![\"WAV](\"https:\/\/www.off-the-beat.com\/wp-content\/uploads\/2020\/02\/WAV-vs-FLAC-1024x487.jpg\")
<\/p>\nWhy WAV and FLAC make fun of saying “the quality of the sound is different”?<\/p>\n
<\/p>\n
By observing the area of \u200b\u200baudio files (audiophiles), sound quality is different between WAV and FLAC!<\/p>\n
![\"FLAC](\"https:\/\/appuals.com\/wp-content\/uploads\/2020\/03\/intro33.png\")
<\/p>\n
There are people who scream. In this article I would like to consider if it is true.<\/p>\n
Differences between
\nWAV (Wave) is a file format that can store linear PCM, and FLAC is a format that can store audio with compression without loss (ie, it can be compressed without loss of sound quality).
\nAccording to the FLAC standard, formats of up to an integer of 32 bits \/ 655,350 Hz can be stored, but the encoder itself supports an integer up to 24 bits.
\nWAV is a former standard created in 1991 and is actually a type of container standard called Riff. Several waveform data can be saved in WAV, and 64-bit floating, etc. can also be stored.
\nIn other words, if both are in the range of 24 bits, you can save exactly the same waveform. This is a perfect combination at the waveform level.
\nIf you are using a music player called FOOBAR2000, you can use a supplement called Binary Share to confirm that there is no difference of 1 bit between WAV and FLAC.
\nFLAC has a CRC checking sum for each frame and the file has an MD5 hash to detect damage to the file. It is possible that WAV does not realize even if it is damaged, as confusing data.
\nBy the way, if you want to save 32 bits or more, you can use a lossless compression format named Wavpack.
\n“Decodify load gar”
\nThe logic of those who argue that there is a difference in sound quality between WAV and FLAC is generally “FLAC requires decoding to PCM for reproduction, and there is a need for decoding, so the CPU load is high and quality of sound is high. It gets worse. ”
\nThe first half of this logic is correct. FLAC is a compression without loss and must be decoded for reproduction. However, when discussing the second half, questions arise.
\nAlthough FLAC is a compressed format first, the load during decoding is one of the lowest among all lossless compressed audio formats.
\nThe codecperformancecomparison site has decoding speeds by format. Among them, FLAC has a speed that far exceeds Monkey’s Audio, Wavpack, AAC and MP3. The decoding speed is almost the same regardless of whether the compression rate is 5 or 8.
\nIt is said that FLAC can decode only by arithmetic of integers (source), and is quite fast. Although WAV does not require decoding, it is nothing more than using the CPU, since reading the file, saves it in memory and the CPU processes the driver and sends the data to the sound card.
\nIn addition, WAV is not compressed, so there is the problem that the file size increases. This means that reading files from storage is clearly more frequent than FLAC. This is nothing more than using more bandwidth of your computer bus. How do you explain the electromagnetic noise that generate these storage and buses? Has anyone verified it?
\nImportance of the ABX test (double blind test)
\nI think most people in audio files (OTA, audiophiles) minimize science and think about things with several conjectures, relying solely on their own senses.
\n“Clear sound” “Soft and elegant acutes” “Natural Low”
\nThese words sound good, but they have to do with falsifier and I can not even start a discussion.
\nI want to compare A and B. At that time, the “ABX test” or double-blind test (double blind method) is useful.
\nLet’s say you have A and B. I want to compare A and B. It can be WAV and FLAC, or an oxygen-free copper cable and Amazon Basics cable. It may be the sound reproduced by the electricity generated by hydroelectric power and the sound reproduced by sunlight, or the difference between iPod and Walkman, cassette tape and disk.
\nIf you really want to make a comparison, hide A and B and ask. There is another collaborator in this. When you start the test, the collaborators first prepare the equipment for A and B, and they clearly touch each sound as A or B. Next, the collaborator hides A or B and reproduces the sound. It can be in the order of Aabbaba or Ababab. Anyway, he plays randomly. And the collaborator asks: “What was the random sound I presented for the first time, A or B?” He can request and re-listen to the sounds A and B at any time during the test.<\/p>\n","protected":false},"excerpt":{"rendered":"