Oppo 播 SACD Output 的問題 ?

chhanthony

keekee 發表於 2013-11-12 13:04
我諗安仔香港都要問翻日本設計部呀
好有可能系部AMP自己轉換DSD為PCM了，但系屏幕依然比你睇系 ...

現時巿面上9*9嘅DAC Chips都係先將DSD轉做PCM再做解碼。

chhanthony

keekee 發表於 2013-11-14 09:26
原來系HDMI既規格裡面有說明，PCM和DSD的sample rate顯示計算是PCM睇到的就系等於它的sample rate，但DSD ...

keekee兄

pwong888

其實如果用105play SACD, 用rca出amp應該set SACD output DSD定pcm會好声D?

samnet1

keekee 發表於 2013-11-14 09:26
原來系HDMI既規格裡面有說明，PCM和DSD的sample rate顯示計算是PCM睇到的就系等於它的sample rate，但DSD ...

謝謝 KeeKee C-Hing 回覆

Additional 資料:

"For One Bit Audio streams, the value indicated by the SF bits shall equal the ACR fS value. For Super Audio CD, the SF bits are typically set to 0, 1, 0, to indicate a Sample Frequency of 2.8224MSamples/s (i.e. 64*44.1kHz)."

This is due to 8-channels of audio being available through HDMI at 8 bits per channel at 44.1KHz. The end bandwidth is 2.8224Mbits  /second, which equates to 2.8224M 1-bit audio."

samnet1

louis523 發表於 2013-11-14 02:17
路過一問，想知DSD係代表咩？thx

http://zh.wikipedia.org/wiki/Super_Audio_CD

See the Red Text :

SACD是Sony與Philips合力研發的音樂碟片規格，是繼CD的發明之後，成功超越CD錄音品質的新產品。

音樂CD（CDDA）採取44.1KHz的取樣頻率，因此在接近人耳聽覺極限的高頻訊號中（接近20KHz）只能取樣出三點；要用這三點來恢復正確的類比訊號有困難性，被認為會造成相位誤差。

SACD的取樣頻率高達2822.4kHz，是一般CD 44.1KHz取樣的64倍，而且SACD頻率範圍更是高達100KHz以上；也因此使得SACD改善了原來音樂CD音質給人冷硬的刻板印象，而以更細膩、更多細節、更柔軟的聲音呈現。

SACD的錄音方式是用Direct Stream Digital（DSD，即直流數位技術）方式錄音，摒除傳統的PCM錄音方式，將所有訊號以每秒280萬次直接把類比音樂訊號波形轉變為數位訊號，也就是所謂的『直接位元流數位』，因此取樣波形非常接近原來的類比波形。

另外，SACD省去位元轉換程式，降低了數位濾波而可能產生的失真與雜訊。還有一個特點就是SACD也可以容納多聲道以及影像，由於SACD自身的定位以及1bit量化DSD直接資料流程在技術方面的簡潔和優勢（多數DAC是處理DSD數位訊號及類比訊號的互相轉換，如果要輸出或輸入PCM格式，則必須加上DSD及PCM訊號的轉換機制，這個機制需要相當的計算量），使得大多數的資深音響發燒友經過親耳聆聽後，主觀感覺都認為SACD在音質上勝過音樂CD。

samnet1

chhanthony 發表於 2013-11-14 10:00
現時巿面上9*9嘅DAC Chips都係先將DSD轉做PCM再做解碼。

現時巿面上9*9嘅DAC Chips都係先將DSD轉做PCM再做解碼。

==> 為什麼市面上的 DAC 這樣 Design??   芯片價錢太貴 ??  

keekee

多謝各位ching加分
我都系搵鬼佬網討論區學下野，鬼佬有時d野又真系幾update既。

chhanthony

samnet1 發表於 2013-11-14 10:43
現時巿面上9*9嘅DAC Chips都係先將DSD轉做PCM再做解碼。

==> 為什麼市面上的 DAC 這樣 Design??   芯片 ...

因為DSD本身係行1bit但1bit係做唔到其他運算，even ESS9018都係行5-6bit，
而pure 1bit嘅DAC亦唔可以用作mutli bit嘅DAC,當中好多比較學術，有興趣睇下下面啲高手過招!


from R33

ESS9018 is a DAC accept 32bit/384Khz data natively. and can accept DSD64 and DSD128x dsd data natively.

the only DAC which is not using off the shelf IC may be only MSB, dCS and EMMLabs. MSB and DCS using high precision resistors for implementation, but one is PCM base and one is 5bit dsd one.

For dcs, when we play 16/44.1 , it still need to convert to DSD for processing , to some extends , it is not a pcm native decoder.

for EMMLabs, it is a different implementation with self developed software for dsd processing and for pcm data , also need to convert before processing.

there are other DAC using BB PCM1795 chipset (which is a 32bit/192khz) , it is also a sigma delta chipset can accept data up to 192KHz. for dsd data, it also accept dsd natively but also convert to pcm internally before processing.

for data over 192KHz or dsd64x, the dac need to downsample to 192khz before doing D to A conversion by PCM1795.

both pcm1795 and ess9018 are sigma delta based chipset.

For DAC implementation, there are two implementations for pcm , one is sigma delta and one is R2R. For circuit design, sigma delta need i/v and LPF at the output pint of the DAC chipset, and R2R does not .

From Hiandy

The basic structure of dCS is a 5-bit processor mechanism.  It's not the 16/24/32 bits of PCM nor the 1-bit of DSD.
The exact internal structure/mechanism of ESS9018 is a commercial secret but I heard that it's actually a 5-bit or 6-bit mechanism.
The R2R mechanism of MSB is also as fuzzy as Andreas mentioned in the interview.

When 1-bit DSD signal is fed to dCS and MSB, a software-based DSP will manipulate the data to a form that can be decoded by their own dac mechanisms.
The same happens for ESS9018, the only difference is that its DSP section is located inside the 9018 chip so we don't see it.

Andreas' explanation is similar to what Mr. Kent Poon told us during a DSD Seminar orgainized for XXXXXX brothers at his showroom in TST a few months ago :
" ... It is all a matter of definition: DSD, or Delta-Sigma Modulation, can be encoded with more than just 1 bit, and PCM can have a very high sample rate. When looking at the criteria of word length and sample rates only, the boundary between DSD and PCM can become fuzzy. Sometimes it is more useful to distinguish DSD from PCM in the frequency domain and look for the characteristic behavior in the higher frequency bands ... But in the end, shouldn’t we use our own ears to be the ultimate judge for what sounds great and what sounds not so great? ... "

In summary, the sonic characteristics of PCM format & DSD format have some real difference.
According to Andreas Koch (& me too) , DSD is the preferred format.
Whether the dac chip can handle DSD natively is not the most important factor affecting the sonic performance.
Both the dCS & MSB can give fabulous performance for DSD file playback.
IMHO it's more dependent on the quality of implementation/execution.

反感啲人除意用個'fake'字貶低人家嘅作法而凸顯自己作為作法高人一等嘅marketing手法，例如真假平衡之爭就此而來。。
如果解碼內部架構同輸入訊號格式不一樣就說成「假」，咁市面上九成九PCM 解碼都係5至6bits delta sigma架構，Delta sigma解碼無論用作PCM或DSD喜非又「假PCM」、又「假DSD」解碼！ 用作錄音的又全是「假PCM」嘅delta sigma ADC！

Native DSD解碼（如emmlabs，playback）對應輸入pcm要作convert 成1bits DSD,好唔好事乎implementation.  EMM要第二代產品先完善PCM-DSD嘅firmware( 當時由Andreas Koch開發), 其他新手唔知要幾耐？有幸感受過未完善前嘅Firmware，播CD明顯不及SACD之餘，比其他純播CD產品亦有不及。

追求native解碼方式我覺得沒有必要。

chhanthony

samnet1 發表於 2013-11-14 10:39
http://zh.wikipedia.org/wiki/Super_Audio_CD

See the Red Text :

以過往經驗話我知，唔需要太過追求是否DSD，因為現時巿場上DSD同PCM嘅產品/錄音都係十分模糊，一切自己聽為實。

chhanthony

samnet1 發表於 2013-11-14 10:43
現時巿面上9*9嘅DAC Chips都係先將DSD轉做PCM再做解碼。

==> 為什麼市面上的 DAC 這樣 Design??   芯片 ...

真正的純DSD DAC，佢嘅缺點又有幾多人頂得順



http://www.lampizator.eu/newdac/lampizator/DSD_DAC.html

http://parttimeaudiophile.com/2013/09/29/dsd-lampizator/