E
easton15
New member
What is the difference between 24bit/96kHz and 24bit/192kHz?
apl
Stand Up Comity
Hi, Easton.
96kHz and 192kHz are, in this case, sampling frequencies. To create digital audio, an electrical signal, such as from a microphone, is measured at time intervals. The list of numbers created by measuring at intervals becomes a digital representation of what the signal looked like. Information between the samples is lost. This begs the question, "How much loss is acceptable?"
By signal processing theory, you can accurately represent about half the sampling frequency. So to cover the audible frequency range, 44.1kHz is adequate.
Great arguments are made for using higher sampling rates, usually based on subjective opinions. However, physics, math, and physiology do not support the need to sample at higher frequencies. But, hey, if it sounds better to you, go for it!
96kHz and 192kHz are, in this case, sampling frequencies. To create digital audio, an electrical signal, such as from a microphone, is measured at time intervals. The list of numbers created by measuring at intervals becomes a digital representation of what the signal looked like. Information between the samples is lost. This begs the question, "How much loss is acceptable?"
By signal processing theory, you can accurately represent about half the sampling frequency. So to cover the audible frequency range, 44.1kHz is adequate.
Great arguments are made for using higher sampling rates, usually based on subjective opinions. However, physics, math, and physiology do not support the need to sample at higher frequencies. But, hey, if it sounds better to you, go for it!
K
kevinb9n
New member
disclaimer: I'm reading the newbie thread because I AM a newbie. So take what I say with a grain of salt.
Let's say you're recording a single track in mono.
96KHz means that you will record one individual "sample" of sound 96,000 times every second. And 24-bit is how much data is recorded in each sample.
So your recording software/hardware might record a sample of
111010100010101011111010
and then, 1/96,000th of a second later, it might record
100101110110001101000100
that is a crapload of information when you think about it.
For example, since 24 bits = 3 bytes, that means it's gathering 288,000 bytes = 280 KB of data every single second.
A single-track, mono recording that's three minutes long would take up 50 MB of space on your hard drive.
Imagine if you had two stereo tracks and one mono. That same three minutes is now 250 MB!
Now, if you want to use 192KHz... double all those values. Now your project takes up half a gigabyte.
So I'd only consider it if you have a hell of a lot of space to burn. Also keep in mind that if you can record at 96KHz and only use up 25% of your CPU power, recording at 192 will probably use up 50%. Applying effects to your work will similarly take twice as long. The amount of RAM this will all consume will probably double.
I doubt it's worth it. CDs are recorded at 44.1KHz, so I suppose there is an argument to be made for sampling at a slightly higher rate to control the amount of information loss during editing and mixing and effects processing and whatnot. But I can't imagine 96KHz not being enough, at this stage in the game anyway.
Hopefully if I've made any egregious errors someone will correct me!
Let's say you're recording a single track in mono.
96KHz means that you will record one individual "sample" of sound 96,000 times every second. And 24-bit is how much data is recorded in each sample.
So your recording software/hardware might record a sample of
111010100010101011111010
and then, 1/96,000th of a second later, it might record
100101110110001101000100
that is a crapload of information when you think about it.
For example, since 24 bits = 3 bytes, that means it's gathering 288,000 bytes = 280 KB of data every single second.
A single-track, mono recording that's three minutes long would take up 50 MB of space on your hard drive.
Imagine if you had two stereo tracks and one mono. That same three minutes is now 250 MB!
Now, if you want to use 192KHz... double all those values. Now your project takes up half a gigabyte.
So I'd only consider it if you have a hell of a lot of space to burn. Also keep in mind that if you can record at 96KHz and only use up 25% of your CPU power, recording at 192 will probably use up 50%. Applying effects to your work will similarly take twice as long. The amount of RAM this will all consume will probably double.
I doubt it's worth it. CDs are recorded at 44.1KHz, so I suppose there is an argument to be made for sampling at a slightly higher rate to control the amount of information loss during editing and mixing and effects processing and whatnot. But I can't imagine 96KHz not being enough, at this stage in the game anyway.
Hopefully if I've made any egregious errors someone will correct me!
apl
Stand Up Comity
Good points, Kevin!
Most mixing and effects are linear processes in a differential equation sense, so oversampling doesn't improve anything. In fact the math going from 48kHz (or 96 or 192) to 44.1 is a major pain, and does introduce artifacts, albeit audibly indistinguishable ones.
kevinb9n said:
Most mixing and effects are linear processes in a differential equation sense, so oversampling doesn't improve anything. In fact the math going from 48kHz (or 96 or 192) to 44.1 is a major pain, and does introduce artifacts, albeit audibly indistinguishable ones.
