Hello guys.
The bit depth and bit rate is a way to say in words, how we transform the analog signal, to digital.
Long story short, Sample rate is the number of audio samples a computer records a given signal (vocals or guitar etc.) per second. It’s time relative. For example 44.1kHz is equal to 44,100 samples of audio recorded every second.
The human ear is to perceive a maximum range between 20Hz to 22.5kHz. That is the range of sound we can hear. Anything lower you feel and anything higher you just won’t hear. In the digital domain we double that frequency and we have an excellent representation of our analogue sound which is 44.100Hz / 44.1KHz.
44.1kHz - Sample rate for CD
48kHz - Sample Rate for Video
88.2kHz - Twice the sample rate for CD
96kHz - Twice the sample rate for video
176.4kHz - High quality HD audio
192kHz - Highest quality HD audio
The bigger the sample rate the bigger the data.
There is no need for audio application to use 48Khz.
The bit rate, 16 or 24, the how precise the sample can it be. Hense, how many bits / sample we have, and it directly corresponds to the resolution of each sample. 16 is the normal and HiFi signal used for commercial products like CDs. 24 is for DVD's and Bluray applications, although in studio recordings 44.1/24 is commonly use in order to have more high resolution recordings of your audio.
That's it for me in order to help you guys in your questions.
ILIAS GOGAKIS
Professional Mixing and Mastering Engineer
BA (Hons) Audio and Music Technology
Originally posted by iliasgogakis on Sat 02 Jan, 2016
Nicely worded, Ilias. A 32 bit answer, I think ;)
#1057 Posted Sat 02 Jan, 2016 2:20 am
Nicely worded, Ilias. A 32 bit answer, I think ;)
Originally posted by MonkeyC on Sat 02 Jan, 2016
Haha, thanks Monkey...i just put my technical side there. Hope I helped a bit.
Wish you all a happy 2016
Wish you all a happy 2016
#1058 Posted Sat 02 Jan, 2016 5:04 am
In my days of audio science in electronics, you couldn't recreate a sine wave of 20khz from an analogue source. At 44.1khz you only got 2 sample points at best and maybe a few more at the higher sample rates. Anything converted using compression (i.e. mpa, mp3) is lossy like a jpeg image. Energy is removed from the high end content like a spectral pump of frequency bands. It fools the ears into sounding good until you lower the bitrate, then it starts sounding like your listening inside a tin can. What's important is the bit depth rate of 16,24,32 bits per sample to reduce aliasing noise when applying effects and gain in a daw. Also if you use effects for vocal or instrument manipulation like melodyne, elastic audio for time and pitch correction, its always best to apply them against dry tracks at the start of an effect chain. before adding reverb or gain effects. Wet tracks create harmonics and overtones that may produce inadvertent sounds like warble and tin can sounds... Cheers!
#1059 Posted Sat 02 Jan, 2016 9:34 am
In my days of audio science in electronics, you couldn't recreate a sine wave of 20khz from an analogue source. At 44.1khz you only got 2 sample points at best and maybe a few more at the higher sample rates. Anything converted using compression (i.e. mpa, mp3) is lossy like a jpeg image. Energy is removed from the high end content like a spectral pump of frequency bands. It fools the ears into sounding good until you lower the bitrate, then it starts sounding like your listening inside a tin can. What's important is the bit depth rate of 16,24,32 bits per sample to reduce aliasing noise when applying effects and gain in a daw. Also if you use effects for vocal or instrument manipulation like melodyne, elastic audio for time and pitch correction, its always best to apply them against dry tracks at the start of an effect chain. before adding reverb or gain effects. Wet tracks create harmonics and overtones that may produce inadvertent sounds like warble and tin can sounds... Cheers!
Originally posted by Sterling on Sat 02 Jan, 2016
Aliasing has to do with frequency, bit depth has to do with dynamic range. They are not related.
#1060 Posted Sat 02 Jan, 2016 10:19 am
Yes they are! raising the dynamics of a low level signal produces distortion from Nyquist errors between sample steps and gain (round off errors). I think I'll keep quiet next time...
(Quote from whatis.com)
The Nyquist Theorem, also known as the sampling theorem, is a principle that engineers follow in the digitization of analog signals. For analog-to-digital conversion (ADC) to result in a faithful reproduction of the signal, slices, called samples, of the analog waveform must be taken frequently. The number of samples per second is called the sampling rate or sampling frequency.
Any analog signal consists of components at various frequencies. The simplest case is the sine wave, in which all the signal energy is concentrated at one frequency. In practice, analog signals usually have complex waveforms, with components at many frequencies. The highest frequency component in an analog signal determines the bandwidth of that signal. The higher the frequency, the greater the bandwidth, if all other factors are held constant.
Suppose the highest frequency component, in hertz, for a given analog signal is fmax. According to the Nyquist Theorem, the sampling rate must be at least 2fmax, or twice the highest analog frequency component. The sampling in an analog-to-digital converter is actuated by a pulse generator (clock). If the sampling rate is less than 2fmax, some of the highest frequency components in the analog input signal will not be correctly represented in the digitized output. When such a digital signal is converted back to analog form by a digital-to-analog converter, false frequency components appear that were not in the original analog signal. This undesirable condition is a form of distortion called aliasing.
(Quote from whatis.com)
The Nyquist Theorem, also known as the sampling theorem, is a principle that engineers follow in the digitization of analog signals. For analog-to-digital conversion (ADC) to result in a faithful reproduction of the signal, slices, called samples, of the analog waveform must be taken frequently. The number of samples per second is called the sampling rate or sampling frequency.
Any analog signal consists of components at various frequencies. The simplest case is the sine wave, in which all the signal energy is concentrated at one frequency. In practice, analog signals usually have complex waveforms, with components at many frequencies. The highest frequency component in an analog signal determines the bandwidth of that signal. The higher the frequency, the greater the bandwidth, if all other factors are held constant.
Suppose the highest frequency component, in hertz, for a given analog signal is fmax. According to the Nyquist Theorem, the sampling rate must be at least 2fmax, or twice the highest analog frequency component. The sampling in an analog-to-digital converter is actuated by a pulse generator (clock). If the sampling rate is less than 2fmax, some of the highest frequency components in the analog input signal will not be correctly represented in the digitized output. When such a digital signal is converted back to analog form by a digital-to-analog converter, false frequency components appear that were not in the original analog signal. This undesirable condition is a form of distortion called aliasing.
#1062 Posted Sat 02 Jan, 2016 11:36 am
There is no mention of bit depth being a factor in that quoted text.
#1063 Posted Sat 02 Jan, 2016 12:38 pm
Round off errors in levels of gain is called "quantization noise". It has nothing to do with aliasing. You address aliasing with a higher sampling frequency and/or an anti aliasing filter. You address quantization noise with dithering and/or a higher bit depth.
#1064 Posted Sat 02 Jan, 2016 12:57 pm
Bit depth is the digital thing. Is like how many photographs in a period of the signal will be taken (of the analogue signal) and will be translated in the digital domain. Normal 16bit is ok for reproduction. For recording 24 is very standard for HD , where is Superd HD they use 32 float bit rate. Have in mind that the higher the bit rate and the sample will give bigger files
#1065 Posted Sat 02 Jan, 2016 2:33 pm
This is an awesome conversation going on here. For my two cents. I always keep the digital format that is used to present audio to the consumer. Which is 44.1 at 16 bits for CD. Now, we also have to deal with streaming services, and who knows what they do to the audio to be the best format for them, not the consumer. That's what I mix at unless asked to do something different. My main reason is that the dithering convertors can vastly alter the final product. Sometimes overall and sometimes it's frequency specific. To date I have not found any software or program that does this with any sense. Some will sound better and others will not. Most of the algorithms either drop the least significant bit or the most or both. In other word you gamble every time you go through a conversion.
That's why I hope for the CD standard formulas for every step of the production. Can we hear it or feel it, that is the question.
I use Apogee outboard gear for my sample and clock rates. I line everything up to one master word-clock. This keeps whatever bit depth or sample rate following the main clock and removes a lot of shuttering and distortion and jitter. By doing this your mix will get to where you want to start much quicker because the audio racks aren't shifting back and forth through out the entire song.
It also gives whatever you're using in the digital realm the chance it needs to playback consistently.
Well, I think Ive muddied water just a bit (pun intended) more. Always follow your ears. They didn't cost you anything and came in your shipping box with you! : - )
TC
That's why I hope for the CD standard formulas for every step of the production. Can we hear it or feel it, that is the question.
I use Apogee outboard gear for my sample and clock rates. I line everything up to one master word-clock. This keeps whatever bit depth or sample rate following the main clock and removes a lot of shuttering and distortion and jitter. By doing this your mix will get to where you want to start much quicker because the audio racks aren't shifting back and forth through out the entire song.
It also gives whatever you're using in the digital realm the chance it needs to playback consistently.
Well, I think Ive muddied water just a bit (pun intended) more. Always follow your ears. They didn't cost you anything and came in your shipping box with you! : - )
TC
#1083 Posted Wed 06 Jan, 2016 8:41 pm
This is an awesome conversation going on here. For my two cents. I always keep the digital format that is used to present audio to the consumer. Which is 44.1 at 16 bits for CD. Now, we also have to deal with streaming services, and who knows what they do to the audio to be the best format for them, not the consumer. That's what I mix at unless asked to do something different. My main reason is that the dithering convertors can vastly alter the final product. Sometimes overall and sometimes it's frequency specific. To date I have not found any software or program that does this with any sense. Some will sound better and others will not. Most of the algorithms either drop the least significant bit or the most or both. In other word you gamble every time you go through a conversion.
That's why I hope for the CD standard formulas for every step of the production. Can we hear it or feel it, that is the question.
I use Apogee outboard gear for my sample and clock rates. I line everything up to one master word-clock. This keeps whatever bit depth or sample rate following the main clock and removes a lot of shuttering and distortion and jitter. By doing this your mix will get to where you want to start much quicker because the audio racks aren't shifting back and forth through out the entire song.
It also gives whatever you're using in the digital realm the chance it needs to playback consistently.
Well, I think Ive muddied water just a bit (pun intended) more. Always follow your ears. They didn't cost you anything and came in your shipping box with you! : - )
TC
Originally posted by tomic on Wed 06 Jan, 2016
I agree with what you said on the "That's why I hope for the CD standard formulas for every step of the production. Can we hear it or feel it, that is the question." The point here is that on the music production there is no point to go above 44KHz. The only reasonable explanation is to choose 24 samples in order to have better conversion on your analog to digital music.
And again that depends on the hardware you have, and the converters, and then a new story begins.
If you are a pro, then you can have a state of the art converters and monitors and then the translation can be successful.
Even if you are working in a home studio you can have a very good system with decent converters but you still need good monitors to make the translation.
The most important though is the room acoustics. All the above is useless, if your room sounds like a cathedral room :/
You need some Rockwool panels in the room in order to minimize the reverberation and the reflections in the room.
Once we resolve the basic physics, then we can continue with sample rates and bit depths :)
Have a good weekend :)
ILIAS
And again that depends on the hardware you have, and the converters, and then a new story begins.
If you are a pro, then you can have a state of the art converters and monitors and then the translation can be successful.
Even if you are working in a home studio you can have a very good system with decent converters but you still need good monitors to make the translation.
The most important though is the room acoustics. All the above is useless, if your room sounds like a cathedral room :/
You need some Rockwool panels in the room in order to minimize the reverberation and the reflections in the room.
Once we resolve the basic physics, then we can continue with sample rates and bit depths :)
Have a good weekend :)
ILIAS
#1125 Posted Fri 15 Jan, 2016 8:54 am
