be the same rate. and I thought 448 would be better for vocs

but from the sounds of things I will be fine just sticking with 441, I have also read that 24bit 441, is still better than analogue quality? ..so for a live rock n roll sounding band 24bit 441 is totally cool?

thanks for all your comments, you have all helped.

buy the way anyone hear about the rumoured release of SX3 in mid sep ...mmmm interesting :)
 
Flight 16 said:
I C, So recording at 24bit 441 is perfectly ok for producing an album quality recording?

There were a lot of well produced, big time albums recoded at 44.1 16 bit, you will be fine.

Flight 16 said:
I was only going to up the samples to 448 when putting them in Cubase for when I do vocals....as all the tracks would have to be the same rate. and I thought 448 would be better for vocs

First of all, it's 48k not 44.8k and no it really doesn't make a difference.

Flight 16 said:
but from the sounds of things I will be fine just sticking with 441, I have also read that 24bit 441, is still better than analogue quality? ..so for a live rock n roll sounding band 24bit 441 is totally cool?

Define 'better'. Even 16 bit has a better signal to noise ratio than most analog tape, but that is not why people choose tape.
 
Back
Top