CD Player vs Computer

[esldude]

I think you get some of the best bang for buck sound quality with USB-SPDIF converters feeding a DAC or using a DAC with a good USB input. I have done a few very rudimentary measurements of noise pickup. At tuner frequencies it is probably worse or at least a hand held AM/FM/SW radio picks up plenty of junk right next to a computer. It goes down pretty far at somewhere around 4-6 ft from the computer it seems.

 

Haven't seen much evidence the EMI/RF upset digital connections or at least not grotesquely. Wouldn't say it has no effect.

 

Analog interconnects even unshielded ones pick up much less junk than I would have guessed. Even wrapped around a PC power supply though odd things were picked up they were down 90-dB or more at least over audio frequencies. Then using a balanced connection seemed to take care of that and any hum until it was buried in the noise at more than -110 dB down.

 

So, if worried about noise in a PC based system one of two or maybe both things make sense to me. One is used balanced connections for your audio equipment. The other is move the computer far away. The Monoprice USB extenders are quite inexpensive, available in lengths up to I think 66 feet, and over those lengths work at least up to 192 khz without problems. Of course you could do both. Finally, just sit your computer at least 6 feet from everything else and use normal interconnects and you have a darn good chance of excellent sound and no meaningful noise involved. One need not become overly neurotic about the noise problem from what I can tell.

[xyz]

It looks like I started a war here. That wasn't my intention at all. I'm glad I didn't mention I have a turntable as well. Thanks to everyone that took the time to post. Let me give a bit more info to clarify. The reason I asked about just playing the CD directly is that, to me, the computer is just acting like a transport. If thats the case, I don't see how you can get better sound by ripping it to a file and then playing it. And that was my question. Mabye it is better to rip it first; I'm just not sure. The way I see it is if you start out with a red book CD, that is all you can end up with from a sound quality standpoint. You can't get more info than what is on the CD; ripped or played directly.

 

esldude,

"Actually, playing from a CD can only hope to equal playing a computer file. The reason being the timing involved. CD playback on a drive has all sorts of mechanical and timing issues. Playing from a file with the correct connection leaving the computer eliminates the mechanical issues, and often has better control of the timing or jitter related problems. I am guessing that is why Barry quoted above hears glass masters, and CD's sound different (those other issues) while listening to a computer all that digital data is correctly copied then played back with more consistency and quality."

 

I'm not 100% sure as to what you are saying, but I'm pretty certain we agree on this point. With regards to jitter, that was an issue that came up a long time ago, and now has popped up again. Before PC audio, the fix for jitter was to put a component between you transport & dac, design the dac and trans in a way that reduces jitter (clocklink, I believe is the correct term) or just use a single box player. Concerning jitter, I know I am OK if I go with a CD player but with a computer, I am not so sure. From the many posts, my best guess would be that playing the CD directly in the PC is like an old dac and trans without any means to reduce jitter and playing ripped file is the same as a single box CD player.

[esldude]

 

esldude,

"Actually, playing from a CD can only hope to equal playing a computer file. The reason being the timing involved. CD playback on a drive has all sorts of mechanical and timing issues. Playing from a file with the correct connection leaving the computer eliminates the mechanical issues, and often has better control of the timing or jitter related problems. I am guessing that is why Barry quoted above hears glass masters, and CD's sound different (those other issues) while listening to a computer all that digital data is correctly copied then played back with more consistency and quality."

 

I'm not 100% sure as to what you are saying, but I'm pretty certain we agree on this point. With regards to jitter, that was an issue that came up a long time ago, and now has popped up again. Before PC audio, the fix for jitter was to put a component between you transport & dac, design the dac and trans in a way that reduces jitter (clocklink, I believe is the correct term) or just use a single box player. Concerning jitter, I know I am OK if I go with a CD player but with a computer, I am not so sure. From the many posts, my best guess would be that playing the CD directly in the PC is like an old dac and trans without any means to reduce jitter and playing ripped file is the same as a single box CD player.

 

Yes, it is a jitter issue perhaps. If you play from the CD transport, there are potential jitter issues from both the mechanical transport and how clocking of it is done in the computer. One of the advantages of the USB to DAC converter if asynchronous in operation (Wavelength, Musical Fidelity V-link, Audiophilleo, Halide Bridge and others are asynchronous) is you have clocking of the data inside these USB-SPDIF converters. Those clocks in those devices playing a file can be generally more precise and greatly reduce jitter vs coming straight out of the CD drive. It is a matter of isolating timing and data flow from other influences. That is why I say a CD at best can hope to equal a file. It is the same data, you don't gain any new data ripping it versus what is on the plastic optical disc. But read real time and fed out from CD that data is not as likely to be well timed as when played as a file with precise external clocks clocking out that data. In a sense playback of a file can simulate the perfect CD transport fairly well. Real CD transports in a computer are not of that level of quality even with the same data in a file vs. on the CD.

 

It is an opinion several others have expressed. But you can always try it like you wish and if it makes you happy then what more could anyone ask for?

 

One other thing is some computer drives read the CD in bursts at high speed. That will be noisy in your listening area if nothing else. You might be able to change settings and force it to play at 1x speed, I am not sure.

[One and a half]

It looks like I started a war here.

From the many posts, my best guess would be that playing the CD directly in the PC is like an old dac and trans without any means to reduce jitter and playing ripped file is the same as a single box CD player.

 

The 'war' usually starts with not interpreting the question to begin with correctly

 

The computer's CD drawer is likely to cost at most USD20 in parts and they are made in the millions of pressed steel, moulded plastic in stark contrast to CD players of the golden hi fi era which had robust frames, supports and vibration reduction. This was in an effort to remove as many mechanical artefacts as possible to avoid it conflicting with the electrical signal.

 

To compare the computer CD's players now in electrical terms, uses the streamed output, directly to the computer's sound card. In this case you have a much inferior mechanical setup to extract the audio from as a stream, let alone before we consider the electrical parameters. I'm not aware of any software that streams this over a USB bus for an external DAC to pick up on. So this stream is connected to the computer's generic although 'reasonable' sound card. For laptops, it's not a choice, but to use the standard card, but a desktop could add an RME or Juli@ 'better' sound card.

 

Given the mechanical limitations within the computer's CD drive, the outboard CD player would usually sound better. If you have a DVD/BD player, compare the analog out to your standard CD player, you will stick with the CDP. That would be before you start on jitter minimisation.

[Jud]

It looks like I started a war here. That wasn't my intention at all. I'm glad I didn't mention I have a turntable as well. Thanks to everyone that took the time to post. Let me give a bit more info to clarify. The reason I asked about just playing the CD directly is that, to me, the computer is just acting like a transport. If thats the case, I don't see how you can get better sound by ripping it to a file and then playing it. And that was my question. Mabye it is better to rip it first; I'm just not sure. The way I see it is if you start out with a red book CD, that is all you can end up with from a sound quality standpoint. You can't get more info than what is on the CD; ripped or played directly.

 

esldude,

"Actually, playing from a CD can only hope to equal playing a computer file. The reason being the timing involved. CD playback on a drive has all sorts of mechanical and timing issues. Playing from a file with the correct connection leaving the computer eliminates the mechanical issues, and often has better control of the timing or jitter related problems. I am guessing that is why Barry quoted above hears glass masters, and CD's sound different (those other issues) while listening to a computer all that digital data is correctly copied then played back with more consistency and quality."

 

I'm not 100% sure as to what you are saying, but I'm pretty certain we agree on this point. With regards to jitter, that was an issue that came up a long time ago, and now has popped up again. Before PC audio, the fix for jitter was to put a component between you transport & dac, design the dac and trans in a way that reduces jitter (clocklink, I believe is the correct term) or just use a single box player. Concerning jitter, I know I am OK if I go with a CD player but with a computer, I am not so sure. From the many posts, my best guess would be that playing the CD directly in the PC is like an old dac and trans without any means to reduce jitter and playing ripped file is the same as a single box CD player.

 

Actually, you can get "something from nothing" in the digital world at times. One of those times concerns the jitter problem. The clock link you talked about to reduce jitter from CD players (referred to these days in its most common implementation as a phase locked loop, or PLL) is not nearly the modern state of the art in jitter reduction. The cheapest method of near-state-of-the-art jitter reduction is to rip the CD to a computer file, then play the file with one of the available "memory players" into a DAC with an "asynchronous USB" input. This avoids the mechanical jitter inherent in disc players and even hard drives. It's a simple thing - physically, massive stuff made of molecules (discs, player mechanisms, laser servos, etc., or hard drive platters and reading heads) is much, much harder to get into perfect time alignment down to tens of picoseconds than a stream of electrons (reading data out of a buffer at the DAC in accordance with a "clock" located very close by in the DAC). Yeah, esldude is right on this one, and I can tell you on my own system it ain't close. That's with a $450 DAC, $50 player software, free ripping software, and a laptop I owned anyway. (I also - pace, everyone - think a good USB cable helps, and you can get one I have used personally and like, the Audioquest Forest, for around $30.)

 

So the basic idea is to get the digital information off the mechanical contrivances where it's harder to control jitter, into computer memory where everything's electrical and easier to control, then through an asynchronous USB interface into the DAC's buffer, where the timing/jitter associated with reading data out of the buffer is controlled by an extremely low-jitter clock (oscillator) right there in the DAC.

[francisleung]

Hi XYZ,

""You can't get more info than what is on the CD; ripped or played directly""

Well based on our experiment, CD tracks ripped and up sampled sounded better than the 16/44.1 tracks.

Please read http://www.computeraudiophile.com/entries/51-COMPARISON-OF-FORMATS-AND-OBSERVATIONS

[sandyk]

Hi Jud

Why did you feel the need to upgrade to the Audioquest Coffee USB cable, and what additional improvements do you believe resulted ?

Kind Regards

Alex

P.S.

I am not disputing any improvements that you may have noticed.Have you been able to try something like the SOtM +5V USB supply, or are you unable to use this product in your gear ?

[Garf]

I would like to know if anyone has compared CD's played in a traditional CD player to a CD played on a computer. I am trying to decide if I should get a seperate CD player or just use the computer. This is not for my main system so I would like to keep the cost down. My budget is $500-1000 for each piece. Has anyone compared components in this price range?

 

My understanding is that you want to stick with spinning discs, based on the information in the original post, and the information in the post above. I know of no audiophile quality CDROM drive for a computer, hence a $500 + budget is a bit overkill. If you are limiting it to CDs, then purchase a separate CD player and avoid the computer in your playback chain.

 

The quality of internal CDROM drives, plus the path the audio signal takes through the inherently noisy computer make it not worth it. If you aren't willing to spend the additional money to upgrade components and build a true computer playback system (ie: rip CDs), then I wouldn't spend anything more than $49 for a decent internal CDROM drive for the computer, and a cheap pair of headphones.

[Jud]

Hi Jud

 

 

Why did you feel the need to upgrade to the Audioquest Coffee USB cable, and what additional improvements do you believe resulted?

 

Hi, Alex. Not so much a need as that it became available for a very good price (including a promotion being run by a financing arm of PayPal to pay over months if wanted, not at interest, but at a considerable further discount!). If I ever wanted to move from the Carbon to the Coffee, this would be the time. I'd been pleased at the step a year or so previous from the Forest to the Carbon, and had heard the Diamond with my DAC in a good friend's system, so I guesstimated I'd like this change, and I was right. I was also curious about the DBS system, which is reminiscent of tech Omega Mikro uses in their cables. But still, would easily have been content with the Carbon if the price on the Coffee hadn't been so good.

 

Additional improvements - Just to sum up succinctly, my collection sounds more like the real thing. Nothing in particular artificially emphasized, but get the sense of being able to hear things more clearly. Not night and day, but a nice little improvement relative to expenditure.

 

Have you been able to try something like the SOtM +5V USB supply, or are you unable to use this product in your gear ?

 

MacBook Pro, so already using battery power. The SOtM, even if I can use it (don't know the answer to that), in my own personal reckoning at this point is too expensive for a better battery.

[sandyk]

Hi Jud

Agreed. I didn't realise you were running on battery, so that's an even more interesting result.

Regards

Alex

[xyz]

Hi XYZ,

""You can't get more info than what is on the CD; ripped or played directly""

Well based on our experiment, CD tracks ripped and up sampled sounded better than the 16/44.1 tracks."

 

Just to clarify, when I say getting more from a CD that you started with is that you can't take a CD, process it, and up and end up with something like a SACD or DVD-A. They are high rez formats and there is more information on the disc. (Thats why the came out with them in the first place.) Another good example is MP-3. If you have, for instance, some MP3's on your PC that you want to burn on a cd and listen to them on a Redbook player, the songs on the burned CD will show a much higher bit rate than 320. The sound quality, though will not be any better. In order to comply with the format, the burning software makes the files look bigger so the redbook player can read the songsin the CD format. So even though the files look bigger, there is no new music data added.

 

Same thing with up sampling. Its how the music is processed to get the higher rate, but still, no new music data is added. As far as sound quality goes, ripped files may very well sound better than the origional. I'm sure there are reasons for this. Whatever the case, though, the better sound is due to something other than adding new info to the music files.

 

I'm sure that most, if not all of you, already know this but I just wanted to make my thoughts clear. Again, thanks for all the replies.

[Jud]

Same thing with up sampling. Its how the music is processed to get the higher rate, but still, no new music data is added. As far as sound quality goes, ripped files may very well sound better than the origional. I'm sure there are reasons for this. Whatever the case, though, the better sound is due to something other than adding new info to the music files.

 

I'm sure that most, if not all of you, already know this but I just wanted to make my thoughts clear. Again, thanks for all the replies.

 

Careful about applying intuition to the digital domain. Upsampling does in fact create new information. For example, the 8x oversampling performed by the vast majority of DACs today provides 7 additional "bits" for every bit in the original Redbook CD. And yes, that additional data does specify what the music should sound like in increments of 1/352,800th of a second rather than 1/44,100th of a second, so there is really eight times as much digital "music data." It's created out of thin air by mathematics/algorithms, but the eventual analog result can very possibly be closer to the reality of the live music than the original Redbook data.

 

Adding this new info doesn't by itself make better sound, but that is simply because neither the Redbook data nor the interpolated 8x data is music. It has to be converted to analog. Conversion to high fidelity analog is considered by most audio engineers more problematic for Redbook data than for 8x data - which is why 8x oversampling is so ubiquitous in DACs.

[goldenpiggy]

Careful about applying intuition to the digital domain. Upsampling does in fact create new information. For example, the 8x oversampling performed by the vast majority of DACs today provides 7 additional "bits" for every bit in the original Redbook CD. And yes, that additional data does specify what the music should sound like in increments of 1/352,800th of a second rather than 1/44,100th of a second, so there is really eight times as much digital "music data." It's created out of thin air by mathematics/algorithms, but the eventual analog result can very possibly be closer to the reality of the live music than the original Redbook data.

 

Adding this new info doesn't by itself make better sound, but that is simply because neither the Redbook data nor the interpolated 8x data is music. It has to be converted to analog. Conversion to high fidelity analog is considered by most audio engineers more problematic for Redbook data than for 8x data - which is why 8x oversampling is so ubiquitous in DACs.

 

I'll throw my 2 cents in...I have a couple of fairly "serious" standalone CDP, an Esoteric UX-3se and Accuphase DP-700. I got them mainly to play SACD, but they do wonders on redbook CD. They each weigh close to 60 lbs. take up a lot of room, and cost a small fortune. They each employ custom-designed transports that weigh nearly 20 lbs. No stamped sheetmetal, but solid blocks of aluminum or cast alloy. Mechanical marvels.

 

Do they sound good? Absolutely. Do they sound better than my computer playback setup? No. My setup is an ancient iMac running PureMusic, feeding an RME FF800 over Firewire, which feeds a Metrum Octave DAC via S/PDIF. The computer setup costs way, way less than either of the standalone players. Way less.

 

I ripped CD layer of various SACDs and play back at 176KHz. It bests the standalone players playing back the CD layer. It is nearly indistinguishable to the SACD layer.

 

I still buy used CDs. First thing I do (after cleaning them)? No, I don't RIP them. I pop them into a standalone player and go through various tracks. The expensive standalone players are in essence preview machines. Pretty to look at though.

 

These will absolutely be the last standalone players I own.

 

Concerning 8x oversampling...yes this is creating additional data, but this is not creating extra bits per sample as in upsampling (like going from 16 to 24 bits.) Rather it is creating extra samples (interpolation) that slot between original adjacent samples. It's done so the low pass filtering can be done in digital domain, and analog filter doesn't have to be so steep (which can introduce phase distortion.) Redbook CD is 44.1Khz, so we're going to have to brickwall anything over 22.05KHz. Looking at it another way...the oversampling pushes the high frequency noise outside of the normal 20-20KHz audio band, where we can do filtering without affecting the audio band.

 

Long live the standalone CDPs...

 

Cheers,

JR

[Jud]

Concerning 8x oversampling...yes this is creating additional data, but this is not creating extra bits per sample as in upsampling (like going from 16 to 24 bits.) Rather it is creating extra samples (interpolation) that slot between original adjacent samples.

Cheers,

JR

 

Howdy, JR. I've seen distinctions drawn between "upsampling" and "oversampling," but often I've also seen them used interchangeably. Nothing wrong with your distinction (though you're bringing in word length as well as sample rate) but the same distinction is not always drawn by everyone using these words.

 

The important part is what you mentioned about filtering.

 

I should also mention that there are DACs, though they are not common, that don't upsample and either rely on upsampling software in a computer or filter the Redbook data.

[phofman]

But, IMO, computer audiophiles are a step beyond, making even vinyl enthusiasts seem pedestrian. It is stupefying how many variables, just in the computer, are reported to make differences in SQ.

 

They are reported to be heard, but not that many reports of confirming the difference by blind listening tests. I would take the unvalidated reports with caution. Having said that, I have no preference of CD player or computer playback, both have its pros and cons.

[beetlemania]

They are reported to be heard, but not that many reports of confirming the difference by blind listening tests.

 

but I don't use DBT as my standard (I trust my experience/ears, what else do I have?; and I have experienced "anti-placebo effect" and, thus, don't fret that I'm prone to placebo effect). If someone claims a using SATA port 2 "crushes" all other SATA ports, I'll raise my eyebrows and brush it off. But if multiple users claim similar results, it catches my attention and I will consider trying it for myself even if it doesn't make a lick of logical sense. In CA, there are a lot of variables claimed to make a difference, most of them are "confirmed" by multiple users. I'd like to think that this is because CA is in its infancy and that things will settle out over the years. But, it also occurs to me that CA marries audiophile nervosa with computer geekdom . . . that can't be good!

[crisnee]

But, it also occurs to me that CA marries audiophile nervosa with computer geekdom . . . that can't be good!

 

Love that!

 

-Chris

[Downrange]

Computer can sound better than any CD player I could ever hope to afford.

 

That said, computers sound very different from one another. Why? I'm not sure, but I've been trying since Christmas to get the Mac Mini (LION) and MacBook Air (LION) I purchased in part for music DLs to equal my old Dell with XP. Not even close, and I've tried all the software. There is something going on besides "bit perfection." Some computers pass more noise, or whatever exactly it is, when they send the bitstream to the DAC, than do others. This is not subtle.

 

Biggest revelation for me was the need to have a galvanically-isolated bridge and a high-end DAC with superior analog circuitry. Without that, the differences shrink away to insignificance as you're really comparing mid-fi systems. The Musical Fidelity V-Link 192 is the single biggest hardware improvement I've made to my system since I began building CA platforms (actually, it's perhaps the single most important component I've ever encountered). It is probably the equal, or near-equal, of the highly reviewed (by TAS and others) Off-Ramp that Empirical sells for much more money. It absolutely reveals the inadequacy of the current Mac platforms (hdwe/sw).

 

My .02.

[phofman]

Just for clarification - CD players do not maintain jitter mechanically, in fact their operation is very similar to PC playback. They have a buffer, samples are read from the buffer by precise clock, and feedback circuitry keeps the buffer optimally full by adjusting CD spinning velocity - e.g. inner and outer borders require different spinning speed. Very similar to how PCI/USB soundcards transfer works - reading RAM buffer by precise clock via DMA, feedback to the software playback chain via IRQs to keep the RAM buffer full.

[esldude]

Just for clarification - CD players do not maintain jitter mechanically, in fact their operation is very similar to PC playback. They have a buffer, samples are read from the buffer by precise clock, and feedback circuitry keeps the buffer optimally full by adjusting CD spinning velocity - e.g. inner and outer borders require different spinning speed. Very similar to how PCI/USB soundcards transfer works - reading RAM buffer by precise clock via DMA, feedback to the software playback chain via IRQs to keep the RAM buffer full.

 

Yes and no. I tried in one thread measuring jitter. My equipment lacked the low noise floor to really measure. Now it was low enough really bad jitter, especially low frequency jitter was apparent using a jtest. I could not find any really to discern on any USB fed Dac out of the couple on hand. Maybe saw some when fed directly from soundcards on laptops and old desktops. Definitely saw some on a Bluray and quite a bit on a DVD player. The weirdest was the DVD player. The first two seconds after changing tracks it was horrendous and mainly low frequency jitter (presumably a mechanical artifact either from spinning or its effect on the power supply. It then got lower as it went, and 18 or more seconds into the track it was pretty low. My resulting FFT of that test signal looked quite similar to what Stereophile got on the same test. It would repeat this anytime it switched tracks on a CD or you skipped it ahead one track. Odd behaviour.

 

Just for clarification vs. published results I think what I had to test with would at least show it there at 500 picoseconds. It would show it above that. It would possibly hint it a bit lower. From what I can tell, and I am no expert or pro at this, I wouldn't see it were it 300 ps or lower. And I may too optimistic at what I saw at that. Still those spinning disc players definitely had some higher jitter.

[phofman]

esldude, I am not saying they have the same performance, just they use similar transfer and control methods. It is quite known that seeking and track switching introduces noise on power supply lines of optical-based devices - motors, tracking/focusing coils. I played with it too a few years ago, the noise was clearly visible on my scope. No surprise it influences the CD/DVD/BR player clock. On the other hand, some audiophiles claim the hear difference between harddrives, some prefer the RAM playback - again the same issue - noise on power lines.

[4est]

"It is probably the equal, or near-equal, of the highly reviewed (by TAS and others) Off-Ramp that Empirical sells for much more money."

 

I have to ask, have you heard an Off Ramp to compare it to the V2? IME, there is quite a bit of difference in devices like these.

[Downrange]

Have you heard the V-Link 192?

 

There's a review on Pg. 3 of this thread. http://www.computeraudiophile.com/threads/8855-MUSICAL-FIDELITY-V-LINK-192-USB-TO-COAX-CONVERTER

 

I do have one of Steve Nugents Turbomodded ECD-1's, and that's pretty amazing. I suspect the V-Link 192 is not quite on the level of the newest Off-Ramp - after all there is huge price difference. The important thing is it's a quantum leap above the V-Link II. This wasn't really evident to me until I assembled the best components from all three systems I have running down here on the CA system. It's a phenomenal device.

[musicbuff]

Why did you switch to dbPoweramp instead of EAC?

[musicbuff]

My Oppo BDP-95 definitely sounds better when I connect my 2 TB hard drive via USB cable (wireworld USB). Both FLAC and WAV sound better than the CD being played on the Oppo. Ripping with dbpoweramp, used to use EAC.

Why did you swith from EAC to dbPoweramp?