What do the bits, bit rate and sample rate of an audio file mean? Part 2

If it’s in a lossless uncompressed format, the bit rate is strictly equal to the number of bits * sample rate * number of channels. And typically, the MP3 bitrate you can see just represents how much capacity the format needs to describe this one second of audio.

MP3 is a lossy compression. In the compression process, some information is lost, but the lost information cannot be represented by the number of bits and the sampling rate. In general, the higher the bit rate, the less information will be lost. Mathematically, bitrate and sound quality are proportional. As for whether you can hear it or not, it depends on many factors. The MP3 algorithm is not complicated, of course, to understand it you have to learn what the Fourier transform is.

There is also lossless compression (representing APE, FLAC, etc.), which also has a bitrate, and this bitrate has nothing to do with sound quality. It also describes how much capacity the file uses to describe one second of audio content, but the same audio content can be compressed to different sizes (compression ratios), similar to zip compression ratios. No matter how big you compress it, in the end it can be restored to the same file. So if you see who is chasing lossless bitrate, you can basically conclude that the product is a bad pen.

What do the bits, bit rate and sample rate of an audio file mean?

For example, the common mp3 format audio source is 16-bit 320 Kbps or Sony Hi-Res 24-bit 192 KHz.

The stored sound data (for uncompressed PCM format) is equivalent to drawing a curve on graph paper, and its points can only be on the grid.

Then the sampling rate is how dense the grid is in the horizontal direction (how many grids per second); the number of digits is how dense the grid is in the vertical direction (how thin and how much storage each sample is stored), 16 Bit means that the range of values ​​of each sample point is from 2 to the 16th, i.e. , from 0 to 65535. The code rate is the multiplication of the two (how much storage the samples take per second).

First, there are two cases, lossy and lossless. After figuring out lossless and lossless, we need to figure out the sample rate and bit depth. After calculating the sample rate and bit depth, we need to figure out the difference between the ratio of bit rate and bit depth to sample rate and its impact on audio quality.

In order to store a continuous physical signal (well, tell me about Planck’s constant…) in a computer, it must be converted to a digital signal. In acoustics, a digital signal is a digital representation of the amplitude of the sound wave at any moment.

Sound waves are longitudinal waves, which are difficult to draw. The following figure is replaced by transverse waves (the concept of longitudinal waves is the phenomenon that the density of air or other media changes regularly due to energy. The peaks represent high density, the troughs represent low density, and the horizontal line is the average density, i.e. silent state)

Using high school physics, waves contain two dimensions, one is intensity and the other is time. “Number of digits” indicates how many levels the sound wave is divided into from the strongest to the weakest; “Sampling Rate” determines the precision of the time axis or the sampling density, i.e. the length of time represented by each red dot, and the code rate is one second The number of dots on the clock, multiplied by the space that each point occupies.

So the so-called 24 bits consist of dividing the intensity of the sound wave by 2 at power level 24, occupying 3 bytes of space. Obviously, the finer the grade, the more details are restored.

The sample rate is generally 44100 Hz for CD (Hertz = times/second), 48000 Hz for DVD, and 96000 Hz as standard. As with the number of digits, the more points you get in a single second, the more details are restored. Why does CD take this value? Because the hearing range of the human ear is generally believed to be between 20 and 20,000 Hz. It is necessary to represent a peak and a valley, and at least two sampling points are required. So the CD can represent the sound of 22050 Hz at most, but this sound doesn’t have any detail, because if there are only two peak and trough points, the mean waveform is completely lost. Therefore, there will be a higher sampling rate.

Audio bit depth, sample rate, and bit rate

About the sample rate. What is the most intuitive effect of sample rate?

Affects the expressiveness of the sound’s frequency range. The higher the sample rate, the larger the frequency range that can be expressed. 44.1KHz sampling rate can represent the frequency range from 0Hz to 22050Hz; 48KHz sampling rate can represent the frequency range from 0Hz to 24000Hz; 96 KHz sample rate can represent the frequency range from 0 Hz to 48000 Hz. The average frequency range that the human ear can hear is approximately 20 Hz-20,000 Hz. Combining the above two, if you see one parameter: 16Bit 44.1KHz , it means this digital audio can express “96dB dynamic range” and “0 Hz -22050 Hz” frequency range; 24Bit 48KHz , it means this digital audio can Performance ” 144dB Dynamic Range ” and ” 0Hz -24000 ”

Hertz” frequency range. (3) Audio bit rate, also called bit rate, or bit rate. Bit rate refers to the amount of information that can pass through a data stream per second, which can also be understood as: how much is used per second The amount of data in bits to represent In principle, the higher the audio bitrate, the better the quality However, if it is lossy compressed audio , different compression algorithms, even if the bitrate is the same, will lead to completely different sound quality results.Typical representative: 96kbpsThe sound quality of WMA audio format is obviously better than that of MP3 audio format than 96 kbps. Why is this? The difference is due to different compression algorithms and different data utilization rates. Another example, if the MP3 is compressed below 48 kbps, it’s already terrible. And if it’s format AAC audio, to the same 48kbps bit rate, the sound quality is obviously better than MP3. For lossless compressed audio, even if the bitrate is completely different, the final sound quality is the same

. For example, if the same WAV file is compressed in FLAC and APE formats, the bit rate of the output file is not the same, but the sound quality is the same. Even if it is the same format, the compression level is different and the bitrate is completely different, but the end result, the sound quality remains the same (but when encoding and decoding, the CPU usage is different and the encoding time is also different).

MP3 is just an audio format.

And the audio has several important parameters, such as KHZ, BIT, channel, KBPS, etc.

And the format is different, the algorithm is also different, so even if the above parameters are the same, the format is different. The sound quality will also be very different. Among them, VBR is a dynamic sampling. For a detailed and complete explanation, please refer to the description below.
After reading it patiently, you will be able to say a thing or two.

Audio sampling explains the digital audio system The original sound is reproduced by converting the waveform of the sound wave into a series of binary data The equipment used in this step is an analog-to-digital (A/D) converter, which samples the sound wave at a rate of tens of thousands of times per second. A sampler records the state of the original analog sound wave at a given time, called a sampler. A series of samples can be connected to describe a sound wave. The number of samples per second is called the sample rate or sample rate, and the unit is HZ.

(hertz). The higher the sample rate, the higher the frequency of the sound wave that can be described. The sample rate determines the frequency range (equivalent to the pitch) of the sound, which can be represented by a digital waveform. The frequency range is often called the bandwidth. To understand properly, audio sampling can be broken down into the number of bits sampled and the frequency of the sample. 1. The number of sampling bits The number of sampling bits can be understood as the resolution of the sound processed by the capture card. The higher the value, the higher the resolution and the more realistic the sound recorded and played back. The first thing we need to know: sound files on the computer are represented by the numbers 0 and 1 . So the essence of recording on the computer is to convert the analog sound signal into a digital signal. On the contrary, during playback, the digital signal is restored to an analog sound signal output. The capture card bit refers to the binary digits of the digital sound signal used by the capture card when capturing and playing sound files. The bits on the capture card objectively reflect the accuracy of the digital sound signal’s description of the input sound signal. 8 bits represent the eighth power of 2–256 and 16 bits represent the sixteenth power of 2–64K . For comparison, for the same music data, a 16-bit sound card can break it down into 64K precision units for processing, while an 8-bit sound card can only handle 256 precision units.

It causes a large signal loss, and the final sampling effect is naturally incomparable. All the major products on the market today are 16-bit capture cards, not 64-bit or even 128-bit as some ignorant marketers advocate, who confuse the concept of polyphony in capture cards with the concept of bits of sampling. Although the EMU10K1 chip used by the most powerful capture card series is claimed to be 32-bit, it is just a Direct Sound acceleration-based multi-audio streaming technology, and its essence is still a 16-bit sound card. . It must be said that the 16-bit sampling precision is more than enough for computer multimedia audio. 2. Audio Sampling Level (Audio Sampling Rate) The digital audio system reproduces the original sound by converting the sound waveform into a series of binary data. The equipment used to accomplish this step is an analog-to-digital (A/D) converter, which takes each sound wave, and each sampler records the state of the original analog sound wave at a given time, which is called sample. A series of samples can be connected to describe a sound wave. The number of samples per second is called the sample rate or sample rate, and the unit is HZ (Hertz). The higher the sample rate, the higher the frequency of the sound wave that can be described.