MP3与无损(flac.APE)的本质区别和音质差异

MP3与无损(flac.APE)的本质区别和音质差异 MP3与无损(flac.APE)的本质区别和音质差异 The essential difference between MP3 and flac.APE and tone quality difference... Everyone is listening to the format should be MP3 now (according to the survey, ninety percent of the oil is listening to MP3., which includes the use of professional players, mobile phones, computers and other songs of the engine oil. While seven percent of the oil will listen to Ogg, AAC and other small format music, while only three percent of the oil will listen to FLAC, APE etc.), lossless music rarely listen to music, a large part of the reason is because it is a portable player at present on the lossless format support is too small, even if there is such a product is often the price not cheap, daunting. A lot of oil, and now do not know the lossless music, and even a bit have not heard (I have a lot of friends are like this), today I simply present the mainstream music format introduced it. When the singer's song recorded into the original file of digital music, then the music file is very large, it is not easy to store, so people put it in compression, almost all popular digital music formats are compressed, like MP3.aac.ogg. flac.APE, but their encoding just not the same way, we often say that the lossless (flac.APE) that is lossless compression, while those in front of the MP3 small format is lossy compression. Lossless compression is like trying to flatten a piece of bread. The volume is smaller, but the quality remains the same. We recorded the sound, in fact most because the frequency is too high or too low, is our ears hear not (some animal such as bats, dogs can hear a part), since not to hear that we still have what? So we have to listen to this part of the voice data delete cannot see, the music data is that part of the left ear can hear becomes very small, easy to store, this compression method called lossy compression (such as MP3). So what's the difference between lossless and lossy sound quality? Have you had any experience like that, the same song, the next MP3 and a FLAC, and then put on headphones, but we can't find the difference in sound quality at all! But if you listen to the MP3 all day long and listen to the whole day without damaging it, it will be different! It's easy to get tired of listening to MP3 ears all day long, but it's no problem hearing it (as long as you don't fall asleep)! What is the cause of this? That's what I said today about the difference between lossless and lossy sound quality! We know that music is rhythm, rhythm and rhythm are different, such as the treble and bass between both the song and how quickly or slowly, it will have an excessive, lossless music feature is both in rhythm with no sound (whether we can hear it), the data is always the steady flow, let your headset in magnetic fast stable vibration, sound natural and true; and MP3 lossy format because of rounding the human ear to hear some of the data, so you in the magnetic block vibration headset is not natural, resulting in a sound vibration, when no sound is shaking this magnet will produce excessive vibration is not smooth, burr murmur, maybe you can't hear, but such a long time listening ear will uncomfortable it (of course, Some good headphones can avoid this glitch, but most of them won't be able to afford them! Often people laugh at the music fans, notice that the songs are innocent and the fever is reasonable! He that listens to the ear does not necessarily have a fever, but he who has a fever will surely hear nothing! 1.wav audio format of the three major parameters, and the parameters for the meaning of audio files The wav file has 4 parameters: sampling frequency, channel number, quantization bit number, and rate of 4 The best of these 4 parameters is the sound track number, so don't introduce this parameter Then the parameters I will introduce are sampling frequency F, quantization bit B, and rate R sampling frequency In the three parameters inside the most important is the sampling frequency, behind the two parameters are based on the transmission and storage process according to the requirements and get, but it is the sampling frequency, signal the analog world to the digital world bridge. Before talking about the sampling frequency, we can first recall the situation when we studied parabola in junior high school. In junior high school, when the teacher taught us to draw parabola, what method was used to draw? If you think back, you should remember, it's the 5 point method. Yes, the parabola can be drawn with 5 points. What is an audio signal is actually cosine wave, but the frequency and amplitude of this Yu Xianbo are variable over time. We want to record this audio signal, it is impossible to record the value of each moment, but we can refer to the parabola method to describe the audio signal exactly as little as possible. And sampling frequency, it is doing this work, that is, in a second we want to record the number of audio signals, how many points can be approximately accurate expression of this audio signal. In signal processing, there is such a theorem called Nyquist's theorem. How do you get this theorem? You don't need to know. This is a signal processing professional. You need to know, like me. Undefined We need to know is that this theorem tells us that if we want to accurately record a signal, the sampling frequency of two times, we must be greater than the maximum frequency is equal to the audio signal that is the maximum frequency. that is F>=2*fmax. In the wav format, F=44.1kHz. We know that the frequency range of the human ear is 20-20kHz, that is to say, if we want to accurately record this audio signal, the sampling frequency is the lowest, at least 40kHz. As for why 44.1kHz, not other frequencies, I'm sorry, I don't know. Undefined But, at least we can be sure that the 44.1kHz sampling frequency can record exactly less than 22.05kHz of the audio signal, this is enough. Quantization Bits Although we have the sampling frequency, we can record the audio signals accurately. However, the recorded audio signals are analog and can not be processed by the computer. Here we are going to come up with a new concept, analog and digital. There is a difference between analog and digital. Let me give you a brief introduction. For example, 0-1 this range. A line segment, we can arbitrarily take a point, the value of this point can be determined, the point of the range of values can be determined, but the point of the number of values can not be determined, and this is the analog. A computer that cannot determine the number of possible values cannot be processed by a computer. The quantity is the same as the rest, and the third is not the same. The number of values can be determined, so that the computer can handle it. 0-1 this range, depending on the accuracy requirements, we can determine the number of values required. And the quantization bit, which is doing this job, determines the recording point of the audio signal and the possible number of its values. We know that the quantization bit wav of B is 16, and this is a 2 digit number. He tells us that the number of points at which a record point can be 2 is 16, or 65536. 0-1- split it 65536 times, and I think the precision is enough. Undefined Rate Now, the sampling frequency and quantization accuracy are told, the turn of the code rate, first drink saliva first. Undefined How do you get the rate? Very simple, that is, sampling frequency, X quantization bit, X channel number, that is, R=F*B*2. R=44.1kHz*16b*2=1411.2kbps~=1411kbps. This is the rate 1411, although the rate is calculated, but he has an exact meaning, that is, it can store the amount of information in a second, remember that the amount of information. Here, you might be reminded that MP3, 320kbps, AAC, 512Kbps, lossless compression format 700+kbps. Then some people wonder, is not it, the greater the rate, the better? For lossy formats, then the greater the bit rate, the better However, there is a premise that the converted song must be a lossless format that is burned from the genuine CD, and the transformation is the same lossy format, such as MP3. Otherwise, you convert a 128kbs MP3 song into a 320kbps bit rate MP3, and the sound quality will not improve. For lossless compression formats, the size of the code rates will be meaningless. The size of the bit rate tells you just how well his compression algorithm is good enough. But I have to remind you that this bit of code rate is just like this. He can't tell you that the saved information is good or bad. He can only tell you that he has saved so much information. Yes, he's actually a warehouse. He doesn't care what he puts in the warehouse. He just keeps it full. Undefined Well, here, the three parameters of the wav format are finished, there may be a lot of people wondering why I talk about this antique wav format, not MP3 ah AAC these lossy format, or FLAC, ape lossless compression format. The reason is simple, because wav is the nearest analog digital quantity, the most primitive data, and the latter format is based on WAV, according to their own characteristics to deal with it. What's more, the three parameters mentioned above are still used in the later format. Naturally, first introduce WAV, undefined 2. compression principle in lossy format In this section, as well as the following lossless format compression principle, Similarly, a 10 second audio, after lossy compression, the rate value of 320kbp, then the size is only 787KB!!! 1/5 for wav format!!! Use as little data as possible to store as much useful information as possible, and do so in a lossy format!!! That's why the format is so popular. Undefined 3. compression principle of lossless compression format With the development of storage technology, the amount of information we can store becomes more and more big, and the storage of wav format becomes nothing interesting. Since there is no pear, why should the introduction of lossless compression format? The reason is simple, since I 40MB can store 2 lossless compression formats, why do I still store 1 wav formats?. So lossless compression formats have developed. Lossless compression formats and lossy formats have one thing in common, that is, compression. The difference is, lossless. Then, how can we do lossless compression? We can refer to second methods of lossy compression. For example, the same 5 record points, 0.4, 0.4, 0.5, 0.5, 0.3, if we want lossless compression, we only need to record 0.4 (2), 0.5 (2), and 0.3 (1). In this way, we only need three record points to record, so we need 5 record points, and it is compressed as well. What's more, lossless compression. That's one way of thinking, but he tells us that lossless compression is possible for information processing. To record an audio complete, do not need to go to WAV format, lossless compression will do. Similarly, the rate of this parameter to compare, a 10 second audio, after lossless compression, the rate value of 727kbps, the size of 1.73MB. It's about half of wav. Large capacity player support lossless compression format, small capacity player is fun, lossy format, each has its own location, technology development is indeed a good thing, undefined When I see someone asking whether WAV, FLAC and ape make any difference, I'll make a summary here. After the introduction of the above algorithm principle, we can see that, if only from the file itself, WAV and all other lossless compression formats are not different from the saved information. In forums, people often see the two questions related to lossless format: lossless compression, whether there is any difference between the format and lossless compression format, and whether there is any difference between wav and lossless compression formats. The first question, I can answer now, yes. But the difference is not in the whole course of the information recording, but in the difference between the compression algorithm and the format adopted by the algorithm. That's why the same song, ape format is smaller than FLAC, because the algorithm is different. As for the acoustics, the performance will come with second questions, in the fifth part Frequency spectrum analysis of 4. audio files This section analyzes different audio formats and different bit rates in the same audio format. Provide a reference for those who worry about which audio format to choose. The audio formats to be analyzed are MP3, AAC, three formats, lossless format as reference format. Since I'm using FB conversion, the MP3 format only has the VBR mode and the highest cbr320. So, May be familiar with the code rate is different. However, I used parentheses to indicate the corresponding bit rate value, which is a probable value, not necessarily, but it can be referred to. The MP3 bit rates are VBR, V5 (~130kbps), V2 (~190kbps), V0 (~245kbps), and CBR's 320kbps. In order to correspond to the VBR schema of MP3, AAC uses the VBR pattern as well The AAC bit rates are Q04 (125kbps), q05 (175kbps), q06 (225kbps), Q08 (325kbps), Q10 (400bps) The reason for this choice is that people are accustomed to the rate values of 128kbps, 196kbps, 256Kbps and 320kbps. When selecting the conversion mode of MP3, the reference rate is as close as possible to the customary rate values. Because AAC is much better than MP3 in coding, the AAC conversion mode is the converted file size as close to the MP3 size as possible. As for the Q10 model, it is in contrast to lossless compression formats. First, compare the format of each audio file with the corresponding bit rate [attach]76054[/attach] Prior to the description, the audio file interception is Eason's ten years (40s-60s) this section, using the online download lossless, interception software using goldwave. First, a volumetric analysis. Obviously, the smallest volume here is V5MP3 (130kbps), corresponding to the q04aac (125kbps). The second grade is V2MP3 (~192kbps), corresponding to q05aac (175kbps). The third grade is V0MP3 (245kbps), corresponding to q08aac (225kbps). The fourth grade is cbrMP3 (320kbps) and the corresponding q08aac (325kbps). Finally, q10aac (400kbps) and the corresponding flac. If the original disk is genuine, its corresponding quality is lower, junior, intermediate, advanced, and highest. First on the highest level of spectrogram Nondestructive spectrogram Q10aac spectrogram Let me explain first, the abscissa is time, the ordinate is the frequency, the whiteness of the point is the intensity of sound corresponding to the time and frequency. So call spectrum. By contrast, we find that the frequency reproduction range of the q10aac in the sound is not different from the lossless, and the dry to 22kHz without pressure. However, the frequency reproduction strength of the sound is defective, and the frequency and sound intensity are absent in some time periods Pictured above Q10aac defect The red circle is part of the missing part. As you can see, q10aac still fails to record perfectly in detail (after all, it is harmful), but, as a whole, the information it retains is very close to lossless. Personally, as a high fidelity format, the high bit rate AAC is qualified. Then the spectrogram comparison chart of high grade Q08aac 320MP3 By comparison, we can find that the spectrum of 320MP3 is higher than 20kHz across the board, the frequencies are removed, while the q08aac is dry to 22khz pear, in detail, the two are similar, I am not on the map, so, this round of AAC won. Intermediate grade spectrum comparison chart Q06aac V0mp3 To the intermediate level, the MP3 format reaches 19kHz in spectral reproduction, while AAC is 18kHz. In terms of sound detail, the two are basically similar, this round is MP3 format wins. The primary grades of spectrogram Q05aac V2mp3 At the primary level, the frequency of the MP3 format averages at 16kHz, and many can go up to 18kHz, while the AAC format is the same. However, in terms of detail presentation, AAC's voice over 16kHz is much more than MP3's.. And below the 16kHz part, the two are pretty much the same. So, this round, AAC wins. Low-grade Q04aac V5mp3 In the lower level, MP3 is one size fits all to 16kHz, and AAC is an average 16kHz, many capable to 17kHz. Below the 16kHz part, the AAC record is not MP3 complete. Personally feel that this bout of play and. Through this comparison, we can find the difference between the concept of MP3 and AAC in lossy compression, and MP3 as far as possible in the frequency range that he can record. While AAC is sacrificing the low frequency part of the details in exchange for high frequency reservations, at low bit rates, who is good, who does not look good, personal choice. At high bit rates, AAC is clearly superior to MP3, and AAC can retain as much as possible in the low frequency part or the high frequency part, while the MP3 is still powerless in the high frequency part. 5. personal recommendations for audio format selection In the third part, I've raised two questions about the quality difference between lossless compression formats and the difference between wav and lossless compression formats, where I'll answer In the fourth part, we understand the advantages and disadvantages of lossy audio formats through the spectrogram comparison, the lossy format selection as the groundwork However, before I answer the third part and recommend the format, I'd like to introduce the flow chart of the audio when it is being played WAV format: wav data stream - "DAC -" filter circuit - "amplification circuit -" output Lossy format: lossy data stream - decoding - "DAC -" filter circuit - "amplification circuit -" output Lossless compression format: lossless compression, data stream -- decompression -- DAC -- filter circuit -- amplification circuit -- output Explanation: the function of DAC is to change digital signals into analog signals. The filter circuit is to remove the unwanted frequency components and amplify the circuit. This is to enlarge the analog signals so that they can be output easily Through the playback flow chart, we can see that wav format playback is the simplest, while lossy formats and lossless compression formats are one step more. In the signal processing, there is such a sentence "error is everywhere."". The implication of this sentence is that the greater the probability of error generation and the greater the accumulation of errors, in each step of the process. For lossy formats, the only advantage of a lossless format in format is that the compression ratio is large enough, and this compression rate is based on sacrificing sound quality, and the sound quality is not as lossless as normal. "Are there any differences between" ape "and" FLAC "?, "Wav is better than lossless compression formats," I now answer one by one Before you answer, we compare the lossless compression format with the wav playback process, and you can see that the lossless compression format has one more "decompression" step than wav. For different lossless compression formats, the algorithms for decompression are different. Those who say that wav is better than lossless compression format make their views reasonable. Why would I say that? No, lossless compression, since it's lossless, there should be no difference. Yes, in terms of the information integrity of the document, yes, WAV and other lossless compression formats do not make any difference!!! The difference is not in the file itself, but in the playback process!!! Because the lossless compression format is one step more than the wav format in the playback process!!! That's decompression!!!! Suppose we can guarantee that the following DAC, filter circuits, and amplifier circuits are the same. However, more understanding of this step of compression may have an impact on the quality of sound. Why do I say this? Decompression may affect sound quality. I don't know the principle of influence, but the error is everywhere, as we have said before. This means that the error generated by this step may affect the overall quality of the sound. As to whether this influence can be ignored, it depends on the manufacturer's skill. Similarly, the lossless compression formats do not look at the final difference in quality, nor is the format itself, but whether the "decompression" is good enough to ignore the impact of the error. With the above broadcast flow is introduced, and the fourth part of the spectrum analysis, we can according to the use of the environment, the format recommended. Regardless of the size of audio files, the preferred choice for quality is of course lossless format. If your front end can support lossless compression formats, and your playback system can recognize the difference between wav and lossless compression formats (that is, you can hear the error of the decompression, wav). Otherwise, on lossless compression format. Concerned about the size of audio files, but also the pursuit of sound quality, the high bit rate aac. However, here is a premise, you play system to be able to listen to the high rate of AAC and 320MP3 difference, otherwise, or obediently on the MP3, do not toss about. If you don't care about the sound quality, just go with the MP3. There's a big premise here. These audio files are really lossless, not fake. Otherwise, for a large warehouse, the contents are still garbage