CD player affect SQ using External DAC?

[Vinny]

Hi guys
I've a dumb question........... without looking at the circuit deeply(I'm no where close to engineer that wouldn't help me even I look at it ops: ), would the cd player affect SQ anymore if using external DAC for decoding the CD? It seems that the DAC is doing all the work and the CD player just need to read the CD to get the 0/1 signal and sent through digital/coaxial signal. Based on these, the CD player would not affect anymore as long as it has a clean signal output from reading the CD to the digital output?

Thanks for answering while it's all elementary ;x(

[DrJRapp]

the timing accuracy of the player (jitter) could effect the sq as could the players ability to convey as loss free bitstream as possible to the external dac.

[wkhanna]

Others here are far more informed than I, and I'm sure they will chime in soon.

CD's have their own DAC (digital to analog converters). In general, the more you pay, the better the chip performance. If you have a 'good' pre/pro or receiver, and 'not so good' CDp, you may get better SQ utilizing the DAC in one of those by using the coax output on the CDp. In this case you are using the CDp as a 'transporter'. Think of it as a turn table and the CD disc as vinyl LP record.

However, the quality of the transport can significantly affect SQ also. Just like with the turn table example, the control of the speed that the disc is moving is critical.

Some component setups on the higher end of the scale use a separate transport that does nothing more than spin the disc, read the digital signal, and send it to the output. These transports utilize complex circuitry and electronic clocks that precisely control the speed to limit 'jitter' or slight variations from the precise speed required to deliver perfect reproduction of the digital data.

This signal is then sent to a separate DAC component which does the processing (converting the digital data to an analog signal). If the transport or DAC unit has a means for controlling the volume (signal level), you can run it directly to an amplifier, which is said to give you the purest SQ since you bypass any further processing that a pre-amp or receiver may induce and possibly downgrade the signal quality.

If I have made any errors in this description, please jump in and correct my mistakes. I'm still in the ‘learning mode’ with much of this Digital HiFi stuff.

Edit: oops, the 'Good Dr.' beat me to the punch while I was composing.

[Vinny]

Thanks :T
It makes me lean towards getting a CD player a little bit while considering between DAC or CD player.
I've been throwing more and more CDs into computer as FLAC and output via coaxial, DAC seems the way to go but CD is still beats the FLAC's quality :roll:

[DrJRapp]

While there are quite a few good DACs out there, you'll probably find that if you buy a good, yet inexpensve CD player such as the Cambridge Audio Azur 640C you'll get internal DACs that rival many out there for musicality..

[Victor]

Vinny,

The laser pick-up mechanism and the signal recovery from the CD medium have been perfected over the years to the point that even the cheapest portable CD player works nearly identical in this respect to the player costing thousands. The error correction protocol has been hard-coded and remains the same regardless what CD player you happen to have. This reality amounts to the fact that you can be certain that the bits are picked up and assembled into a digital word with the same accuracy, regardless of how much you paid for your player.

The CD Player’s clock accuracy might be an issue, while what is never really an issue is the jitter, because the CD player is a fundamentally synchronous system. Since there is nothing there to synchronize, the jitter phenomenon is not a problem.

How can you tell that the clock in your CD player is good or not good? Well, - the clock accuracy problem will manifest in an inaccurate D/A conversion process, which is definitely measurable and possibly audible. The mundane THD test will immediately show this problem to be evident or totally nonexistent. For instance if your CD player indicates a 96 dB of S/N, then you are in great shape with a nearly 16 bit of requisite analog output. These days the CD clock accuracy is taken for granted, - it is accurate to a few parts per million even in the CD players that are given away for free.

There is very little reason to use an outboard D/A box, unless you are suspecting that the D/A process in your CD player is somehow flowed. If you use an outboard DAC, then the jitter may play a role due the clock recovery from a SPDI/F stream being less then accurate and there is no way to truly accurately recover the clock from the SPDI/F signal. Many outboard DACS do a stellar job here, but it comes at a price. The Benchmark DAC uses a Sample Rate Converter to significantly attenuate jitter and it is the most elegant method of dealing with this problem.

Finally the resolution of the D/A chips used today in most budget CD players are an order of magnitude better then the resolution of the analog circuitry that immediately follows the D/A chip. This is true for any CD player regardless of how much you pay for it. With this in mind spending more may buy you an additional 1 bit of resolution because most expensive CD Players are actually 16 bit accurate and most budget CD players are 15 bit accurate. Those are the best and the worst specifications, and even those numbers are questionable because I have seen $200 players that can output 16 bits and $2000 players that cannot. It will definitely not buy you a better sound considering how much money you must spend to get that 1 bit. While it remains to be seen (or heard!) if you can actually hear that extra bit.

Regards,
Victor

[whoaru99]

Thanks :T
It makes me lean towards getting a CD player a little bit while considering between DAC or CD player.
I've been throwing more and more CDs into computer as FLAC and output via coaxial, DAC seems the way to go but CD is still beats the FLAC's quality :roll:

With large drive space at low cost as it is these days, is it even worth using any compressed format, lossless or not, for music servers?

[Vinny]

With large drive space at low cost as it is these days, is it even worth using any compressed format, lossless or not, for music servers?

No it doesn't worth it. I should have done wave and get no compression at all.... but one of the hard drive is still down(160 out of 250,160 and 80) and the other two are flooded with the back up that FLAC is the best I can do atm. I'll get them back to wave format once I get my 160gb back on track (get my lazy butt moving is one big factor ops: )

[Vinny]

While there are quite a few good DACs out there, you'll probably find that if you buy a good, yet inexpensve CD player such as the Cambridge Audio Azur 640C you'll get internal DACs that rival many out there for musicality..

Thanks for the info about the 640C! I've been looking into Rotel before and compare to the DAC. I don't have enough to get both of them home and compare. But I guess I can do it with Cambridge one if I can find a good deal for it :B

[Vinny]

....

Thanks, now it seems getting more sense on why people are still paying a lot for a CD play but not working in transporter/DAC way like pre/amp do while two different unit can step up easier

[Shawn Parr]

Interested in jitter, and how it affects things?

Here is the jitter article by Bob Katz. Pretty in-depth.

Some of the more informative parts are toward the end, especially the information from Sony Music about dealing with dither during CD manufacturing. Especially the section labelled "Can Jitter in a Chain be Erased or Reduced?"

The two important things to take from that article is:

1. The only way to be sure you are getting the best possible from your stuff is to experiment and use your ears.

2. CD players have jitter, and it can have an effect on things. This is why the Benchmark DAC is as popular as it is, it is very effective at re-clocking.

From my experience if you are using appropriate equipment and properly spec-ed cables (75 ohm with 75 ohm connectors) there is very little jitter introduced by the digital connection. Jitter noticed from that situation tends to be the jitter of the master clock being imposed on the rest of the chain.

[wkhanna]

I, too have been looking closely at the CA 640C v1.0 and v2.0. The main differance being the v2.0 uses two seperate DAC's, dual 24-bit/192 kHz Wolfson WM8740 DACs, one for each channel.

Is this correct, and does this feature affect the 'jitter' issue?

[JDH]

I, too have been looking closely at the CA 640C v1.0 and v2.0. The main differance being the v2.0 uses two seperate DAC's, dual 24-bit/192 kHz Wolfson WM8740 DACs, one for each channel.

Is this correct, and does this feature affect the 'jitter' issue?

I'm using a Slim Devices Squeezebox 3 connected to a Benchmark DAC 1 then connected to a Rotel RB-1080 power amp, this setup gives me flexibility and sound quality at a reasonable price.

[DrJRapp]

I, too have been looking closely at the CA 640C v1.0 and v2.0. The main differance being the v2.0 uses two seperate DAC's, dual 24-bit/192 kHz Wolfson WM8740 DACs, one for each channel.

Is this correct, and does this feature affect the 'jitter' issue?

Actually version 2 uses the 2 DACs in push pull, not one for each channel. This theoretically should reduce distortion significantly.

[wkhanna]

Thank you for the clarification, Good Dr.!!

[Victor]

Interested in jitter, and how it affects things?

Here is the jitter article by Bob Katz. Pretty in-depth.

The two important things to take from that article is:

1. The only way to be sure you are getting the best possible from your stuff is to experiment and use your ears.

2. CD players have jitter, and it can have an effect on things.

Shawn,

I read the article by Bob Katz. It is written from a recording engineer point of view and it focuses mostly on the jitter at the recording end of the process. The article does not break any new ground but it is a fairly well composed synopsis on jitter.

Mr. Katz does not say that “…The only way to be sure you are getting the best possible from your stuff is to experiment and use your ears…”

What he does say, - “…one of the difficulties is finding measurement instruments capable of quantifying very low amounts of jitter. Until we are able to correlate jitter measurements against audibility, the ear remains the final judge…”

He implies and correctly so that jitter is extremely difficult to measure. Particularly the jitter in the picoseconds. He then says that it is because he can’t measure it, - the ears are the only choice left. He also does not hide the subjectivity of the assessment by ear by indicating that it is HE who can hear it and not ALL. He is implying that in the absents of a standard test his results are his alone and should be understood that there might be other reasons for his observations, - other then jitter.

Furthermore he does not say that, - “…CD players have jitter, and it can have an effect on things…” What he does say is that he can “…hear the subtle (but important) sonic differences between jittery (and less jittery) CDs…”. He is talking about the media not the CD Player.

There are several points that Mr. Katz makes in his article that I would argue with or simply disagree with on the scientific grounds alone, but this is not the issue here.
The issue is how concerned we need to be with jitter when it comes to the playback chain, particularly in the case of a CD Player.

Jitter can be understood and seen as one understands and sees noise, because it is a manifestation of noise. As such, the Math tells us that regardless of the type of the jitter, - correlated or not, the THD test will immediately show if it is a problem in a particular system.

I wonder, - is there a reason to suspect that the particular CD Player has an unacceptable level of jitter due to, let’s say, a noisy power supply or less then stable crystal oscillator, or whatever the case might be? Well, - if the published THD+N is around 16 bits then I would have to say, - NO, - the possible clock jitter, the system power supply and the oscillator are all within the required design guide lines.

My point earlier was that a CD Player being a synchronous system is immune from jitter issues. I based this on the fact that in general there is no evidence to suspect that the master clock oscillator in the typical CD Player is jittery. While bad designs are certainly possible it is not a norm.

Furthermore, in a typical over-sampling CD player the over-sampling process introduces an immunity to any jitter possibly present in the clock. The Sigma-Delta type D/A converter which these days are found in virtually all CD Players also has a built in algorithmic immunity to jitter.

If I had to estimate, I would say that the distortion of the signal coming out of the D/A in a typical CD Player has components due to clock jitter in the 10’s of picoseconds range. Can you hear this distortion? Mr. Katz says that he can. He also says that it is definitely due to jitter. Please notice that Mr. Katz’s tests were done during the recording or A/D process, where jitter may account for the unacceptably high level of THD which might be audible, while also producing bad recording. Mr. Katz never really says that the jitter in the D/A process of the CD Player is just as audible.

regards,
Victor

[Kobus]

Victor, I must just say that while you opinions are sometimes not what the typical audiophile want to hear, I do appreciate them and thank you for your effort and time spent giving us informative reading.

Regards


Kobus

[Shawn Parr]

He implies and correctly so that jitter is extremely difficult to measure. Particularly the jitter in the picoseconds. He then says that it is because he can’t measure it, - the ears are the only choice left. He also does not hide the subjectivity of the assessment by ear by indicating that it is HE who can hear it and not ALL. He is implying that in the absents of a standard test his results are his alone and should be understood that there might be other reasons for his observations, - other then jitter.

He also says:

Listening tests
I have participated in a number of blind (and double-blind) listening tests that clearly indicate that a CD which is pressed from a "jittery" source sounds worse than one made from a less jittery source? In one test, a CD plant pressed a number of test CDs, simply marked "A" or "B". No one outside of the plant knew which was "A" and which "B." All listeners preferred the pressing marked "A," as closer to the master, and sonically superior to "B." Not to prolong the suspense, disc "A" was glass mastered from PCM-1630, disc "B" from a CDR.

If clocking systems are so good in modern CD players then why would jitter recorded onto the disk during a completely digital process be a problem? Maybe:

David Smith (of Sony Music) was the first to point out to me that power supply design is very important to jitter in a CD player, a CD recorder, or a glass mastering machine. Although the FIFO is supposed to eliminate all the jitter coming in, it doesn't seem to be doing an adequate job.

Furthermore he does not say that, - “…CD players have jitter, and it can have an effect on things…” What he does say is that he can “…hear the subtle (but important) sonic differences between jittery (and less jittery) CDs…”. He is talking about the media not the CD Player.

Ah, but the data is RAM buffered and re-clocked by the playback device. Thus any jitter you hear coming off the playback device is directly effected by the playback device itself. An adequate clocking system would correct the jitter recorded onto the disk, this is exactly what the DAC1 does.

My point earlier was that a CD Player being a synchronous system is immune from jitter issues. I based this on the fact that in general there is no evidence to suspect that the master clock oscillator in the typical CD Player is jittery. While bad designs are certainly possible it is not a norm.

No digital devices are immune from Jitter. The only place jitter is not a concern is when making digital copies, as the playback device, if utilizing a good clock source, can re-assemble the data without the digitally propagated jitter. This is the entire point of a FIFO in a digital playback device.

If I had to estimate, I would say that the distortion of the signal coming out of the D/A in a typical CD Player has components due to clock jitter in the 10’s of picoseconds range. Can you hear this distortion? Mr. Katz says that he can. He also says that it is definitely due to jitter.

Actually this article is pretty old (2002). Modern good clocking systems are sub-10ps. The average nowadays is between 6-9 ps. A cd player is definitely in the 10s of picoseconds, which in reality probably means 10, 20, 30, or more. 20ps of jitter compared to 8 would be noticeable by anyone who listens to high quality music. Mind you my wife probably wouldn't hear it, but she also doesn't notice when entire channels are missing.

Please notice that Mr. Katz’s tests were done during the recording or A/D process, where jitter may account for the unacceptably high level of THD which might be audible, while also producing bad recording. Mr. Katz never really says that the jitter in the D/A process of the CD Player is just as audible.

He discusses A/D at the beginning, but all the tests involving CD players, and most involving DAT machines were only testing D/A and digital only record loops (DAT to DAT for example).

One very important concept here is the FIFO. When a CD player gets the data from the disk, it is buffered, then sent to the converter based on the clock of the CD player. Thus any jitter on the disk that was created after the A/D stage, should be corrected for. These tests show that in many playback devices this is not the case. This seems to be due to a combination of a lack of good clocking, and the electrical infrastructure to support the clock (powersupply, etc.).

I finish up with some real world experience. We did a test at one location where I worked. We took a Protools system, and [b]with playback only[b], i.e. A/D was not part of the process, we clocked the system using the internal clock, and compared it to being clocked via an Aardvark Aardsync 2.

The test was done all day long, with groups of 5 to 10 people. A total of over 60 people ended up listening to the test. Of that, maybe 2 or 3 didn't hear any difference, and most heard enough difference that they would consider it a major change.

I also see and hear a number of sources talking about jitter and comparing THD specifications. One aspect of jitter can't really be tested via THD measurements. THD is added harmonic content to the signal.

With jitter, since it is affecting the clocking of the digital system, it can change the signal without necessarily adding additional harmonic information.

Imagine a system with really bad jitter specifications. Now imagine we are playing back a 100hz wave. With THD, we are measuring how much 200hz, 300hz, etc is being generated. But with how jitter effects our clock that 100hz tone itself may fluctuate up to 101hz and down to 99hz. (Remember this is a hypothetical example that has been way over-exaggerated). THD+N measurements can't detect this type of distortion as there is not any harmonics created from it, only the fundamental signal itself is changing.