所谓mp3的音质（The sound quality of mp3）

所谓mp3的音质（The sound quality of mp3） 所谓mp3的音质(The sound quality of mp3) The sound quality of the so-called mp3 (2007-02-16 20:18:20) Throughout the mp3 users now, compare the eac's the general standards for grab rail + lame compression, experienced in the process of making friends in this can figure out some tips, in view of the different music use different parameter Settings and the compression ratio, from the standard 128 KBPS to maximum of 320 KBPS, but the rate between the gap and the effect of the difference between how much?? How is the most appropriate compression ratio, CBR and VBR should choose the better, etc. These problems are often discussed, in order to find out these details, I specially made a test, target-oriented here and everyone share once I have some feelings. "Gran goldberg, concerto" the first rail, Munich Bach orchestra, eac, grasp the rail software compression software CD, ex player software fooba2000 v0.8 version, headphones listening equipment was sound, beautiful er6 and schur e3c. Because of the details of the classical repertoire, the band is large, and the demands on sound quality are relatively high, so it is obvious that the details of the different processing methods can be reflected. (the spectrum control chart is presented later) I used the rac to catch the track, then used the lame mp3 encoder (vision 1.92 engine 3.92) engine in the software to process the wav files, and I tested the parameters on the lame parameters one by one to select the good results: First parameter to the thread priority choose top and bottom, respectively, under the same other parameters of compression contrast, found that whether the thread priority level has no effect on sound is, the generated file size is the same, compared to sound the same, so the parameters on the quality does not have any effect. The second parameter is the version, can choose to have mpegI, mpegII and mpegII. V, in the same way, the other parameters are determined, with the three options for compressing the three times, compared to come down, although the three ways of files are the same size, but still mpegI actually listen to feeling a little better, in the low frequency is slightly small compression ratio, but high frequency distortion slightly more, listen to the voices and a kind of the pop music is more suitable, to listen to classical type with mpegI is good also, the bottom of the sound is better, but if it is a high frequency of solo music, violin, etc. Can recommend mpegII V type, there will be a better effect. The third parameter is the most important, this is the bit rate, choose him directly affect your mp3 file size and listen to the feeling, the distortion of high compression ratio, compression ratio of small distortion is small, but how can I find a our own balance both can accept? This needs carefully grope in the experiment, considering the low bit rate voice file quality is not very suitable for playing music, so the lowest at 128 KBPS, in turn, 128, 192, 256, 320, a total of four types of fixed rate file to compare and test. Still more rough 128 KBPS compression ratio, the high frequency part after compression distortion is very apparent, sounds hollow, dry, harsh, and often have a fool fool's voice, the simple sense of Musical Instruments is very poor, part of the instrument are changed flavor, may even listen to wrong, The volume of a piece of music in 3:39 is 3414kb. Although the volume is small but the sound is not satisfactory, there is a large defect. 192 KBPS bit rate compression effect but is much better than 128, the first sound is solid, at least not the sort of empty feeling, high frequency distortion are much smaller, the sound is compact, small noise, clean, and reached a relatively ideal listening effect, just due to compression was more severe, the details of the performance is still not very good, the simple sense of Musical Instruments, especially the wind instrument is very hard, not real, a lack of musicality, compressed size is 5123 KB, I think the compression rate in 128 ~ 256 m to use the better capacity of mp3 players, both can satisfy the basic listening, is suitable in size, 128 m can store about 128 minutes of music, 256 m double as 190 minutes of music. 256 KBPS compression ratio than 192 nature is the sound quality more on a higher level, take the music started 10 seconds, said the cello part of low frequency obviously particles are much smaller, sound more natural, smooth texture and texture is also more clearly, more details, the rendering of atmosphere is more outstanding, the back of the parts of rotation and expressive up many tunes, definition of the small signal also improve some big signal, sound resistance to listen more carefully, but at the same time, the file size is increased to 6831 KB, it is for the 256 m mp3 player can afford, calculate is not difficult to know, according to a 256 bit rate can hold about 135 minutes of music, generally speaking is enough, 128 m is a little less, can only support a little more than an hour, so recommend the use of 128 m 128 bit rate. 320 KBPS is the maximum rate, if you can provide the final generated file is 8592 KB, is about 8.4 M, compared to wav files for 37 M basic compression ratio was 4.5:1, but the generated mp3 files have sound distortion is very small, compared with other rate 320 natural advantage is obvious, timbre, details and so on are very exquisite, is basically reached the same copy of the CD sound quality, especially in the CD player with mp3 function and sounds, basic no difference, but I use a more upscale earplugs, resolving power is higher, coupled with my experience and ability of music and equipment, contrast wav files or listen to the many different, first of all; Compressed mp3 sounds a little shrink feeling, as a whole is dry, no sounds fresh dynamic wav files, details at the end, harmonic and spatially, the degree of separation is not as good as wav files to high quality, but on the tone color is very close to, only the expression difference, digital taste too strong. So if you are using a ipod etc. This kind of micro hard disk type walkman player, I recommend that you use the compression ratio of 320 KBPS, such feelings can get relatively better listening, directly to wav, of course, is the best ~ ~ no compression, no loss, but now nothing walkman also support ape the lossless compression, otherwise can choose a variety of again. It's all about the compression ratio of the fixed code rate, which is called CBR in the lame. Actually lame the biggest characteristic is that it provides the user with variable bit rate VBR compression way, in this way some pause, simple signal will automatically reduce the bit rate to reduce file size, is a very good coding way, but how to choose the VBR lowest and highest bit rate range to get the most appropriate documents and the sound quality? It is a need to experiment to solve the problem, also considering the value on the basis of 128 KBPS, so we chose 96 KBPS to 160 KBPS for a range of compression, compressed file size is 3801 KB, more than 128 KBPS CBR is only 387 KB, but the sound is definitely raised a big level, first of all, high frequency distortion of little more than half, at least while detail place also a lot of noise, but the first listening is better than 128 KBPS too much, after compression rate is an average of 147 KBPS, is also very save space. Then tested 96 KBPS to 192 KBPS and 96 KBPS to 224 KBPS and 96 KBPS to 256 KBPS and 96 KBPS to 320 KBPS, found that they and the maximum CBR compressed sound very approximate, namely 96 KBPS to 96 KBPS VBR voice approximate 192 KBPS of CBR, but is 4481 KB in size in the former and the latter is 5123 KB, so much as a compromise method of the pursuit of high quality and save space, VBR is useful too. On the other hand, because of the change of code rate, the stability is naturally less than that of CBR. Fourth, the model parameters, a stereo, J - stereo, Forced - stereo and mono, contrast tests of the standard stereo effect is best, although compress the file size is the largest, but consider the smaller file size difference and the sound quality difference and listen to the feeling, still feel stereo is more ideal. Fifth, compression way, the software has a VBR - old, VBR - new, and the other two, but only two kinds of works, in front of the contrast between old and new two VBR encoding, found that if in sound quality, and the comparison of the old is exquisite, but old coding compression speed is slow, about five or six times the new encoding, a tune made about 3 or 4 minutes, very bad to use, and the file size is big by about 10%, so it is recommended to use VBR - new, more convenient to use. The mp3 player is not a word made out of thin air, but an abbreviation for a technical term, the abbreviation of the mpeg-1 audiolayer-3, what does that mean? Let me read it to you: the mpeg-1 audiolayer-3 is the audio part of the third layer protocol in the international standard technology protocol for mpeg1, which is an audio format, doesn't it look very complicated? It doesn't matter to us, first of all, the word mpeg - the acronym for Moving Picture Exp - erts Group, which means dynamic image compression processing Group. The group is quite severe, specialized research and development of dynamic and static video (with audio) in almost all the international technical standard, we see from TV to film, from the VCD and DVD application of their research achievements, mp3 is part of their mpeg1 agreement, audio part in the third layer, so as to mp3, As far as I know the birth of the mp3 player is the mpeg group staff found in the case of not conscious, it is as a way of fetching is produced, CD tracks in the test of the time up to an affiliate product, but is this small accessory product evolution is out of our current most popular music formats, so why? Let's go and see. First is small size: according to the different compression ratios can get the file size of your size, but compared with the original wav format, that's too much smaller volume, the file is small in size, storage space is small, to play the hardware aspects and so on volume, cost will drop significantly, so there will be a good market, plus there are mp3 file download, compared to the CD player users still buy software, nature and save a fortune. Again: follow one's inclinations, small volume of the player can easily carry, you can use what I love to use, and because the player circuit part of the relatively low power consumption, so the single battery work time is long, save a lot of replacement battery trouble, and be in the home can put their favorite tunes together to listen to, and not under the control of the molding software of audio and video publishers, freedom is very big, this reminds me of the jay Chou for m-zone advertising slogan of "my site, I decide!" Really cool ~ ~ ~!!!!! After saying such a bunch of relevant words, we must let the people wait, let us go into the main body, to thoroughly dissect every aspect of the mp3. I. basic knowledge: Digital compression audio and mp3 technology: To say mp3 I have to talk about digital compression audio, well, digital compression audio is just a compressed digital audio, but what is digital audio? What is compression??? Below are some more professional explanations, and interested friends may have a closer look! In the digital world of computer, voice is stored in a digital coding way, different from the traditional analog audio in life, because the computer can only record the number of 1 s and 0 s, so you have to quantify the analog audio sampling, according to nye chief, NYQUIST sampling theorem, with twice the rate of a sine wave frequency sampling can completely restore the waveform truly, therefore a digital recording the sampling frequency of the wave is directly related to its highest frequency index. For example, using the sampling frequency of 44.1 KHZ sampling, can restore the highest 22.05 KHZ frequency -- -- -- -- -- the value of the auditory slightly higher than the limit, so the a/d conversion often use more than 44.1 KHZ sampling frequency, but the acoustic data file size for sampling frequency (in Hz) and sampling rate (generally) measured in bits and differ somewhat, sampling frequency, the higher the storage space is required, the greater the sample of the same ratio, the higher the storage space is required, the greater the this creates difficulties in actual operation, thus the emergence of compression technology becomes inevitable. Digital compression is to complete the sampling of the original digital audio files have to "lose weight", by using some effective algorithms and way, remove redundant information in the file, so can reduce the volume, Ease of use, the concrete operation is too complex and subject (even look at me this computational mathematics learning difficulty), so not to explain one by one, only here briefly explain to involve in the compressed MP3 compression coding technology, MP3 compression using the five important technologies: minimum auditory Threshold determination (Minimal Audition Threshold), the Masking Effect (Masking Effect), a storage tank (Reservoir of Bytes), joint stereo (JointStereo) and Huffman coding (Huffman Code). Depending on the computing of these technology computers, we can greatly reduce the file size of the digital audio, and the new audio format generated after compression is the familiar mp3. Ii. Principle of compression: No matter what technology it has its own reason, audio technology, because it is directly related to our sense of hearing, compressed mp3 why sounds or very close to the voice of the CD?? It has a lot to do with human anatomy. Experiments have shown that the range of sound frequencies that humans can hear (audio) is 20Hz ~ 20kHz, but the human ear does not respond to the sound of the entire audio frequency band. 2 ~ 5kHz is the most sensitive frequency band of the human ear. According to its characteristics of the entire audio spectrum into multiple critical band, because the human auditory system is based on frequency to distinguish of sound energy, any tiny voice frequency will be critical for masking effect and the band is covered within the volume high voice. The MP3 is not quantified to eliminate the sounds that the human hearing system cannot detect and achieve the purpose of compression. As you can see, MP3 is a lossy audio compression encoding. So no matter how high the rate of mp3s it's lossy, the quality of the sound is lower than the CD, but the number of drops is acceptable, which varies from person to person. Iii. Sound quality after compression: That's the most important topic of all. What kind of standard can a compressed voice achieve? So I'm going to give you a graph here, to look at the reference values in the standard of mpeg. It can be seen that, since the compression ratio is different, the quality of the sound files obtained is gradually improved, and the 128kbps basically reaches the CD quality. Is that so??? I have to answer is: of course not. This is only an official reference data table, in the course of practical use, the bit rate is 128 can't show the CD sound quality, although mp3 compression using software algorithm in the process of different will have some differences, but the compression ratio is the most fundamental to limit its performance. On the market is not just only a lame a compression software, this is not only a mp3 audio compressed format, other company also has a lot to use compressed format, just because of a variety of reasons not popularized, such as mp3, but as a member of the family of compressed audio, I also introduce you to here, can let people know more about things. 1, the APE: APE format: Monkey "s Audio (www.monkeysaudio.com), simply saying that APE is an Audio lossless compression format that is half the size of a wav file, but the quality is not reduced at all. The quality of the CD can be restored to the maximum extent, depending on the quality of the track. If grasp the rail, the quality of ape sound quality will be similar to CD this is out of reach for mp3 lossy compression, such as, nondestructive, of course, price is the increase of volume, ranging from 200 MB to 400 1 CD so if there is enough bandwidth, ape to net CD to bring more convenient, solved many classical music enthusiasts the needs of high quality CD. A 80 minutes of music CD size is about 700 m, for such a big size of transmission, obviously it is very inconvenient, then with the Monkey "" s Audio - to WAV files compressed into half (or less) so as to the size of the APE document, and ideally, when you put the APE decompressed into WAV, quality without any damage. Advantages: high compression ratio; Sound quality; After loading the plug-in, you can play it directly with Winamp, which is better than any of the parameters of Mp3 or Mp3 Pro. Faults: compression and decompression time too long. 2, MP3 Pro To reduce the distortion, MP3 Pro this new MP3 format appears. It will be separated into the audio frequency in the low frequency and high frequency of two channel and the traditional MP3 encoder is responsible for the low frequency part, and the other using SBR technology (Spectral Band Replication copy)/frequency decoder is responsible for the high frequency part of the last two play at the same time, in the MP3 Pro player software to get significantly more than the same Bitrate of MP3 sound quality. Not only is the high frequency details rich, the trembling phenomenon is also imperceptible, very close to the original WAV file. MP3 Pro USES of SBR technology, the high frequency isolated and music are not directly coded so simple, but according to the analysis of low frequency signal transmission by the original encoder, so as to rebuild the high frequency signal. 3, Real Audio Real Audio, extension RA: this format is really the soul of the web, and its powerful compression and minimal distortion make it stand out in many formats. As with MP3, it is also designed to solve the network transmission bandwidth resources, so the main goal is compression ratio and fault tolerance, followed by sound quality. So the new music auditions we encounter on the Internet are all in this format. 3, WMA WMA is the file format of Windows Media Audio coding, which is developed by Microsoft. WMA is not for single market, but network! The competitor is the famous Real Networks in the Internet media market. Microsoft claims that WMA can reach the sound quality of a CD at a rate of just 64kbps. Unlike previous coding, WMA supports the anti-replication feature, and she supports the addition of protection through Windows Media Rights Manager, which can limit the playback time and the number of playing times and even the playing machines. WMA support flow technology, that is, to read while playing, so WMA can easily realize the online radio, because it is Microsoft's masterpiece, therefore, Microsoft on Windows added support for WMA, WMA has a good technical features, under the promoting of Microsoft, this format is accepted by more and more people. Compared with ra, copyright is probably the most hated place for users. 4, the ACC: AAC (Advanced Audio Coding, Advanced Audio Coding) is the technology that dolby LABS provides for music communities. AAC claims to "have the maximum capacity of 48 channels, with a sampling rate of 96 KHz, and can provide the equivalent of the itu-r broadcast quality at the data rate of 320Kbps." Compared with MP3, its sound quality is better, and it can save about 30% storage space and bandwidth. It is developed in accordance with mpeg-2 specifications. Panasonic mp3 products are adopted the encoding, and compatible with mp3 format, of course, I find it is very nice aac audio formats when using, 128 KBPS aac enough to compete with 224 KBPS mp3, almost half the space is small, but on the space and structure after aac and mp3 encoding style is different, like it or not belong to many things. 5. ATRAC 3 / ATRAC 3 plus: ATRAC3 (AdaptiveTransformAcousticCoding3) developed by SONY of Japan, was adopted by the MD ATRAC upgrade, the compression ratio (about 2 times of ATRAC) and sound quality with MP3. The principle of compression includes simultaneous masking, aging masking and other loudness curves, which are roughly the same as MP3. The copyright protection feature of ATRAC3 is OpenMG. At present, the portable player corresponding to ATRAC3 is mainly SONY's own products. However, in February 2000, the company signed a patent licensing agreement with Fujitsu, Hitachi, NEC, Rohm, sanyo and TI to manufacture and sell ATRAC3 for the use of LSI. Compared in MDLP compression recording CD and mp3, ATRAC3 had the voice of the advantage of, is not very convenient to use, copyright issues, however ATRAC 3 plus due to the advanced algorithm, on the sound quality has been further improved, but only on SONY's products are widely used, the more depressed. 6, VQF: VQF refers to TwinVQ (transform-domain Weighted INterleave Vector Quantization), which is an audio compression technology developed by NTT Human Interface Laboratories of Nippon Telegraph and Telephone group. The technology is supported by the famous YAMAHA company. VQF is the extension of its file. It and MP3 players are of a similar implementation method by using is really algorithm to compress voice, but it compared with MP3 compression techniques have fundamentally different: VQF to music rather than sound is the purpose of compressing, therefore, VQF adopted is a known as the "Vector Quantization coding (Vector Quantization)" compression technology. The technology will vector quantization audio data first and then the audio waveform similar waveform in unity and smoothing, and strengthen the outstanding part of the human ear is sensitive, finally, the vector data processing after the quantification to be compressed. In my own use, I feel that the VQF is a lot better than the mp3 volume of the same sound quality, but the high bit rate is not so good, so few people use it.