You don't want to use the DAC on the sound card... whoooa, no! Junk. Get a soundcard with an external Coaxial (RCA) and/or TOSLINK digital output. These skip the cheapo onboard DAC (only meant to blip and tweep) on the soundcard and will send as good a digital signal as they can out to an offboard unit.
On the cheapo side of things, I have a Soundblaster Live 5.1 ($40-$50) with a Hoontech (Korean) daughterboard for the 5.1 that gives you Coax/TOS digital outs ($40) feeding a Musical Fidelity X-ACT DAC ($100), going into a Jolida 102b (coming next week - replacing a parasound SS setup). For very little dough, a knockout setup. Doesn't take much imagination, just get yourself out of the noisy computer chassis in digital, do the conversion in an outboard DAC, then you know the rest...
How much are you looking to spend?
Roland Edirol UA-5 is a nice soundcard with AD/DA on a USB cable. Works quite well. Costs $250.
I use a SB Audigy II, which has 3.5mm s/pdif digital out, with an adapter (3.5mm mono to mono RCA), run to a Theta DSPro Basic DAC.
It sounds very nice. Much, much better than the analog outs from the sound card.
Many sound cards use this 3.5mm s/pdif out.
I would recommend, when the oportunity presents itself, that you add a dedicated transport. Any CDP should be much better than an internal CD-ROM drive in a computer.
I may upgrade to an M Audio brand card soon, I have heard they are very nice for the money. Will probably still use an outboard DAC though.
"Any CDP should be much better than an internal CD-ROM drive in a computer."
Resolution Audio used a CD-Rom as their transport in their highly regarded opus 21, so this is not necessarily true, and could be chalked up to snobbery.
While I am not familiar with Resolution audio or their setup, I am familiar with the run-of-the-mill $20.00 CD-ROM drive found in most OTC computers.
1) CD-ROMS are designed for data, NOT audio. Inspect the audio outs on any CD-ROM drive, certainly not adequate by my standards.
2) The signal path inside most Computers is data oriented. i.e. Horrific for audio, low quality jumpers here and there, poor quality connectors, poor quality cables, indiotic routing, etc.
The original poster requested information on how to best setup an outboard DAC with his Computer, I gave an informed opinion on how to best achieve this. Along with this information I included a basic observation. In mine, and every other system that I have heard, using a CDP as a transport is FAR FAR superior to the internal CD drive. Outboard DACS or analog from the Comp, either way, a dedictated transport was superior.
I have not listened to some of the very high end sound cards, however, if memeory serves me, the original poster didnt make mention of a high end music server/workstation.
So, I still believe even a "snobbish" JVC or brand-X CDP from say, Kmart, will outperform most (read: NOT ALL) internal CD-ROM drives, since, obviously they are designed with audio in mind.
Granted most CDP will sounds better than most CDRom drives, but there are probably some exceptions out there. I don't diagree with your general statement, just your absolute terms. FWIW RA pulled the CDrom because of an audible whine, but it still sounds pretty damn good!
yes, CDROMs transfer data. If you don't do D-A conversion till the audio data gets out of the computer, then nothing is lost along the way. A bit is a bit. We are in digital world here. Until the digital data is converted to analog signal, then nothing is lost. Some error checking is performed when the CDRom is reading data, either by the operating system, or the IDE controller, or the CDROm itself.
Who uses CDROM's built-in DAC these days? Windows XP reads audio data in digital by default.
yes the computer transfers and works with data. Digital audio is one form of data. Data integrity is very important in computer design. It is an insult to a computer engineer to say that the computer does not maintain data integrity.
Once you got the data out of the computer, then it is pretty much up to the DAC to give you an interpretation of the digital data.
Boogie, maybe you should take it over to the "cables" forum and explain to all the nice Audiphiles that their ears are all wrong.
They, for some reason, think that digital cables sound different. 1s are 1s and 0s are 0s, data is data, by your own admission.
And thats just the cable issue you didnt address the "transport" at all. I guess all those audiophiles who think they make a difference must be wrong as well.
"It is an insult to a computer engineer to say that the computer does not maintain data integrity." Who said that. Let me read my post again. Umm nope.
Well since we are drawing conclusions....
It is an insult to an Audiophile to conclude that all his hard earned money and time are misguided on useless CD transports and high dollar Digital cables.
I guess dejittering is a waste of time/money as well?
Jposs, I said "should" sound better. Definately not absolutely will. Should isnt absolute by any stretch of the imagination.
While trying to not delve into semantics, I had a problem with "any", not "should". But I dont think its necessary to argue with you, because I dont think we disagree.
You're right. It is an insult to an audiophile to conclude that all his hard earned money and time are misguided on useless CD transports and high dollar digital cables.
My point is, CD-audio data is read correctly in digital by CDROM drives, and nothing is loss until it is manipulated or converted to analog. Whatever is stored on disk is read "as is" by the stupid CDROM drive. Hehe, you want pure signal path, the digital signal path is as pure as you can get. :) Data integrity is always the case in the computer. If it is not the case, imagine corrupted files and data happening all the time as you're working with a computer. Since corrupted files and data happen like once in a blue moon on a healthy computer, then you can be sure that data that contains audio information is kept in its full integrity as well.
Well, we're not talking about error-correction stuffs when reading the CD by the CDROM, but the same error-correction is used for other types of data (i.e. spreadsheet, documents) as well.
At the end of the pure digital path, you got a DAC. This is where the action is. Some DAC do tricks like upsampling and others do tricks like dithering. As long as the CD audio data is concerned, it remains pure until this stage.
Dejittering? for what? a bit, is a bit.
Jposs, I figured we were on the same page, I just wanted anyone reading our stuff to garner good usable information. That, and the snobbery thing sorta lit a fire under my butt, but its all in good fun, without tort and retort life would be too boring. =)
Boogie, I think we are from different schools on this issue. Some, feel data is data, others, such as myself, believe there is something more, shall I say, ethereal.
I will acknowledge that it stands to reason that the data should be just that, data, but too many times a simple change, such as a transport drive or digital cable change, made a difference in the sound. Sometimes a very large difference.
I feel that by default musics 3 demensionality is more than simple data, it conveys both Time and Space.
I would like to be clear that, simply because I said most computers and their CD-ROM drives probably arent the best medium for garnering musical data, doesnt mean that they(computers) somehow dont have data integrity. They are marvelous machines, in fact after audio gear, they are my second hobby, and probably command more of my time than the audio stuff. I have and maintain several. I do not questions the computers data integrity. As a matter of fact they are designed for just that. However, data integrity, while critical, is not synonimous with high quality audio, IMO.
I think I am going to start a thread, in the digital forum, to gather some opinions on this matter. I think it will be enlightening and fun.
I see. Then I suspect that these transports doesn't maintain the purity of the signal, i.e. they manipulate the signal somehow.
I really want to see some strong argument that a bit is more than just a bit. In the analog world, we know how more expensive gears manipulate the signal to sound "expensive" with all the filters along the way. In the digital world, unless then data is manipulated, then it will remain exactly as what recorded on the CD.
I'm not questioning "high-quality" audio since my point is stressing that the digital data read by CDROM is as pure as what is on the CD. We all know that pure data doesn't necessarily sounds good, as well as expensive speakers never have real flat response. My point is audio signal purity, which comes from data integrity kept in a computer.
I sold my california audio labs transport/DAC pair, and am using my CDROM and my computer as the audio source. The signal is kept in digital from the CDROM till it gets out from the computer box, and converted to analog by the Stereo-link 20-bit DAC (CD audio is 16-bit).
I would never turn back.
Guys, what's wrong here is that "CD-audio data is read correctly in digital by CDROM drives". It's correct for audio extraction with CRC check, multiple re-reads, and error handling, but not for playing CD in CDROM drive. When you play audio CD real-time, the received audio data will almost never be 100% lossless. In this case much better built CDP or Transport easely outperfom any CDROM. But if you do audio an extraction from CDROM to hard drive using EAC and Plextor or another very good drive, that HDD may become a perfect source.
data is data, audio or other data (word document, software, etc), they are all data.
If you say you can't rely on CDROMs to read audio data, then you also can't rely on the CDROM to read the Microsoft Office installation CD that you paid $700 for it. I don't know. Does not make any sense to me.
If I may make an analogy Boogie, if that were the case, symphonies would only hire Musicians who can read music well and would care less about the ability to actually play the instruments.
"Reading" the Data is merely the tip of the iceberg.
Data IS data, but the way how they read it are different. I would suggest you to read some FAQs about 'how cd-player works'. Shortly, when you read data CD, your drive works on, let's say, 24x speed with buffering, it has enough time to read, check CRC, and re-read if necessary up to several times. And re-tries ARE necessary, due to the nature of CD media. When you play an audio CD, your drive (except Meridians?) has only one attempt for anything, obviously no CRC check performed. The result is not 100% accurate data retrieved, resulting in jitter, which directly affects sound quality.
On the other hand, audio extraction (from CD to HDD with appropriate software: with CRC, re-reads, variable speeds, etc.) is much closer to the way how data CDs are processed, and thus a perfect copy may be obtained.
Very interesting discussion, EAC is Exact Audio Copy, done by a cool computer dude thats probably gonna sell his now-free-shareware once its perfected. I read up on his digital extracting/burning software, (its recommended in Stereophile magazine by at least 2 writers there, for cd burning,) I must add that bits are bits but what happens when you MOVE bits is a whole nuther matter. To take bits from a disc and put them on your hard drive, even with all the numerous re-checking of data that takes place does put the integrity of the original in question. However, I would think that this is the lesser of the evils. Now that the data has to be re-read from the hard drive, we have a whole nuther question. Even if the bits are exact, when the bits are sent from the HD to the external DAC, is this going to be a truer transfer than a stand alone cd player? That is the one of the questions. To me at least. Jitter, dropped bits is another question. But what about the overall musicality of the bits? Does this exist yet? Has it been distorted yet? or has this "distortion from original, true bits assemble" not taken place yet? Is it the DAC or the data stream or both as a source of distortion, is the question I'd pose. We need to establish this to get on with the comparison.
This is still a much debated question in many various forms going on in audio. I cant fully side with "bits is bits" myself. Too many experienced audio people still struggle to account for differences and resort apologetically to subjective "magical" terminologies. Alotta audio dudes still feel that these indefinable, unscientific terms are essential in order to describe the differences that science thus far can't. Tubes vs. Solid state anyone?
Talking about whether CD-ROM drives are truly accurate misses the point of computer based systems. The original question was about putting together a high end computer based system and if that's the case the CD-ROM is not important in playback since the music will be stored and played from the computer's hard disk. The role of the CD-ROM will be solely to rip the CDs to the hard disk. If properly setup, this type of computer/hard disk playback is of very high quality. Systems of this design are routinely used in recording and mastering studios.
As far as soundcards go, I recommend RME
. It can be configured with optical (Toslink), RCA (SPDIF) or XLR (AES/EBU) digital I/O.
Well yes Onhwy61, but he did say, "this computer and its TRANSPORT will serve as the source...." although I agree with you, a harddrive based playback system probably would be better. It truly is an interesting topic. I am curious where the integration of HD technology with digital audio playback will take us.
Would RME soundcard give me better results than inexpensive devices like xitel or link 1200?
Seems like a waste of the RME's capabilities just to use it as a link to an external DAC. Do we have to pay that much simply for a quality link?
Gonglee3, for 2 channel applications most of the RME product line is overkill, but they do offer superior performance. Is the subtle increase in performance worth the additional cost and complexity? I think we each have to answer that question for ourselves.
Even though you don't plan on using any onboard DACs on your PC's sound card, a good PC sound card with built-in clock synch features and digital I/O will provide excellent high end audio results when you interface your hard drive music source with a high end DAC with word clock synch connections. This will dramatically reduce jitter that can arise from interfacing problems. Lower cost interfaces like the Xitel are similar to other low cost solutions from Edirol and M-Audio which do not take into account jitter generation. If you are serious about PC audio, then consider the RME Audio or Lynx Studio Technologies Model 2 or Model 22 sound cards.
"Sound on Sound", an excellent professional audio magazine from the UK, gave an excellent review of the Lynx Studio 2. The article's author explains why the internal clock provides a deeper sound stage, clearer detail and a more natural sound than sound cards which do not possess a clock mechanism. Keep in mind that the review just evaluates the card's DACs, which goes to show that onboard DACs can sound just as clear and focused as an outboard DAC. Even bypassing the DACs on the card, you can still take advantage of clock synching features. Just go to http://www.soundonsound.com and search for "Lynx" and "RME" to read their reviews. I find that their evaluations and opinions of sound quality coincide with mine. In this magazine, you'll never read a rave review of SoundBlaster sound cards for audio quality.