I wonder is this the right place to post this. This matter of fact came to my attention recently. I downloaded the flac version of 1830m album from here and it failed in Audiochecker by Dester. Then I downloaded another alac version of 1830m from here as well and it also failed in Audiochecker. Then I was wondering maybe this album quality is just bad. Then, I came across the following comment by "Zaku Abumi" at his UZA flac upload:
"I am using the program you are mentioning. The mixing engineer cut off the frequency at 21.5 khz as in 1830M. If you check the graphic of the audio files, the peak level is constant. That does not happen when it comes to transcodings."
Zaku Abumi seem to be using another different program named LoselessAudioChecker. So I went and downloaded that program and both of the previous 1830m album passed the test. So to verify the fact that the mixing engineer did something different since 1830m, I went and rip my recent AKB cd (So Long and Sayonara Crawl) that I owned using EAC. Both rips really do failed in Audiochecker but passed in LoselessAudioChecker. So this verified Zaku claim that the mixing engineer really did something different since 1830m.
Then I tried to test the reliability of this new program named LoselessAudioChecker. I convert my rip that passed both Audiochecker and LoselessAudioChecker to lossy and then back to loseless with few methods. And at the second tried, the reconvert loseless file failed in Audiochecker but passed in LoselessAudioChecker. Then I already lost confident in LoselessAudioChecker reliability.
So to me, it seem that Audiochecker is more accurate than LoselessAudioChecker. Now here come my question and doubt... Why did the mixing engineer of AKB change the method??? Does this new method affect audio quality??? I will try the ABX test between the 1830m version and single version when I am free to see whether this affect the audio quality. I wonder anyone tried the ABX test already? Do you heard any differences?
Finally am free to do some ABX test. So I pick out the tracks from 1830m and their single for these 2 songs: Sakura no Ki ni Narou & Kaze wa Fuiteiru.
First part of the ABX test, instrument test. I focus on the instrument part of the song and did 3 ABX tests for each song. Full marks on both songs. The single version have more power in their instrument especially bass. The bass in the album version lack a little bit of power and sound a little bit quieter or soft.
Second part of the ABX test, vocal test. This time I focus on the vocal part of the song and did 3 ABX tests for each song. 2/3 correct for the Kaze wa Fuiteiru and 1/3 correct for the Sakura no Ki ni Narou. Wasn't really that impressive -.-''' Maybe those correct are just by luck at all for the vocal test...
So it seem that I am more sensitive to bass and not really that good at vocal T_T However, I do heard significant differences between the album version and single version. Though it is just the bass for me, but it still prove there are audible differences.