How inexpensive could a pre-pro be? (1 Viewer)

[Roger Kint]

Even today, CD playback is inferior to vinyl.

Larry B:

This may be a dumb question, but can you explain how CD playback is inferior to vinyl?

 

[Larry B]

Chung:
Thanks for the info. Even though we did not see eye-to-eye, I enjoyed our discussion. In my mind, there's nothing like a civil, mature exchange of ideas.
Roger:

Larry B:
This may be a dumb question, but can you explain how CD playback is inferior to vinyl?

There's nothing at all stupid about that question though, since it deals with a subjective topic, it's a bit difficult to answer. In simple terms, vinyl sounds warmer, richer and, if you will, "less digital" than CDs.
Until about a year ago I listened exclusively to CDs, but was constantly intrigued by those who maintained that vinyl was better. I wondered if they really were onto something, or if they were simply dinosaurs who couldn't give up the past. To satisfy my curiosity, I arranged a demo at my audio shop. Specifically, we identified a number of pieces of music which between the two of us (the salesman and I) we had on both CD and vinyl. Once the demo began, it took me all of 30 seconds to realize how much more satisfying the vinyl was. I ordered a rig and once it when it was set up, my GF, who (as you might guess) normally has little or no interest in matters audio (though she loves music and used to play), immediately remarked at the improvement. I should mention that my CD playback equipment is better than average.
If you're truly interested (as I hope you are), I strongly recommend that you go audition vinyl for yourself; only then will you realize how grossly distorted, artifical and fatiguing CDs actually are.
There is one last point I should make, and that is that a $200 CD player will sound better than a $200 turntable. This is because it's cheaper to mass produce a CD player, than a turntable. However, once you get into a "reasonable" range for a musical source (say, $1,000), the turntable will blow the CD player out of the water.
If nothing else, I hope I've intrigued you enough to go listen.
Regards,
Larry

 

[Ricky T]

Bruce and Larry,

I'm surprised that you guys say that Sony ES gear absolutely stinks in analog. Don't the classic 7ES and now best in their classes 777ES, XA777ES, 333ES, 555ES CD/SACD players convert the digital source into analog stages before passing the ANALOG signal to an analog volume control?

Larry,

In the 2 channel world, it's debatable between an int and stereo separates. It all has to do with the implementation of the design (inside the int) and probably how much power you need. Chung point is, ALOT of 2 channel audiophiles use integrated amps, and don't really seem concerned about the internal amps degrading the sound. In fact, many seem to be more concerned with an internal tuner degrading the sound.

 

[Larry B]

Ricky:

In the 2 channel world, it's debatable between an int and stereo separates. It all has to do with the implementation of the design (inside the int) and probably how much power you need. Chung point is, ALOT of 2 channel audiophiles use integrated amps, and don't really seem concerned about the internal amps degrading the sound. In

fact, many seem to be more concerned with an internal tuner degrading the sound.

I think the internal amps probably do degrade the sound but, if well designed, an integrated amp can still sound provide a fulfilling musical experience.

Larry

P.S. We're practically neighbors - - - I'm in Hackensack!

 

[Samuel Des]

In simple terms, vinyl sounds warmer,

Couldn't this be due to lower SNR?

 

[Larry B]

Samuel:

Sorry to have ignored you; I seem to have missed your post.

Not sure I understand. Do different DSP modes interfere with each other? Or are you referring to hardware interfering with each other?

I was referring to the latter. And the "interference" is, I suspect (though I'm not an EE), not just in the form of noise; it's also in the form of optimization. High-end manufactuers have known for years that moving a circuit from one physical place to another affects the quality of the sound (although audible, I doubt this is measurable). Some of them spend countless hours (and thus lots of money) optimizing this. I sincerely doubt that this receives much attention from the designers of most (though admittedly, not all) preamp processors.

Larry

 

[Larry B]

Samuel:

Couldn't this be due to lower SNR?

I have no idea. But I do know what I hear.
Larry
P.S. Lack of jitter is probably also a very significant factor.

 

[John Sturge]

And, you could *always* go out and spend that $500 on a receiver

Ick, Ick ,Ick! Man I must be be the biggest hater of "congested electronics not properly isolated in a box" $500 is not going to get you quality at the price range. Sad.

My first and last receiver was HK-520, Hated the sound. I'll never go back.

Opinion and that's all it is.

 

[Scott H]

audiophiles no longer had to contend with wow-and-flutter

And said hello to jitter.

I share this not to disparage chung, but to indicate a different conclusion using similar components (Sony ES and Mondial amps). I consider myself a budget minded audiophile, I have been around very high-end stuff learning how to listen since 83, and in HT my primary focus is two-channel music reproduction. In the early 90s I had an upper model Sony ES CD player. I liked it. I found it more revealing, less veiled than the Denon it replaced (speaks were B&W, btw). But when I aquired better amplification, an Acurus A250, that ES player was gone. It was replaced with an NAD 512. My current, always changing two-channel system still has that NAD. I have critically listened to the ES players compared to competing NAD, Rotel, and Marantz players throughout the years and I would personally choose any of them over Sony, with one caveat - I would consider the ES players as a transport mated to an outboard DAC.

 

[chung]

And said hello to jitter.

Not sure how one can quantify the effects of jitter. Does it show up in distortion measurements using test tones on a CD, or noise measurements? I am lucky to have never heard the ill-effects of jitter. Now wow and flutter, that's jitter blown up hundreds, if not thousands of times.
Differences in CD players are 2 orders of magnitude less than differences in LP vinyl music reproduction equipment. I am sure if you look for them, you will find those differences. But such differences are so small compared with speaker differences, room differences, etc., that most listeners would not be able to reliably favor one CD player/DAC over another, if the designs are close in vintage/performance. If you compare a 1983 and a 2002 CD player at the same price point, yes you will find some difference.
Again, I recommend this article on subjectivity:
Link Removed

 

[Craig_Kg]

I have no idea. But I do know what I hear.

Larry

P.S. Lack of jitter is probably also a very significant factor.

Jitter and ringing are the big complaints with CD vs vinyl. Jitter can be tamed by reclocking the output while the best solution for ringing (according to audiophiles) is a low pass filter with a smooth rolloff starting around 15kHz best implemented by Wadia. Note that the filtering removes the higher frequencies that CD can reproduce but vinyl cannot. The ringing is caused by the sharp cutoff (36dB/octave) at 20kHz to suit the 44kHz sampling rate of CDs. The superior reproduction of DVD-A and SACD is primarily due to the higher bandwidth allowing a smoother high frequency rolloff (thus eliminating ringing) rather than extra high frequency information being directly heard. DVD-A is still susceptible to jitter but the SACD medium should be immune to it (but has an amplitude distorion element instead).

 

[Craig_Kg]

Not sure how one can quantify the effects of jitter. Does it show up in distortion measurements using test tones on a CD, or noise measurements? I am lucky to have never heard the ill-effects of jitter. Now wow and flutter, that's jitter blown up hundreds, if not thousands of times.

The problem is that our brain uses the phase difference across our ears at high frequencies (>10kHz, I think) to determine the source of a sound (amplitude is used at lower frequencies). Jitter (and ringing even more so) confuses the brain by smearing the very high frequencies, degrading the sonic image and increasing fatigue.

 

[chung]

Jitter can be tamed by reclocking the output while the best solution for ringing (according to audiophiles) is a low pass filter with a smooth rolloff starting around 15kHz best implemented by Wadia. Note that the filtering removes the higher frequencies that CD can reproduce but vinyl cannot. The ringing is caused by the sharp cutoff (36dB/octave) at 20kHz to suit the 44kHz sampling rate of CDs.

Oversampling has removed the need for sharp filters at the output of the DAC's. That solution has been implemented for at least 15 years.

 

[Saurav]

Now wow and flutter, that's jitter blown up hundreds, if not thousands of times.

Absolutely. AFAIK, the jury is still out on how human perception reacts to all this. For instance, we all know that the frequency response of human hearing is logarithmic - doubling the frequency appears as steps of equal magnitude (octaves). On a similar note, I don't think we've totally understood human response to various forms of distortion. Jitter and wow/flutter are both distortions which make the reproduced sound different from the real thing. The same applies for frequency response anomalies, compressed dynamic range, and so on. If you take all of these parameters, I don't think we completely know which of them makes the human mind go "that sounded like a real voice". So, we know that jitter is orders of magnitude lower than wow/flutter, but what if we are orders of magnitude more sensitive to its effects, whatever they are? Just speculating...
I read something a while ago which made me think... you're walking by an apartment with an open window, and you hear a piano playing. You can usually tell if it's a real piano or a recording, even though you can't see inside the apartment. What is it about a stereo that makes it sound "not real"? Most of the information is presented correctly - you can differentiate a piano from a flute, for instance, and some people can even tell a Steinway from a Bosendorfer. You can identify the notes that are being played, you can set the volume to a realistic level. So... what is it that a stereo does not do right?
And to get somewhat back to the discussion... I personally prefer vinyl because it sounds more right to me. I fully know that the measured performance is much worse than CDs, and I completely do not care My measure of "accuracy" or "fidelity" is a totally subjective "does it sound like the real thing". So far, the closest I've come is a system I heard at a dealer - vinyl, single-ended tube amps, time- and phase-aligned full-range speakers. I don't know what roles those individual designs/technologies had to play, but the overall presentation was amazing.
Oops... put in 'quote' instead of 'url'

 

[Saurav]

Oversampling has removed the need for sharp filters at the output of the DAC's. That solution has been implemented for at least 15 years.

I thought a sharp filter was needed before the ADC to prevent alaising? I'm not sure how oversampling works... does the ADC sample at higher than 44.1kHz, and then the data is filtered to 22kHz in the digital domain? If not, then you still need a sharp filter during the digitization process, right?

 

[Craig_Kg]

Correct. Oversampling was introduced to reduce jitter - not ringing, so the sharp filtering remains (in the recording process).

 

[Scott H]

Not sure how one can quantify the effects of jitter.

Measurement of jitter is done in areas ranging from CD manufacture and duplication to fiber optic data transmission. In fact, high-end CD DAC manufacturers may list the crystal precision of their clocks (in ppm) and output jitter (in psec).

Btw, interconnects can induce jitter.

 

[chung]

I thought a sharp filter was needed before the ADC to prevent alaising? I'm not sure how oversampling works... does the ADC sample at higher than 44.1kHz, and then the data is filtered to 22kHz in the digital domain? If not, then you still need a sharp filter during the digitization process, right?

You need anti-alias filters before the signals are sampled, and after the reconstruction through DAC's. In both cases, oversampling is used to relax the difficulties of designing a steep filter with good phase response.

The output of a DAC has the spectrum of the desired signal, plus that signal modulating the harmonics of the sampling frequency. For instance, if a DAC is clocked at 44.1 KHz, the output will have the signal, plus the signal around 44.1KHz, and other harmonics of 44.1KHz. The signal is band-limited to 20 KHz, so the modulation at 44.1KHz can go as low as 44.1KHz-20KHz=24.1KHz, which is pretty close to 20KHz, the upper band-edge of the original signal.

So if your output samples are at 44.1KHz, you would need to have a filter that passes 20 KHz but highly attenuates 24.1KHz and above. Such filters are difficult to implement without causing phase distortion in the passband.

Oversampling now raises the aliasing signals to harmonics of the oversampling clock. In the classic case of 4X oversampling, the first harmonic is at 176.4KHz. It is much easier to design a lowpass filter that passes 20KHz and rejects 156.4KHz and above. That filter can now be a linear phase filter, with a -3dB frequency much higher than 20KHz. Ringing to a square wave input is now much reduced. The other advantage of oversampling is that the quantization noise is being spread over a much wider bandwidth, so that the total in-band noise is smaller.

The same principles are used in the sampler of the ADC. Oversampling there also makes the design of the anti-aliasing filter much easier. Also, since you can afford a very expensive filter before the ADC, phase distortion is less of a problem. One can also design a digital filter to equalize the phase response of the sampling filter.

The benefits of oversampling are well-known for many years, and have been applied universally nowadays. Interestingly, the early Philips CD players used 4X oversampling, because 16-bit DAC's were expensive. You can use a 14-bit DAC with 1/4 LSB errors, and employ oversampling to achieve the same results as if you had 16-bit DAC's.

I am not really sure if deviation from linear phase is a problem in audio. The tests I have seen indicate that human hearing is not that sensitive to the phase of the signal.

 

[chung]

Oversampling was introduced to reduce jitter - not ringing, so the sharp filtering remains (in the recording process).

No, oversampling is used to reduce noise, to relax the accuracy requirement of the DAC's, and to remove the need for a very steep filter and therfore making ringing much less. I don't believe oversampling has any effect on jitter.

In the CD player, the sampling clocks are phase-locked to a crystal reference. Whether you use 44.1KHz or multiples of that frequency, the frequency stability is the same.

 

[Scott H]

No, oversampling is used to reduce noise, to relax the accuracy requirement of the DAC's, and to remove the need for a very steep filter and therfore making ringing much less. I don't believe oversampling has any effect on jitter.

Oversampling was specifically implemented to address electronic jitter. Oversampling can not cure mechanical jitter however.