Re: [討論] 原來CD 轉檔 音質真的不一樣

※ 引述《vincentsoon (羅漢果)》之銘言：
: 各位發燒友大家好
: 承前幾篇CD串流那篇文
: 本著好奇心聽看看4個檔案是否不同
: 結果還真的差蠻多的
: 真是太神奇了@@
: 測試器材： fiio x3 + Corda classic + B&o A8
: 這首歌可以著重聽以下我覺得有差異部份：
: 背景一直敲的三角鐵聲定位
: 鋼琴殘響
: 0:36 呢喃聲
: 1:49 右耳細節
: 2:46 呢喃
: 3:33 樂器分離度
: ————-
: 有興趣朋友可以聽聽，真的不一樣☺
: 比較明顯好壞的可分成這兩組
: CD、AB
: 以下結錄那篇文章部份&檔案連結：
: https://reurl.cc/Mda88L
: 使用D100轉檔有比較好聽嗎?
: 很多人很關心究竟用好的Ripper來轉檔，是否聲音會比較好？畢竟只是將數位訊號抓取出來
: 方法1: iTune + 普通CD-ROM
: 方法2: iTune + Melco D100
: 方法3: dBpoweramp + Melco D100
: 方法4: iTune + Melco D100 + LPS線性電源
: 下載點：
: A: http://www.my-hiend.com/leoyeh2/2018b/01%20I%20Love%20Being%20Here%20With%20Y
: B: http://www.my-hiend.com/leoyeh2/2018b/01%20I%20Love%20Being%20Here%20With%20Y
: C:
: http://www.my-hiend.com/leoyeh2/2018b/01%20I%20Love%20Being%20Here%20With%20You_
: D: http://www.my-hiend.com/leoyeh2/2018b/01%20I%20Love%20Being%20Here%20With%20Y
: -----
: Sent from JPTT on my iPhone

Bit-Compare File Ａ＆Ｂ、Ａ＆Ｃ、Ａ＆Ｄ、Ｂ＆Ｄ

※

Differences found in compared tracks; the tracks became identical after
applying offset and truncating first/last samples.

Comparing:
"G:\01 I Love Being Here With You_A.wav"
"G:\01 I Love Being Here With You_B.wav"
Compared 13812120 samples.
Differences found: 27526587 values, 0:01.034104 - 5:13.199977, peak: 1.999969(+6.02 dBTP) at 4:11.182721, 2ch
Channel difference peaks: 1.785767 (+5.04 dBTP) 1.999969 (+6.02 dBTP)
File #1 peaks: 1.000000 (0.00 dBTP) 1.000000 (0.00 dBTP)
File #2 peaks: 1.000000 (0.00 dBTP) 1.000000 (0.00 dBTP)
Detected offset as 661 samples.

Comparing again with corrected offset...
Compared 13811459 samples, with offset of 661 discarding last/first samples
from total of 13812120, discarded samples were silent in file #2.
No differences in decoded data found within the compared range.
Channel peaks: 1.000000 (0.00 dBTP) 1.000000 (0.00 dBTP)

Total duration processed: 5:13.200
Time elapsed: 0:02.123
147.55x realtime

※

Differences found in compared tracks; the tracks became identical after
applying offset and truncating first/last samples.

Comparing:
"G:\01 I Love Being Here With You_A.wav"
"G:\01 I Love Being Here With You_C.wav"
Compared 13812120 samples.
Differences found: 27505161 values, 0:01.033968 - 5:13.199977, peak: 1.982147(+5.94 dBTP) at 5:05.843764, 2ch
Channel difference peaks: 1.722321 (+4.72 dBTP) 1.982147 (+5.94 dBTP)
File #1 peaks: 1.000000 (0.00 dBTP) 1.000000 (0.00 dBTP)
File #2 peaks: 1.000000 (0.00 dBTP) 1.000000 (0.00 dBTP)
Detected offset as -6 samples.

Comparing again with corrected offset...
Compared 13812114 samples, with offset of -6 discarding last/first samples
from total of 13812120, discarded samples were silent in file #1.
No differences in decoded data found within the compared range.
Channel peaks: 1.000000 (0.00 dBTP) 1.000000 (0.00 dBTP)

Total duration processed: 5:13.200
Time elapsed: 0:02.117
147.93x realtime

※

Differences found in compared tracks; the tracks became identical after
applying offset and truncating first/last samples.

Comparing:
"G:\01 I Love Being Here With You_A.wav"
"G:\01 I Love Being Here With You_D.wav"
Compared 13812120 samples.
Differences found: 27526587 values, 0:01.034104 - 5:13.199977, peak: 1.999969(+6.02 dBTP) at 4:11.182721, 2ch
Channel difference peaks: 1.785767 (+5.04 dBTP) 1.999969 (+6.02 dBTP)
File #1 peaks: 1.000000 (0.00 dBTP) 1.000000 (0.00 dBTP)
File #2 peaks: 1.000000 (0.00 dBTP) 1.000000 (0.00 dBTP)
Detected offset as 661 samples.

Comparing again with corrected offset...
Compared 13811459 samples, with offset of 661 discarding last/first samples
from total of 13812120, discarded samples were silent in file #2.
No differences in decoded data found within the compared range.
Channel peaks: 1.000000 (0.00 dBTP) 1.000000 (0.00 dBTP)

Total duration processed: 5:13.200
Time elapsed: 0:02.099
149.23x realtime

※

All tracks decoded fine, no differences found.

Comparing:
"G:\01 I Love Being Here With You_B.wav"
"G:\01 I Love Being Here With You_D.wav"
Compared 13812120 samples.
No differences in decoded data found.
Channel peaks: 1.000000 (0.00 dBTP) 1.000000 (0.00 dBTP)

Total duration processed: 5:13.200
Time elapsed: 0:00.441
710.88x realtime

※

努力會讓你美夢成真

但太努力有時會讓人產生幻覺...

做比較請愛用 ABX and 進行一定次數以上
然後不要太相信自己XD

檔案Ｂ＆Ｄ完全相同沒有任何差異，很正常因為是同一台 CD&軟體 RIP 出來的
原始當事人聽起來有差的話是因為使用的電源不同所以是類比域的差異
用同一套設備兩個檔聽起來有差就...可能是宇宙射線的影響？

其它檔案之間只是因為使用的 CD Player 及軟體不同所以數據有偏移（音軌間的空白
分配問題），但檔案中的音頻數據也都是相同的沒有差異

--
人間五十年、化天のうちを比ぶれば、夢幻の如くなり
^,,,^ 一度生を享け、滅せぬもののあるべきか
(ミ‵ω′)\m/ NOBUMETAL DEATH!!(乂'ω')

--

XD

人類是有感情的、很容易受影響的物種，安慰劑會有用就是證明 所以沒盲測是不會有公正的結果 但就算盲測也會受心情起伏影響

最下面的那篇文章很好的說明了"盲測"是如何被影響的實際案例 那天大家都很開心www

所以必須做一定次數的盲測 如果測個十幾次有九成、幾乎每次都能準確的分辨出這是Ａ、Ｂ、Ｃ或Ｄ檔 那差異就是真的而不是幻覺

老實說每次的播放應該都是獨一無二的、但人類真的能分辨每次的不同否？ 以市電來說 60 Hz、會產生 60 Hz＋120 Hz、180 Hz 等的諧波加入訊號中 每次的播放這些諧波都會以不同的時間與音樂波形結合 換了更好的供電能聽出差異很合理 但每次 Replay Track 的開頭對到市電的波峰還是波谷能明辨否？ 人類在無音室中能清楚的聽到自己的心跳與呼吸聲，有些人還能聽到血液流動 所以很明確這些聲音的大小都在人類的聽力閥值之上 之後每次按下 Play 前是否要把脈、憋氣務求心跳、呼吸與播放同步 以創造一致的聆聽體驗？要更精確的話可能還要測個血壓、血氧嗎 所以 Easy 一點、不要太過講究這些有的沒有才比較能享受音樂本身不是嗎