CD v.s LP - When comming from the same MASTER

In my system:

Sarah McLachlan Afterglow sounds ***much*** better on LP
Radiohead OK Computer sounds better on CD

I agree with AROC, once it goes to CD medium which is limited to 16/44.1, something is lost in the mix.

While it is true that the entire 24/96 digital master would not be fully transferred to the analog master, I would argue that the lost is not as significant in the analog domain.

I have a couple of digitally recorded classical LPs, which actually sound better than the analog recordings of that period, but when transferred to CD, sound lifeless. Whereas the LP version is a completely different story.

The sense of continuity in the music is better preserved in analog, whereas down sampling corrupts this flow and the music becomes disjointed.

Apart from the theoretical arguments about whether or not LP's mastered from digital sources sound better than CD's from the same sources, as a pratical matter, even if you have the "best" playback systems made, you have certainly set it the parameters for your "best". The variables available in turntables, IC's, cartridges, motors, power supplys, etc, far outstrip the variables in digital playback. Based on that conclusion, it is totally reasonable to me that Mr Porter would be able to tune his analog system to sound "better" than CD's, without regard to the software issues involved.

"Better", or "different"? That there are differences cannot be argued. Of course the two products were mixed down differently, perhaps even by different people. For the LP there must be RIAA equalization and blending of the LF material to mono (to avoid vertical groove modulation) and in most cases there will be peak limiting and compression so that the softer passages are out of the noise floor. (Remember, Albert, that not all people have $75,000 turntable systems that miraculously remove surface noise, and they buy most of the LPs).

So I conclude that there will be differences, and some will prefer the LP version.

Albert: We're talking about different things. You are saying that LPs usually sound better to you than CDs. A lot of audiophiles agree (including me, sometimes). What I am saying is that if you start from a 24/96 master, a CD will sound closer to the master than an LP. (You may still prefer the LP, however.)

Also, even when LPs are made from digital recordings, it's unlikely that the LP and the CD will be made from the same master. So you aren't making apples-to-apples comparisons of the two media, in any event.

Two comments, the D/A converters used in mastering houses (dCS, Prism, Apogee, EMM, etc.) are readily available to well heeled audiophiles. Vinyl is a pale comparison to the analog master tapes. If redbook digital sucks, then vinyl just sucks a little bit less.

I used to subscribe to that theory too, it's just didn't pan out that way when I actually listened to some digital vinyl. To me most digital vinyl just sounds more dynamic and natural than most CDs. Too bad I don't have duplicate copies on each format to compare. But you can't write off digital vinyl on theory alone. It sounds like NTSCDan has actually listened for himself, which is good. I respect that. To him the hightened noise floor in unacceptable. To me the increased naturalness of the digital vinyl is more acceptable that CD. We must listen for different things. Too bad we cannot afford the Meitner CDSD/DCC2 or Reimyo cdp-777. Then maybe I wouldn't need digital LPs of things I can find on CD for less.

Regardless of what's in theory. In the end, depends on how the process was done, either one can be better or worst than each format. Even the same album in either format can sound different from the same format. It is also depends on which press it came off. We had the experience with 2 different copies of Nora Jones album in LP format and they sound different. One sound better than CD and the other sound worst than CD. As for CD, it probably had less variation between the press but I don't know the answer to that.

So Ntscdan, you have NEVER heard an LP cut from a digital master that was better than CD?

Guess you don't have a copy of Alison Krauss, "Forget About It" because the vinyl KILLS the CD.

I too am a vinyl guy and have little use for digital. However, when the original event is recorded as digital and I have (only) choices of having that master converted DIRECTLY to analog at the mastering lab, or lossy reduced to redbook format and then converted BACK to analog at my home, I will take the studio transfer.

There is no way a CD has the same high resolution capability as the recording studio. The CD (redbook) format was settled upon over 20 years ago while modern recording studios continue to push the resolution of digital closer and closer to analog.

Digital is a "sampling" of the original, so the best possible digital would be "unlimited upsampling".

The only available "unlimited upsampling" is analog. These arguments continue because of differences in the resolution of different peoples system.

I also have a high end vinyl playback system and I beg to differ. The vinyl that I mistakenly purchased that is cut from digital masters is actually much worse than the CD equivalent because of the increased noise floor for one. Plus digital masters do not and never will have the warmth of analog masters which is the main problem. Cutting in onto vinyl doesn't solve the problem it just makes it worse. Unfortunately I think there are some vinyl folks who are so obsessed with the "superiority" of their kilobuck vinyl spinners that common sense goes out the window. Don't get me wrong, I consider myself a vinyl guy and have little use for digital, but for current recordings recorded and mastered on Protools etc. you are deluding yourself with the vinyl versions. The only caveat to my comments is that I do not listen to classical music. I think Stereophile or one of the other audio rags should do an objective story on this.

Pabelson, are you speaking from a theoretical point of view?

I ask, because had you made the comparison with a high end turntable and CD, you would not make that statement.

All of my recordings from digital masters are better on LP than CD, but then again I have one of the best turntables available. As long as there is an exception, it's wrong to state CD will ALWAYS come closer to the hi-rez digital master than LP.

If your claim is based on components in your system, that's a different matter.

If the master was something higher than 16/44.1 and a lot are 24 bit word length and up to 96k, then CD cannot capture all of that data, so in theory the LP will have the edge...

Except that the LP can't capture all of that data, either. In theory, it can capture some of the energy above 20kHz--but there isn't much, and very few people making LPs even try, for obvious reasons. As for dynamic range, LPs have a much higher noise floor and much lower dynamic range than CD, so the LP loses even more data than the CD does.

Bottom line, a CD will always come closer to a hi-rez digital master tape than an LP will.

Short answer the LP, period. If the master was something higher than 16/44.1 and a lot are 24 bit word length and up to 96k, then CD cannot capture all of that data, so in theory the LP will have the edge since it was converted to analog directly from up to 24/96, and the cd must first be down-sampled to 16/44.1k, then converted to analog.

I would have agreed with NTSCDan, but I listened with my ears, and lots of digital LPs do sound as good or better than CDs. This is with my focus on classical music. Electronic, amplified, rock, or pop music from digital masters might not far as well on LPs as pure acoustic music does! Sad but likely true!

I agree with Albert that I'd rather have the mastering lab to the D-to-A conversion for me.

I am missing something here. If the master was recorded digitally there is zero point in buying a vinyl pressing.
If you think I am kidding go and pick up a couple of 80s CBS Half Speed Masters. Some genius thought it would be a good idea to digitize an analog master and then cut vinyl from it. The results are dismal as is all vinyl cut from a DAT or other digi source. Why would you want to add all the inherent problems of vinyl to ones and zeros? As for the "shorter path to vinyl from a digital master" claptrap in my opinion. Anyone that buys vinyl mastered from a digital source is a sucker, as are all the chumps buying little wee files on itunes that aren't much better than AM radio. But that's another topic.

"It also helps to know that back left is usually where the bass drum is."

Not entirely true all the time. Some examples of the bass timpanis on the right rear of the soundstage are :

Beethoven Symphony No.1 Karajan BPO 1961/2 cycle on DGG
Beethoven Piano Concerto No.5 Stephen Bishop, Colin Davis, LSO on Philips

Metralla....To represent a time-domain waveform by Fourier coefficients, the waveform must be periodic, and remain of fixed spectral content (at least during the time interval of the FFT). Music isn't like that. There are momentary wavefronts with steepness equal to that of a sine wave at well over the 20 KHz upper frequency of human hearing.

There is an example that caused me to form an opinion and that was The Beatles Anthology series.The awful early recordings on Anthology 1 clearly sound better on the Vinyl version than on the CD version,where they sound especially harsh and strangled.Both the LP and CD were produced from the same digital master which they do with Beatles releases these days.In Stereophile Archives online in the "Think" pieces,John Atkinson has analysed and discussed what is on Analogue and Digital recordings.("What's Going On Up There")-extremely informative.L.P's are better.

Apologies - I should not have used the word "periodic" when referring to music in the above. It's slightly misleading.

Regards,

The Nyquist criteria calls for sampling to be at twice the highest frequency of interest: if this is true the analog waveform is represented without error. BUT...Nyquist was talking about sine waves. Music is not a sine wave.

That's not quite true. The Nyquist Theorem states:

To represent in the digital domain a signal containing frequency components up to FHz it is necessary to sample it at LEAST at 2F samples per second.

It is true that sampling at less than 1/2 the highest frequency gives rise to aliasing, creating components in the output after DAC that were not there to begin with. Obviously undesirable. But Nyquist used the phrase "at least" - sampling could be higher.

In theory, the waveform does not really matter. Music is a periodic waveform and the Fourier Theorem states that:

Any periodic waveform can be represented as a sum of harmonically related sine waves, each with a particular amplitude and phase ...

The mathematics is solid. Problems arise in the implementation.

Regards,

Twl...You make sense. Mostly.
The Nyquist criteria calls for sampling to be at twice the highest frequency of interest: if this is true the analog waveform is represented without error. BUT...Nyquist was talking about sine waves. Music is not a sine wave. That's why the CD sampling frequency of 44.1 KHz is not adequite. In my experience (non audio) sampling at about four times the highest frequency of interest was useful. (Higher a waste of time).

Analog recordings TRY to reproduce the signal in a continuous manner, but HF filtering gets in the way. A LP can be "read out" with an optical microscope instead of a phono pickup. A few exceptional recordings have groove modulations up to 22 KHz. If the wiggles are not in the vinyl you can't say that the audio system responds.

I agree that there are problems to be overcome with the analog playback system, and also the recording of records.

My take on this is that at least the analog system has the continuous waveforms recorded into it and can attempt to play them back, and the difficulties in this task are largely in the realm of the playback system to extract from the grooves. If the TT/arm/cart are up to it, almost all of the information can be brought into the system, albeit with some distortion caused by the recording process, and some caused by the playback system.

With the digital system, the sampling process is, by definition, not continuous, and as a result only a portion of the actual music is even recorded. So the many of the significant problems with digital actually occur at the master recording stage and the redbook conversion, with certain amounts of the music being left on the floor during the sampling process. Once this occurs the music is limited at that level, and cannot be recovered. So, even a top-quality esoteric digital player that makes a near perfect reproduction of the master is limited in information by the recording itself.

Maybe some feel that the Nyquist theory is valid and this is not an issue. In the gross sense, I think the Nyquist theory is valid. But not in the ultimate sense. There is a point where analog recordings can exceed the informational transfer of digital recordings, provided the analog recording is decent enough to begin with, and the analog playback equipment is decent enough to extract more information from the record than a digital recording can make. It is at this point(and above) where the analog medium begins to show its stuff. This is why people make the move to analog gear. If there was no difference, they wouldn't do it, because digital is clearly more convenient and more accessible in every way.

So, analog with all its warts(and there are many) can still provide a higher degree of music information transfer into the audio system than digital can. At the lower, more mass market levels, this may not be obvious because the analog gear is not precision enough to show the difference. At the higher levels, it is clearly obvious. I recently had a discussion with a person who has a listening buddy who has had some of the finest digital gear in existence, price no object. He currently has the latest Meitner gear, which is really, really good. He also has a very fine turntable(that cost about half what the Meitner digital gear did). The turntable exceeds the Meitner gear, and all of the other digital gear.

This could not be so, if the digital stuff fully captured the recording, and that is why I state that the digital recording processes are the limiting factor. I am fully aware that the digital stuff has a theoretical dynamic range advantage. I am aware that the background noise can be less with digital. I am aware of the "lossless" conversions concept. I know about the Nyquist theory. I know that digital has all kinds of advantages at the consumer end. And I accept all of that. But the bottom line is that it is limited by the very sampling technology that makes digital possible in the first place. It does not record the full waveforms. And when you reach the higher levels of performance, the advantages of analog(in music information transfer) show themselves.

So, what does this have to do with the main topic at hand. Just this: that the limitations of digital are primarily in the recording stage, and the limitations of analog are primarily in the playback stage. To refer back to the original question, I submit that the digital master will be more limited by the recording, the result will be that the advantages of analog playback will not be as well able to show themselves, compared to using an analog master. Therefore the differences between the CD and the LP from a digital master will be less distinct.

I know that this may be somewhat controversial, but I'm sticking to it.

An afterthought...Some of the very first stereo LPs, Audio Fidelity "Dukes of Dixieland" did not sum the bass, at least that I can detect on my system. Some of the very first stereo recordings (Vanguard in particular) also stuck with two microphones, and that also was better than early attempts at multitrack recording. Sometimes progress goes the wrong way.

Pabelson and others...Summing of the bass (like the drastic RIAA equalization) is a topic that vinyl fans like to ignore. I believe that the reason for suming is to enable non-audiophile phono pickups to stay in the groove. Large vertical groove modulation would raise havoc with VTF, and even cause the stylus to hop out of the groove. I think that the summing is rather gradual in slope, but begins around 200 Hz. Whether this is a problem or not is another thread. Many people sum the bass to a single subwoofer. (Not me).

Horizontal groove modulation is what needs to be limited so as to save space on the vinyl. This is done, first by the RIAA equalization, and secondly by variable groove spacing (according to the momentary LF signal content).

I think, it can't be answered in general.
The Problem with CD has nothing to do with the digital Master in general, most differences are made through the next processes. High speed burned or low speed burned, with or without reclocked burners and so on. It is more or less endless ( check out all these "remastered" issues, sick ).
Some CD's sound very good, some not and I think, it is the same with Vinyl, which has it's source from a Digital Master. Some CAN sound very good and some don't, depends on the Mastering Lab I guess.
Anyway, here I go for the CD, when I am not sure about the result, then I don't waste time on it.
( You know, I can freeze it, or paint the outer ring with a text marker or I can cut the outside for a better angle or ..... )

I don't quite follow the point about the necessity of summing deep bass to 'fit it on a record'. A full wavelength at 20Hz on a 33 1/3 RPM record traces 10 degrees. Put another way, you can fit 36 wavelengths of a 20Hz signal around a groove. Together with RIAA deemphasis, ~20db at 20Hz, this wouldn't seem to be a problem. What I don't know is what physical groove modulation is needed for a 0db level at 20Hz. Perhaps you're right.

At what frequency do they start summing to mono for LPs?

This of course completely begs the question of the necessity for stereo bass that low, unless of course your listening room is the size of a football stadium.

Sean: I'm not arguing that CDs can theoretically reproduce sound perfectly. They can't. I'm arguing that, practically, they come close enough. Yes, there is jitter, but all the research I've seen suggests that the level of jitter in a typical (and I do not mean high-end) CD player is an order of magnitude below hearing thresholds. DACs and anti-aliasing filters have improved since the early days, and I haven't heard or heard of a bad one in the last 15 years--except at the high end, where occasionally a designer is too busy being innovative to get the basics right.

The proof of this is in the listening. Remember that this thread started from the premise of a digital master tape. I would argue that you would have to listen very hard to discern a real difference between that master tape and a CD played on a $100 DVD player. Whereas it would be relatively easy to discern a difference between master tape and a $10,000 vinyl rig. That tells me that the implementation of digital, while still not perfect, is pretty darn good.

And, to get back to the original point, that means it's probably not the flaws in digital, such as they are, that are responsible for the fact that many listeners prefer the vinyl version. It's more likely that the technical flaws of vinyl are, counterintuitively, part of what makes that medium sonically appealing.

Cmk: I mentioned summing bass as an example of the alterations (distortions, if you will) necessary to cut a vinyl record. I agree that it is not audibly significant in most cases. By the way, the reason bass often sounds directional is because we aren't listening to pure tones, and sonic components that are higher in frequency can provide directional cues. It also helps to know that back left is usually where the bass drum is.

I think the point missed here is the fact that regardless of which format (digital or analog) the master is recorded, the mix down to the end user takes a VERY different path, depending on which format were talking about.

If the event must begin as digital, I sill prefer high resolution digital converted directly to analog by the mastering lab. The conversion is one step and there is no deliberate "lossy" re sampling as with Redbook and SACD.

I have thousands of LP's, most are analog source. Music released in the last decade are all over the place in terms of what the recording studio used. I suspect the latest releases such as Alison Krauss (one of my favorites) was recorded to a hard drive rather than tape. That particular recording is the work of Doug Sax, so likely the master to analog was done in the most direct way possible.

I have no doubt that any one of us with the master digital hard drive (or tape) in our own system, would hear performance above and beyond anything we have ever experienced.

The argument comes from differences in opinion as to what quality REMAINS after Sony and similar vendors convert that original digital format DOWN to comply with Redbook or SACD standards.

In my opinion, the studio conversion from the best available digital format, direct to analog is still the best representation of the original event. This is the shortest path to analog that we all experience when we listen.

This is a great topic, and something to seriously think about for anyone who is interested in getting into vinyl. Nowdays it seems like the analog 1/2" master is becoming extinct, so therefore it seems logical to think with an LP copy of even a hi rez 24/192 source would only ever sound possibly as good as the SACD or DVD audio of that same source. This is a very thought provoking topic, unless you already have a nice colletion of vinyl or are interested only in picking up older stuff why would you bother with vinyl playback at all.

Pabelson, actually the real impact of sampling at a higher frequency is not to extend the frequency response, but to make the implimentation of the anti-aliasing filter simpler, so that a filter with less passband ripple and less phase error can be used.

I agree with you that extra high frequencies above 20kHz do not make a difference.

I think this goes to the heart of your arguments ... in THEORY redbook CD is perfectly capable of producing perfect sound. Provided there is no jitter, the ADCs and DACs are perfectly linear, and most importantly, that you can implement a perfect brick wall anti-aliasing filter in the frequency domain, which creates the sin(x)/x function in the time domain to perfectly reproduce the analog waveform from the train of samples.

It is in the implementation that the CD playback falls short.

I agree with you that vinyl is a deeply flawed format, and I personally think that vinyl and CD replay are roughly on a par, with different deficiencies. I suspect that those who prefer vinyl are in some way more bothered by the distortions caused by the deficiencies in real-world digital recording and playback caused by timing jitter, non-linear DACs and ADCs and approximations to the brick-wall filter.

That vinyl should be considered "accurate" by the same people who argue endlessly about how different cartridges can sound at different tracking weights, how different tonearms sound with a different counterweights is equally ludicrous. But vinyl does have moments where it sounds more real than CDs, so I guess I kind of like the distortion introduced by the medium.

Well Pabelson, from real life listening sessions, LP has always produced superior soundstaging AND bass separation. While from a technical standpoint, what you say about suming the deep bass on an LP maybe true, but below roughly 80hz bass is non-directional, so having true stereo bass all the way down to 20hz is just academic, maybe good for the technician in the studio, but with less real life applicability than made out to be.

Take the Telarc digital recording of the Cleveland Symphonic Winds conducted by Frederick Fennell - Holst suite No.1 and 2. On the LP the bass drums hits are definitely on the far L back of the soundstage, and they do go deep, maybe not down to 20hz, but certainly below 40hz, enough to shake my floor and sofa. And the bass isn't the one note type of bass, but its quick, well-defined and full of impact.

Now I also have the Telarc CD version of this excellent recording, and let me assure you, the LP version sounds significantly better, its not funny.

To 'hear' sound above 20KHz has nothing to do with believing.

Here is a very interesting link to this topic :

http://www.cco.caltech.edu/~boyk/spectra/spectra.htm

Styx

In my experience LPs from a digital masters are usually notably better sounding than the CD. There are exceptions like Nora Jones first release where the LP sounds no better than the CD. I have a great sounding Phillips classical CD where I found the corresponding digital LP. What a dissapointment. The LP was considerably worse.

I think that there is more to the story than just differences in the medium. The path between the Master and the signal that feeds into your preamp is a long one with plenty of opportunity for distortion and grunge, regardless of the medium (LP, CD, SACD...).

From the original master recording (digital or otherwise) separate CD and LP masters are created. They include the obvious stuff like separating into tracks with fade in/out and inter track silence. At this step it is typical for bunch of equalization and compression to be done. The munging of the signal at this point is customized for the medium and how good or bad depends on the skill and ears of the mastering engineer. In addtion the quality of the equipment used for the mastering process may be different for different mediums.

There is a lot of nonsense about digital being a perfect medium without distortions. It is true that digital information can be replicated and transferred without degredation or distortion. However, distortions unavoidably introduced for both the capture (A to D) and the playback (D to A) conversions. Distortions introduced from a digital process are different than those in analog. But it is nonsense to assert that digital does not have distortions. The distortions from digital are more difficult to measure but I would assert that they are easier to hear and more harmful to musical content.

So getting back to the topic LP and CD are two different methods of recreating the signal from the original recording. Both create distortions. Those distortions are different and to my ears LP provides the playback that is the most true to the original.

LPs made from digital masters are no doubt compromised but I find that they still sound quite good and often benefit from analog on the playback side.

When the master is digital, the inherent advantages of the analog recording is not there, and therefore I see no advantage to the LP in that circumstance.

The sound was already limited at the digital master stage, so it can't be made "better" later.

No, Cmk, it has nothing to do with set-up. There isn't room on an LP for two channels reaching all the way to 20 Hz--at least not if you want more than a few minutes of music per side. So it is standard practice to sum deep bass to mono. Same thing with dynamic range. It's inherent in the medium.

I disagree with Pabelson's assessment of the LP having "no stereo bass, limited dynamic range". Quite possibly the LP playback system is not properly setup if one experiences these problems.

Sorry, Sean, but I must disagree. A 24/96 master tape does indeed have higher resolution, but on a practical level that means only two things: greater dynamic range and wider frequency range. The latter is only important if you believe that people can hear sounds above 20kHz. As for dynamic range, I've never heard of a master tape with a dynamic range greater than 90dB, which CD can accommodate just fine, thank you. And even if you find such a master tape, you'd have to compress it MUCH more to make an LP, which can only offer about 70-75 dB of dynamic range.

As for the sonic differences, they are not subtle at all at any price range. I'd say the difference between a CD player and a turntable at price X will be greater than the difference between CD player at price X and CD player at price 10X.

None of this is meant to bash vinyl, which I love as much as the next audiophile. I'm not arguing that one medium is better or worse than the other--just different. Better or worse is up to your own taste.

"A CD should sound exactly identical to that digital master". Incorrect, because nobody masters at redbook level anymore. It's all done with higher sampling rates and greater number of bits, and then it's downsampled (decimated more correctly) for redbook. So the digital master contains much information that is not on the CD.

But for the record (no pun !) I think on similarly priced turntable / CD player the differences would be very subtle unless you had a very expensive system.

Just to clarify, I was talking about starting from a digital master. A CD should sound exactly identical to that digital master. An LP won't, and the difference is, quite literally, distortion. (And I like the sound of some of those distortions myself.)

There is no comparison in sound quality from an MFSL lp and an MFSL cd. I'm sure the quality of the source being used would make a difference. Vinyl beats digital hands down in my system. Maybe I'm one of those who thinks that the degradations and distortions sound good. Then again I believe vinyl has no equal in sound quality...........

The LP version would suffer all the degradations and distortions that are inherent in that medium--no stereo bass, limited dynamic range, surface noise, etc. But some of those distortions actually sound good to people, so it's quite possible that the LP will sound better.