電腦數碼音樂重播專欄 (2) - 電腦音樂系統 : PhotoBlog

PhotoBlog - 影音

更多影音消息

本文作者：Kent
發佈日期：2011-04-30 02:13:05

在Facebook 專頁按「讚好」，免費影音資訊自動送上

CAS (Computer As Source) 電腦音樂系統 (2) :

上期有介紹新朝代 CAS 電腦音樂系統特集, 大多介紹的都是桌面型小型 CAS 系統。其實 CAS 的主要概念便是使用電腦作為訊源來重播音樂, 而並非只是代表小型桌面系統。

傳統的音響系統:

CD 轉盤 Transport -> D/A 解碼 -> 前級音量控制 -> 後級電流放大 -> 喇叭 (圖一)

CAS 電腦音樂系統只是改變了由 CD 轉盤轉變為電腦:

電腦 -> D/A 解碼 -> 前級音量控制 -> 後級電流放大 -> 喇叭 (圖二)

如上期介紹的新興 CAS 玩法, 是圍繞桌面上應用的小型系統。這些小型系統使用了簡單, 直接的方式來製造出高性能價格比的高水平音效。主動式 (Active) 喇叭除了需要空間少 (內置後級放大器) 之外, 好的主動式設計具備了較被動式 (Passive) 準確的電子分頻線路, 二段獨立放大後級等不少優點。而解碼器已經擁有音量調節控制功能來代替傳統前級。

電腦 -> 音量控制 D/A 解碼 -> 主動式喇叭 (圖三)

以上三種方法, 用簡單的說法, 如果每個部份的質素都一樣, 那越短線路, 傳真度便越高。這與 CAS 的概念有所配合。傳統音響發燒友對 CAS 有種又愛又恨之感覺, CAS 概念上與傳統音響也有十分矛盾的地方。但有二點是不能否定。

第一: 上期已提及 CD 永遠只可以提供 16bit/44.1kHz 的訊號。過去十多年製作人已成熟地使用上 24bit 高解像錄音 (是 CD 的 256 倍) 方法, 有些更應用上高取樣 (Hi-Sampling) 形式收錄。在重播解碼技術也同樣支援。對音響上有要求的發燒友, 應該會如我一樣被高解像音響所迷住。那豐富低音量訊式包圍一雙喇叭的音場, 定位上的準確性, 音色上的模擬 (analogue) 感, 使用同一樣解碼, CD 重播便顯得平坦而乏味。

第二: 現代錄音全都是電腦化, 過往的 ”母帶”, 已經在 15-20 年前被電腦檔案 file 所取代。無論是製作 CD, DVD, SACD 或 Blu-ray 這些媒體, 其實都是由一個正式的 ”Master 母帶檔案” - 改動而成。

電腦母帶檔案是全個製作的 - ”種” (圖四)

DVD 要加 Dolby Digital, DTS 等的壓縮, 之後 authorizing。差一點的 CD 便把 16bit 以上訊號切去。好一點的 CD 要善用聴覺較敏感的頻率來突出 16bit - 24bit 訊號的效果。賣 iTunes Store 要作 AAC / MP3 大壓縮。

製作時期, 錄音室內是由電腦重播作鑑聽。 Studio 內的電腦被稱為 DAW (Digital Audio Workstation). 重播音樂跟 CAS 一樣是以電腦作訊源。在我作為錄音製作人來說, 由這個原檔案作重播, 必定要比之後做成 CD 媒體, 再由 CD 轉盤重讀碟上的數碼資料來得直接無誤。 CD 的 BLER (Block Error Rate), Pit-Jitter P1/P2 影響完全消失得無影無蹤。電腦作為訊源播放音樂檔案根治了 CD 重播的缺點, 可惜也有自己要解決電腦上的問題, 下次再談。

筆者自我介紹

Kent Poon 潘建章於 1993 在北美開始學習錄音技術。1997 加入世界最大音響學會 AES (Audio Engineering Society), 是其中最年輕會員之一。由 2003 回港擔任威達公司技術顧問至 2008 年。Kent 於早期”音響之路”、”發燒音響”、 ”HiFi音響” 撰寫數碼技術文章而為本地音響愛好者所認識。 2010年在香港成立 Weiss 亞洲區辦事處。執筆之時正聽着 Mari Kodama <Beethoven Piano Sonata>, 發現了法國 Devalet d-premier 是當世數碼重播之頂峰。

Last modified: 2011-08-08 02:46:08

發表您對 < 電腦數碼音樂重播專欄 (2) - 電腦音樂系統 > 的意見

26 Comments

#1 - ameba :

2011-05-03 11:12:51

Thanks Kent for the good article!

It would be perfect if you could use the term '質素' instead of '質數'.

#2 - DWS :

2011-05-03 11:35:17

Thanks so much ameba, I am learning as I said.

#3 - lesterTam :

2011-05-16 10:22:40

Hi Kent,

I'm a newbie to hifi world, after reading your article about mac+iTune as source in designwsound.com (it suggest that
the output of this combination is bit transparent), I have some questions about DAC:

1.Can i understand it as this way: the process for convert wav/CD to digi signal is very easy to achieve "perfect"
so a LowEnd CD player is more thean enough if you use SPFID/Coaxial (so no D/A convet) to your amp and let your amp to do the decode part,
for HiEnd player/source we should aways use RCA output to connect to the amp (assume the build in DAC is better then the DAC in the AMP_)

2.Assuming (1) is correct, that means it shoud have no difference for below 3 settings, right?
a. BD Player --HDMI--＞ AV Amp
b. Mac+itune(rip CD as AIFF) --SPFID--＞ AV amp
c. Mac as CD player --SPFID--＞ AV amp

3.I know that jitter can be a issue for digital signal, is it noticeable? or just very minor for ppl without golden ear just like me?

#4 - batmanames04 :

2011-05-16 11:44:30

"1.Can i understand it as this way: the process for convert wav/CD to digi signal is very easy to achieve "perfect"
so a LowEnd CD player is more thean enough if you use SPFID/Coaxial (so no D/A convet) to your amp and let your amp to do the decode part,
for HiEnd player/source we should aways use RCA output to connect to the amp (assume the build in DAC is better then the DAC in the AMP_)

2.Assuming (1) is correct, that means it shoud have no difference for below 3 settings, right?
a. BD Player --HDMI--＞ AV Amp
b. Mac+itune(rip CD as AIFF) --SPFID--＞ AV amp
c. Mac as CD player --SPFID--＞ AV amp

"

bit transparent doesnt mean same jitter level to DAC, sound should be different with a , b and c

#5 - lesterTam :

2011-05-16 12:06:39

Thanks batmanames04 ,

so if i have a good DAC to reduce in the receiving end, ie source -＞ good DAC--＞ AV AMP

the difference between A,B,C should be very small right?

#6 - batmanames04 :

2011-05-16 12:10:24

yes, lester Tam hing, thats right.

provide you have a good dejittering DAC, b and c come close and better than a

#7 - lesterTam :

2011-05-16 14:22:00

thx again,
So jitter really does matter in the final output result..
one more question, i see some device combined the transporter & DAC in one box(like logitech transporter, Squeezebox...) in that case there are no jitter problem anymore right? and it should achieve good sound quality(compare to CD) assume that the DAC inside is good enough. right?

#8 - batmanames04 :

2011-05-16 14:41:20

Hey, LesterTam you know jitter is the biggest enemy to digital audio

regarding transporter it is streaming audio device with build in DAC, sound just so so but from measurement, still a jittery device

transporter jitter measurement after transporter internal dac

#9 - batmanames04 :

2011-05-16 14:45:33

for your comparison weiss minerval DAC

#10 - batmanames04 :

2011-05-16 14:51:12

dcs scarlatti dac

#11 - DWS :

2011-05-17 10:03:06

hello lesterTam, thanks for your comments.

＞ 1. convert wav/CD to digi signal is very easy to achieve "perfect".

It is easier to achieve "perfect" with today computers than 10 years ago.
However the new DVD/CD-Rom has inferior quality is true. Today best CD/DVD/Blu-Ray roms are still vintage Plextor Premium2, Pioneer Bluray Model6 with pure read2 technologies. Computer ripping and matching ability is really high.

＞ LowEnd CD player is more thean enough if you use SPFID/Coaxial (so no D/A convet) to your amp and let your amp to do the decode part,

This is more likely a videophile approach. Audiophile usually does not use has DA built into amp. (I can think of the lovely Devialet)

＞ HiEnd player/source we should aways use RCA output to connect to the amp (assume the build in DAC is better then the DAC in the AMP_)

Yes, high end CD player should be better quality than the amp. And then you will find high end audio stereo amplifier is better quality.........that's how the journey is started.

#12 - DWS :

2011-05-17 10:07:14

＞ 2.Assuming (1) is correct, that means it shoud have no difference for below 3 settings, right?
a. BD Player --HDMI--＞ AV Amp
b. Mac+itune(rip CD as AIFF) --SPFID--＞ AV amp
c. Mac as CD player --SPFID--＞ AV amp

If a,b,c provide the exact same data, the differences will not be locate in the data content but the timing accuracy.

3.I know that jitter can be a issue for digital signal, is it noticeable? or just very minor for ppl without golden ear just like me?

Don't worry about it until you have discovered the "changes". And I have to say it should be difficult to determine if it is "different" or "better/worse".

#13 - DWS :

2011-05-17 10:09:56

＞ So jitter really does matter in the final output result..

Yes it does matter, but how much is different for different setup and listener.

＞ transporter & DAC in one box(like logitech transporter, Squeezebox...) in that case there are no jitter problem anymore right?

The transporter still needs to grab music from the computer and decode them with correct data content and timing accuracy.

#14 - DWS :

2011-05-17 10:20:17

Batman hing transporter measurement with internal DAC. This measurement only reflects using its internal DAC. If you are using transporter (let's say with Weiss Minerva), then the Weiss Minerva can remove/reject those jitter. But can Weiss Minerva performs as good when using transporter as source, that will be another measurement set.

Go with your mind and heart about sound and music, focus on listening, theory is only right if applies correctly.

An example batman hing posted dCS Scarlatti and Weiss Minerva DACs. If you focus on the jitter (ps value) dCS has 37psec and Weiss has 64psec. Both are extremely low value which should bother anyway, but the fact is dCS has lower ps value.

Moreover, the nosiefloor of dCS is stick with -140 while Weiss has higher noisefloor too.

However one major point is the first peak value of Weiss jitter is around 2.5kHz while dCS is around 300Hz. In some people view, this is more important than 2 points in above.

So again, go with your ears.

#15 - batmanames04 :

2011-05-17 14:59:48

DWS c hing

yes, i believe using transporter spdif feeding weiss minerval will get much better jittering meassurement result. as you know there is good degittering algorithm inside weiss minerval comparing to other dac selling in the market

for comparing weiss minerval vs dcs to jitter

in engineering design
1 we count for
1`number of spikes happening
2 noise floor (lower the noise floor, you can hear more bits)

3 also we measure the spreading of center spectrum at the base

combine 1 , 2 and 3 or simply we can say the integration of phase noise curve of the spectrum

so that doesnt mean msb the best

the goal of engineer is to design a dac with zero jitter output , that means a very sharp long spike at the center frequency

#16 - DWS :

2011-05-17 16:52:20

Sorry that I can only provide short answer.

＞ 1 we count for number of spikes happening

but if the spikes is as lower than -135dB, then it does not matter because -135dB is the noisefloor or almost all converters.

＞ 2 noise floor (lower the noise floor, you can hear more bits)

This is only jitter measurement not SNR measurement. The reference tone is not reference 1kHz and -60dBFS that from the stardard. Therefore in jitter measurements which you posted, the lower noisefloor does not mean you can hear more bits.

Many psychoacoustics reasons are lies in this area. I agree lower noisefloor in SNR measurement, higher dynamic range, less THD+N are results of more truthful to the source.

But I also would like to comment that psychoacoustics reasons are more or less important because it affects how we listen. And there are others reasons too such as driving range, output impedance, output levels, and other analogue areas etc.

#17 - batmanames04 :

2011-05-17 18:01:14

＞ but if the spikes is as lower than -135dB, then it does not matter because -135dB is the noisefloor or almost all converters.

I am wondering why MSB getting -160dB noise floor and curious they measure msb with AP

#18 - siulee400 :

2011-05-17 19:01:29

Hi DWS,

"Today best CD/DVD/Blu-Ray roms are still vintage Plextor Premium2, Pioneer Bluray Model6 with pure read2 technologies."

For Pioneer Bluray Model 6, do you mean the 6 series of:

1) BDR-206MBK
http://pioneer.jp/bdd/products/bdr_206m/index.html

2) BDR S06
http://pioneer.jp/bdd/products/bdr_s06j/

As I heard from some c-hing that any CD-Rom's ripping quality (pure CD-Rom with no writer feature)will be better than those of DVD Rom, DVD Writer nowadays.
Does the pionneer Series 6 Bluray writer can be better than the CD-Rom for ripping CD purpose only? Why it is good?

#19 - siulee400 :

2011-05-17 21:20:53

And which one is better?

1) BDR S05 (Made in Japan)
2) BDR S06 (Made in Japan)
3) BDR S06 (Made in China)
3) BDR 206MBK (Made in Japan)

#20 - DWS :

2011-05-21 06:39:09

Except the one made in china, the rest are better.

#21 - DWS :

2011-05-21 14:37:17

＞ I am wondering why MSB getting -160dB noise floor and curious they measure msb with AP

batman hing. the photo you posted and the from their manuals are measured by Spectraab, not AP.

#22 - batmanames04 :

2011-05-21 15:29:29

agree

#23 - batmanames04 :

2011-05-22 09:40:26

msb technlogy power dac gold4