#1 - stephen :

15 年來我都希望把錄音中最好的質數帶給音響愛好者。＜＜＜===

最喜歡這句...support

#3 - ff4300 :

Support and looking for more in-depth comments on PC vs Mac, Firewire vs USB, FLAC vs Apple Lossless...

#5 - skto :

support

#7 - DWS :

Thanks guys.

I always wish to collaboration with more parties. (I am doing so too).

Indeed it is quite difficult for me to write chinese articles these days, but I wish you enjoy them. These articles are published on AV Bi-weekly magazine. (This week is the first time I missed....sorry)

You will find clone of these articles on DWS and other areas too. As time goes by I wish to share with you on my audio journey, and there will be free audio samples, takes or other audio materials for download. This is what I count on web base media.

From my view, Devialet d-Premier is one of the most unique audio products for a VERY long time in hi-end market. I simply cannot find anything better in term of its feature and sound. (of course its sexy look too.) A lot of people may view it as very expansive product. I have opposite view that it's a bargain in hi-end sector. Its amplifier driving ability can compete with the best amplifiers out there, and you cannot separate its DAC and Amp ability. Devialet is not a DAC + AMP, it is merging 2 as one. I am a happy user.

I absolutely have no negative feeling of LP or analogue. Before anything is converted to digital, they are ALL analogue signal. Microphone, mic-preamp, line preamp, are all analogue circuits. LP is one analogue media, open reel is another. I have honor to listen Nat King Cole open reel safety master and Living Stereo masters, I can still remember strongly how do they sound. Capturing and transfer these materials to digital domain do not necessarily means inferior quality as well. We have very good quality A to D and D to A converters these days.

Modern production is not suffered by technologies but people who abuse it. I wish to share more in the future. Speaking of it, I am now writing this with following view. I have already worked over 30 hours, will have to get more things done before tomorrow 8am........guess I am enjoying it.

Next 2 days can share with you all Korea HiFi show. Thanks review33.com for such collaboration. Tomorrow night I will listen to Medea+ with Magico Q5, nice.

#9 - DWS :

.

#11 - DWS :

.

#13 - sugizo :

Who is she？

#15 - mcc_2001 :

Lossless CAS platform is definitely the furture of music format but there are a very big risk.....that's how to prevent the illegal copy of such music files. The industry should have a scheme to prevent a second copy such as the SCMS (The Serial Copy Management System) as DAT adopted before.

#17 - sjyw :

CAS = "Computer As Source" ?! Where did you learn this? @@"

I only know CAS as "Computer Audio Source"

OMG. So surprising an "audio professional" (although only self-proclaimed) making this mistake.

#19 - momei :

The term CAS.
I first learned about it in Kent Poon's web, HiFitrack and Design With Sound. It stands for computer as source.
Then I came across the term elsewhere that is supposed to mean Computer Audio System.
Then we have the web Computer Audiophile.
It is not in Wikipedia.
The "C" stands for computer, that is, using it or a varied form of it, for example, a storage system or a server to play digital tracks.

#21 - cpu8088 :

"""AS far as I know, this is mostly a Hong Kong/Asia term. Countries with English as second or third language tend to make up short forms for convenience. It is not as commonly used by the American/UK audio press for instance."""

totally agree. at first when i encountered this term i had no idea what it was.

traditionally computer is the cpu with other items put in one box. these other items like sound card, memory hard disks or storage media are peripherals. one can stores data from a remote storage device and that can be the source. so computer as source is misleading.

i can see the trend will go with pay as you play on demand. it is more like renting instead of owning the software. who still want to have piles of cds and keep your library showing locations/numbering.

there are many software which may help split big wav or flac files using splitter or cue. split files can be transfer to dvd discs as data and then rejoin together at later stage.

#23 - DWS :

System 2

#25 - DWS :

JBL with Mark Levinson

#27 - DWS :

Wilson with Meridian

#29 - DWS :

I don't invent the term. I didn't know what it means at the first time, nor understand why someone would like to make scene of it.

CAS to me has been "民安隊". I also heard Eric (Editor of AV Magazine) mentioned it. Our government has a web page: www.cas.gov.hk.

People I met (consumer audio world) speaks it as "CAS System". At the first time I thought they mean "CARS System". Singapore, Malaysia audiophiles all speaks this way. They do not pronounce as independent C. A. S. system but "CAS".

I agree cloud base music library will become the majority of mass culture too.

Computer Audio System
Computer As Source
Computer Audiophile System
Computer Audio Server

as you like.

#31 - batmanames04 :

hey, I mean cas comparing with cd transport playing 16/44.1 to dac


one more thing, our human ears cannot hear more than 20khz sound , why 24.96 (around 48khz) sound better than 16/44.1?

#33 - cpu8088 :

there are many theories about hearing beyond 20khz

here is a good read

http://www.earthworksaudio.com/wp-content/uploads/th_world_beyond_20khz.pdf

#35 - cpu8088 :

"""I mean cas comparing with cd transport playing 16/44.1 to dac"""

assume u have a cd with no scratch and u rip the data to a hard disk with perfect bit which should be the same as u download the same music data file from internet, this should be the same as when you play a perfect cd thru a cd transport.

the difference lies with scratched or imperfectly made cds.

when u rip a scratched cd to hard disk it all depends on your chosen software how to handle the errors and make corrections by guessing. this error correction can be made many times like 20 to 50, for example.

with cd transport it is a different matter. cd transport has limited time frame to make error corrections in may be say 3 to 5 times read and then the inbuilt software will make rougher guess work and transmit the info through to the digital analog converter. a good cd transport may handle error corrections better than crappy ones but no match to more reliably ripped data.

does this make sense?

#37 - batmanames04 :

thanks momei, i need sometime to digest it

#39 - hercules :

"A CD has a bit rate of 1,411 kbps and a file size of roughly 10 MB. In other words, the CD master only contains 1/12 of the data in the native or original master, meaning 11/12 of data have been filtered out! A CD in comparison with its 24/96 master, has 5/6 of the data filtered out. "

The data size doesn't directly reflect the amount of musical information, a 16/44 recording of piano have more musical information than a 24/192 recording of "silence", but the size is much smaller.

The data size only reflect the space occupied for storing sample, more room for more information, a store room can store a bicycle or tyres of the MPV but not a MPV, but a big garage can store many bicycle with one or two MPV at the same time.

#41 - cpu8088 :

in video u will view all black or all white using 576i format data the same as 1080p format data, irrespective there is upconvert of 576i to 1080p. however when there are motions and multi colour images it is a different story. most people can see true 1080p better than upscaled 578i to 1080p.

why is it so hard to hear similarly in audio with upsampling of 16/44 format to hi rez to approximate true hi rez?

#43 - batmanames04 :

My DIY DAC hear almost no different between 16 44.1 and 24/96 material if they are derived from the same source

#45 - cpu8088 :

1996 thats 15 years ago. there was no diamond beryllium or other tweeters with exotic materials. could those old tweeters output properly the frequencies above 20khz? how good was the room treatments for the tests? if it was one that absorbed a lot of high frequencies of course there was less chance to hear properly.

#47 - cpu8088 :

from wiki:

Theorems have two components, called the hypotheses and the conclusions. The proof of a mathematical theorem is a logical argument demonstrating that the conclusions are a necessary consequence of the hypotheses, in the sense that if the hypotheses are true then the conclusions must also be true, without any further assumptions.

The concept of a theorem is therefore fundamentally deductive, in contrast to the notion of a scientific theory, which is empirical.

#49 - batmanames04 :

try to audition a high end DAC, your ears can hardly distinguish 16 44.1 and 24 96 recordings

#51 - DYreal :

If the file is playback from computer then the sound quality maybe affected by the quality of the computer.

I think that it is better to record the same sound with original file in 16bit/44.1Khz and 24bit/96khz into a CD then play it by a CDP.

#53 - momei :

CPU8088,
Regarding beryllium tweeters: not intending to challenge you but merely to tell you the fact. In the 70's Yamaha already made beryllium tweeter for its NS1000 loudspeakers. They were available in Hong Kong and some audiophiles bought them.
The Yamaha tweeters are distinguishable from that of current Focal ones in that the Y ones are convex whereas the F ones are concave. I have read posts here about your undue favour towards Focal speakers. Could you tell us what process Focal uses to make the tweeters? Vapour deposition or what? Apparently the Focal tweeters are thinner than the Yamaha ones. I do not have information what process Yamaha used and I guess it might be heat formation, that is, putting the metal powder in a mould and heat it to the melting temperature in an oven. Talking about Focal Utopia, I suspect, based on what I deduced from listening, the tweeter has dispersion restrictions merely because of the design of the tweeter cabinet that has the width and depth the same as the woofer cabinet. Other designs such as the Karma Exquisite, the KEF Fat Lady all have the tweeter housing tapered to the smallest size to help dispersion. Or in the B&W and Tannoy Royal designs, place it on top.

#55 - batmanames04 :

but most 33 members know you like jm lab beryillium tweeter

#57 - batmanames04 :

2)our tweeters can sound different when modulated with signals 20 kHz.

????

#59 - momei :

Batmanames04,
It seems better for me to know a bit more about your DIY DAC and the tracks you used by which you found no difference than to ask you to go straight into Test 2.
Does your DAC support 24/192 or even DXD? Does your computer out port support 24/192 and what interface or connector you use to play those tracks? The port and interface could be bottle necks and is 24/192 shown on the monitor screen to confirm such resolution is being played? I am trying to see whether by listening to 24/192 you may be able to hear the difference. Also please list two to three tracks you used previously, the source and the genre (type of music).
I forgot where I read about the story. Morten Lindberg (proprietor of the 2L label in Norway) took some of his tracks to the US to meet up with equipment manufacturers and importers and played them two different resolutions. At the end they all preferred the higher resolution version and took copies of those.

#61 - batmanames04 :

momei if you like hear no big different with your dac, you need to provide low jitter source to your dac

please install amarra into your mac, the software can provide much low jitter source

#63 - momei :

cpu8088,
1. Your answer suggests you are not familiar with the vapour deposition process and which of the tweeters you mentioned are made from it. You could see I asked you with an inquisitive intention to learn more. I understand Richcom and HiFi Gallery people have visited the Focal factory and I thought you might also have been there to provide answer to the question. HAVE YOU EVER BEEN THERE?
2. See if you are also able to tell us whether the fluid magnet used in the bass cones of Focal Utopia is licensed from someone or some company. I ask because Japanese Kondo is using similar technology in its speakers.
3. I do not dispute that the Focal beryllium tweeter is amongst the best if not the best itself. But then you say the Focal one is pure beryllium suggesting the Yamaha is not. Well I won't say that because I don't know. All I know is that an alloy could be better than a pure metal depending on the intended application. We all know expensive diamond necklaces and rings do not use pure gold but karat or alloy because pure gold is too soft to be suitable for the intended purpose.

#65 - cpu8088 :

momei

here is a link about beryllium

www.goodsoundclub.com/Docs/Truth_beryllium_diaphragms.doc

doing vapour deposit is old technology already.

i dont know anything about fluid magnet used in the bass cones of focal utopia. i understand that focal grand utopia is using electro magnetic bass drivers which is not a permanent magnet but it turns into a magnet by induction of electricity. how this termed as fluid is beyond me.

the focal beryllium is in sheet form not deposits and bonded together on surface of other materials.
when talking about diamond we talk about pure diamond. when we talk about diamond ring we talk about the ring made of gold or brass. 2 different matters.

#67 - cpu8088 :

"""it means it upsample all kinds of 16/44 or 24/96 sample rate to around 96khz."""

should be talking about native resolution files comparison.
there are n+1 upsampling methods or algorithms

#69 - momei :

Batmanames04,
I listened just now to tracks 3 and 4 of Guitar Fever and the solo renditions of the same two songs and more.
They are not my cup of tea. Joey Tang plays well but I do not like using electric guitars as reference because there is no standard sound out of it, affected by the controls in the guitar itself and those in the amplifier.

#71 - DWS :

You guys are really technical, such many discussions on going. I try to insert my comments.

＞ " can computer audio source (cas) really win cd transport for dac?"

If the source is 16bit/44.1Khz, it can be no difference. CD cannot goes above 16/44.1kHz, computer can.

＞ Our human ears cannot hear more than 20khz sound , why 24.96 (around 48khz) sound better than 16/44.1?

24bit has higher resolution than 16bit, and even though there is no real benefits in terms of dynamic range, it helps really much for post processing headroom. This has nothing to do with our listening ability. On 96kHz vs. 44.1kHz, indeed there is no benefits on extend high end range, but this allows different manufacturer to design their own kind of filters, which affects quite high on quality. Well, this is actually going to be next 2 topics of CAS column, stay tune.

＞ Secondly, the Low Pass Filter roll off will push beyond 22khz to 48khz @ 96 sampling frequency, the sound deterioration near the top end (16-20Khz) of the audible range, the sound quality may improve.

Right, this is a good idea.

＞ there are many theories about hearing beyond 20khz

Yes, but they will not change the case that student learnt from primary school our hearing is 20Hz to around 20kHz. There is nothing evolutionary from David's paper, even though I am using his 50kHz bandwidth microphones.

＞ Could you share with us the DAC part of Devalet dpremier compare with Weiss DAC2?

You cannot separate Devialet DAC's with its amplifier, the internal DAC of d-premier has unique I/V conversion which outputs high voltages in Class A. This output stage blends together as the input stage of the amplifier. The Devialet ADH technology works in DSP area to balance between the voltage output (from Class A), and current output (from Class D). In short, the Devialet works at least par with the best amplifiers (pre+power). For example DAC2 with Soulution, Boulder, Vitus, Edge etc. It is very refine and yet well control.

＞ from 20khz to 40khz there is only one octave.

Octave usually describes musical note/pitch. 20kHz is not a musical note, can't really say only one octave above.

＞ Similarly to 200w to 400w is just 3db difference. not much at all.

It is not as simple as this.

＞ The data size doesn't directly reflect the amount of musical information, a 16/44 recording of piano have more musical information than a 24/192 recording of "silence", but the size is much smaller. The data size only reflect the space occupied for storing sample, more room for more information, a store room can store a bicycle or tyres of the MPV but not a MPV, but a big garage can store many bicycle with one or two MPV at the same time.

I agree. It is very difficult to say that HiRes is better than CD if they are not engineer in the same process. I can use the most colorful paint to draw picture, but never can compete with Gogh using the worse paint.

＞ why is it so hard to hear similarly in audio with upsampling of 16/44 format to hi rez to approximate true hi rez?

It may be difficult with iPOD listening, it may not be difficult for high end system playback. But again listening is very subjective. It is true and only ture to the listener him/herself.

＞ quoted from BobKatz (great recording engineer) Mastering audio the art and the science the ultimate listening test: is it the filtering or the bandwidth?

Bob done more than this. He has done extensively on the AES panel and meeting. The case is the filter at high freq. affects the listening, not the extra bandwidth.

＞ My DIY DAC hear almost no different between 16 44.1 and 24/96 material if they are derived from the same source.

The source is originally 16/44.1 or 24/96? I can hear differences but that's just me. See my above comment.

＞ if you believe nyquist theorm, a good dac should hear no big different bewteen 16 44.1 and 24 96 material ( provide they are from the same source)

No one can yet challenge Nyquist for sure.

＞ the view of Bill Schnee to balance that of Bob Katz who is associated with Daniel Weiss. Bill advocates 24/192. In my book I have heard and read about Bill Schnee more than Bob Katz though I do not know in reality who is more famous.

They are both being at the top hill of the audio world. Bill is more practically Bob is well balance between listening and theory. Bill is really getting there to have AES spec. for Blu-ray audio. I talked with him during 2009 and he has been doing so. I am on another camp that computer file will be the one who works in high end audio.

＞ try to audition a high end DAC, your ears can hardly distinguish 16 44.1 and 24 96 recordings

This has some grounds. I think I understand where you are heading to. But I can only agree if these are 2 different recordings. For the same one, I can hearing quite a big differences.

＞ I think that it is better to record the same sound with original file in 16bit/44.1Khz and 24bit/96khz into a CD then play it by a CDP.

If the originally recording is 16bit/44.1kHz, you record this source with 24/96kHz, you will not gain any improvement, usually worse.

＞ there are reasons why a hi-rez recording will sound different. 1) Down sampling might introduce artifacts. 2)our tweeters can sound different when modulated with signals 20 kHz.

I agree with point 1, but I cannot hear above 20kHz.

Then I read many lines about listening. You all have to beware that there is no right or wrong in listening. Taste is the matter. Everyone has his/her own like or dislike on various sounds and of course music. It is great to share but no conclusion can be made.

#73 - hercules :

I believed Kent still using trusty Weiss SARACON for redithering and sampling rate conversion, which support DXD and DSD conversion to CD resolution.

http://www.weiss.ch/p2d/p2d.html

#75 - batmanames04 :

studio use 24/96 analog to digital converter for pcm recording?

#77 - momei :

Many thanks Kent for the detailed answers.

#79 - batmanames04 :

a photo for you guys to relax

#81 - hercules :

Is Thunderbolt backward-compatible with USB and FireWire?

Third-party vendors will sell adapters, available sometime this spring, that let you connect USB, FireWire 400, and FireWire 800 devices to Thunderbolt ports. Thunderbolt won’t make these legacy devices any faster, however—they’ll still be limited to the performance of their built-in components. For example, a FireWire 800 device still won’t be able to transfer data faster than 800 Mbps.

-------------------------------------------------

I expect the thunderbolt port running in firewire 800 compatible mode to work with Weiss DAC will make no difference as ordinary firewire port.

The real difference and advantage can't tell until real thunderbolt device available.

#83 - DWS :

and Batman hing,

thanks for posting STS company. They are our good customers. Actually this workflow can be enhance when the ADC2 is working in 96kHz and maintain at 96kHz until finally Pow-R with SRC back to 44.1kHz.

The main reason is both DNA-1 and DS-1 are working in double sampling rate (88.2/96). This means each of them will upsamples the 44.1/48 signal before processing and downsamples back to 44.1/48 at its output.

#85 - DWS :

Dear ackcheng,

MAN202 uses Intel SSD, Intel chipset and other technologies. Indeed it is what you describes. Next week we will have an important meeting in Munich to solid the idea of MAN202 core technology as platform.

A platform means it can be an on going development path on audio DSP, which Daniel's team has been best at. We also have to set goals in order to have products ready for the market in different time schedule.

We also will discuss the AFI1 development. AD/DA etc. Samuel Groner will lead the analogue circuit designs (He designed DAC202, OP1-BP, and coming OP1-JFET soon.

#87 - ackcheng :

Thanks! That's ver informative!

#89 - ackcheng :

With the technology available with Weiss, he should try to use SHARC as the DSP processor instead of intel chip. This will make subsequent processing software alot more powerful. Also make the competitor difficult to catch up. I am afraid that he will sacrifice the quality for profits.... As mentioned in the interview, even if the Dac is of poorer quality, it will sell because of the brand name.

#91 - DWS :

Dear ackcheng, you mis-understood his point on the "poor quality" DAC.

What he means is nows today people are rating the DAC chipset more than anything else in DA designs. So that even Weiss products do not use the best "specification" DAC chipsets, it will still sell because it will reach the quality standard in our own perspectives, and more often better than other companies using the best "specification" chipset.

We all know how much it costs for a DAC chipset, but solid engineering always more important.

Further about the Sharc DSP. We have those in ALL professional DSP boxes. The best example is SFC2 (sampling rate conversion). It uses Sharc DSP calculates in 40bit fix and floating points.

Few years ago we launched our first software based software "Saracon". Saracon does not only much cheaper (1/5 prices) than SFC2 because lacks of all hardware costs. Saracon calculation is double precision 64bit floating points. It gets rid of AES/EBU standard of max. 24bit (-144dBFS). Therefore Saracon is objectively a better product (PCM to PCM in any rate reaches -180dBFS THD)

Daniel doesn't need to try using Sharc, he is consultant of TI of using Sharc in audio applications. What actually we are looking is similar to Nvidia GPU calculation in graphic card. One of this card can provide floating point calculation better than hundreds of SHARC DSPs, very scary spec. The world fastest company by China is making use of GPU calculation instead of CPU. But this is overkill for audio application.

Intel CPU calculates in 80bit floating points. Since it does not run Windows/Mac OS, the customize Linus has most direct access to resources and execution.

I do concern your comment (but always respectfully) because Daniel is not a business man and does not focus on profits at all. He will be last man to sacrifice quality.

#93 - ackcheng :

Thanks Kent for your clarification,

"Daniel doesn't need to try using Sharc, he is consultant of TI of using Sharc in audio applications"

Actually, I noticed that he uses SHARC in his Powerhouse so I thought using this technology in MAN202 would be a natural progression! Well, may be there are something better around the corner!

Using GPU may even be better than CPU for the kind of calculations we need. Something like CUBA can hack the GPU for this purpose! Someone is already working on it at DIY level!

#95 - ackcheng :

It is a pity that the powerhouse project is killed. But it is good to know that you guys have a clear direction! Well, I'm a professional, so now as a customer, I will leave the tough part to professional like you! DIY is just for fun, for me!

#97 - batmanames04 :

old weiss medea ( ) with sharc processor ( exactly same version sharc processor as accuphase DF 35 digital crossover)

#99 - DWS :

BW102

#101 - DWS :

BW102: Modular system

#103 - DWS :

This modular controller for the DSP moduler racks. Later on there is computer and provides software controls with a Mouse and Keyboard.

Good old times.....20 years ago. The DSP still supports up to 24bit/96kHz and you can still see it in almost all top mastering studios.

#105 - DWS :

no, but it is a very very good product.

#107 - DWS :

Since university I have to learnt the curve of the NS10, I never have Yamaha speakers.

My current one is ATC-SCM100ASL. Thought of a while to update.....

We have protools but not my usual DAW setting.