
Comparison of compressed files
The playback environment has gotten a bit matomo, so I listened to it again and compared it.
![]()
The file is the same as the one used in the previous item. Sound from Creative’s SB Audigy LS sound card is produced by Sony LBT-V610 (commonly known as Liberty – a model circa 1990 so this is a 17 year old player, and the speaker is a rear bass reflex type 3-way 2 -unit with a volume of about 20 liters) and plays on the speaker. Like last time, I used Lilith version 0.991 for playback. As a pseudo-blind test, listen to each of the two sound sources to be compared, then switch to loop play mode, press the next song button repeatedly with your eyes closed, and don’t know which one will play first. , and then I also tried playing it and trying to guess which file was which.
For the Wave file (compressed to ape format and distributed) and the MP3 file, the previously used file was used as is. Lilith was used to encode Ogg Vorbis (v1.1.0) (faster and with better sound quality than the audio encoder). ACC used iTunes. ogg and ACC only change the bit rate from the default setting (ACC’s “Use VBR Encoder” and “Optimize for Audio Files” options are not checked). Lilith was used to play the files, and the ACC file and its comparison files were played using the VLC player.
First of all, it’s compressed vs. uncompressed, but the air changes slightly even in MP3’s 256 kbbs (lame) mode. The compressed file has a subtle sense of purpose, or a slightly unnatural “missing sound sensation”. I did the pseudo blind test 10 times and made a mistake once.
Compared to MP3s, lame’s 128kbps exceeded 128kbps in the afternoon. I have not done a pseudo blind test because it is so different that it is incomparable. At 198 kbps and 256 bps of lame, 198 kbps is a little less lively and a bit boring. I did the pseudo blind test 10 times and made a mistake once. The difference with Wave is also widening.
The comparison between ogg (setting 0.6: 220 kbps nominal but equivalent to 192 kbps in terms of file size) and MP3 (poor: 192 kbps) is quite tricky. There were too many mistakes to pass the pseudo blind test (I did it 5 times and lost motivation when I made 3 mistakes). 192kbbs from AAC didn’t do a pseudo-blind test because Lilith couldn’t read the m4a file, but it feels brighter than ogg or MP3 (even though it’s a pseudo, I can’t say anything because I haven’t made a blind though when I compared it with the Wave raw, you may have intentionally overproduced the shine).
Ogg (0.4 setting: 172 kbps nominal, but equivalent to 128 kbps in terms of file size) and MP3 (poor: 128 kbps) are quite different, and the pseudo-blind test was performed 5 times and 1 error. MP3s have more weight in the sound than hardware (theoretically MP3s can cut high frequencies, so that may be the effect). With AAC (128kbps) and MP3 (poor: 128kbps), I haven’t done a pseudo-blind test, but AAC clearly sounds good.
After many tests, I noticed that the sound recorded from SoundFont as a test sample was not very appropriate (because it was pre-processed) (at the moment, the spectrum analyzer “sounds to the limit”. Confirmed that “is”). If I have the opportunity next time, I would like to compare by instrument, studio recording / live recording, and music genre. However, I was able to judge “whether the sound changes or not” for the moment, so I would like to get it right for the moment (it was quite difficult both physically and mentally).
For the moment, I will write what I have concluded in the previous Chombo. AAC is excellent at 128 kbps. This is probably the result of combining the AAC concept with the low bit rate limits. 192kbps is a close battle, but it seems that AAC selects “high sound quality as a compressed file” instead of fidelity to MP3 and ogg, which are slightly bright because they are true to the original sound (because it has a waveform after the compression)., It should have been like this before compression. ”
The proper use of 160/192 / 256kbps is quite delicate, and it is persuasive to say that “the size does not change that much, so there is a lot of space at 256kbps” and “the difference in sound quality is rarely a problem”. So the size. At 160 kbps, which is small, there is a reason. There is a problem with the use, but the ratio of taste and mood will be large. By the way, 160: 192: 256 = 1: 1.2: 1.6 = 0.833 ..: 1: 1.33 .. = 0.625: 0.75: 1.







