The truth about high-resolution audio compared with std. CD 44k?

[Diminish]

@ STHLS5: "Diminish, I have the earlier edition of Pohlmann but unable to find the info about 4.3218Mbps for EFM coding."

C'mon, I'm not making this stuff up! One place it can be found is in the chapter entitled "The Compact Disc", p.244 in the 4th Edition. "Each audio disc stores a stereo audio signal comprised of two 16 bit words sampled at 44.1 kHz; thus 1.4 million bits / second of audio data are output from the player. Other data overhead such as error correction, synchronization, and modulation are required. Altogether the channel bit rate, the rate at which data is read from the disc, is 4.3218Mbps."

The figures given by HUG-1 simply were not correct. The 150,000 Bps figure for CD ROM is in Bytes (not bits) per second. Internet connections as well as CD players are concerned with bits per second. More notably, the Yellow Book standards for CD ROM are not particularly relevant to a discussion on Redbook vs. hi-res audio.

[STHLS5]

@ STHLS5: "Diminish, I have the earlier edition of Pohlmann but unable to find the info about 4.3218Mbps for EFM coding."

C'mon, I'm not making this stuff up!..
...

Thanks. I found it. It was never my intention to suggest you are making up those figures.

You mentioned earlier “ the disc itself is read at 4.3218Mbps............this is well over 4 times the data rate of CAT5 network”. I am only interested to that statement and not about the general discussion about CD vs Hi-Rez.

In the wikipedia (Redbook standards) it is stated that the raw bitrate at optical pick up is 4.3218Mbps but it is not exposed to the application reading the disc. The actual data read is 2048 bytes per sector x 75 sectors per second = 150KiBytes/s (1228.8 kbit/s).

Even if we agree that the transport reads 4.3128Mbps, it is still about 0.54MB/s which is far below than the earlier 10base Ethernet capabilities. So how can it be over 4 times the data rate of CAT5 network?

(Note: 8 bits (b)= 1 byte(B))

ST

[Diminish]

@ ST "In the wikipedia (Redbook standards) it is stated that the raw bitrate at optical pick up is 4.3218Mbps but it is not exposed to the application reading the disc. The actual data read is 2048 bytes per sector x 75 sectors per second = 150KiBytes/s (1228.8 kbit/s)."

I'm not sure what you mean by the distinction "not exposed to the application reading the disc". The optical pickup is what reads the disc, correct? If you intend the DAC section as the "application reading" the data and define the data to be read as the SPDIF audio stream, I agree completely. I've said several times now that the transport is concerned with much larger amounts of data than the DAC section. This, along with the fact that this process exists in mechanical, optical, and electronic domains, represents the entire basis for my argument. Your Wikipedia article says "By comparison the bit rate of a "1x" data CD is defined as ... (1228.8 kbits/s)" The 150KBytes/sec. figure was also for the data CD. So, the difference is that the data CD has a slightly slower data rate (1411.2kbit/s versus 1228.8kbit/s), so what?

Now, in terms of the comparision to networking. I misread HUG-1's 1,000,000,000 bits/s figure, and regarded it at 1,000,000 bit/s. Since large bit / byte rates are normally given with Kilo or Mega prefixes, I misinterpreted his unit by a factor of 10 which lead to an improper conclusion. I was completely wrong in the statement that a CD transport's data rate exceeds that of such a network, however, I don't see where this nullifies my baseline contention.

One major difference between a modem and a CD player is the importance of timing. For accurate playback, not only must the DAC decode the correct value, but it must do this at precisely the right time. Any deviation from this is known as "clock jitter". This is a difficult thing to measure because the time intervals are so minuscule. There are also different types of jitter (random and deterministic are two types) with deterministic jitter being the more sonically detrimental. Disparities in the pico second rage (1/1000000000000 of a second) can be audible.

Internet protocol involves data being sent in discrete packets where timing is much less crucial. Another reason why this is not a fair comparison is the method of data transfer. CAT5 cables carry parallel data where bytes of data can be transferred through many different paths. Generally speaking, the more channels available for transmitting data, the faster the transfer and greater the throughput. The SPDIF specification is for serial transfer over a 75 Ohm coax.

I am not and was not contending that a CD transport represents the largest, or fastest method of data transfer. Neither am I saying that it is state of the art in 2011. One has only to look at a Blueray mechanism to see far greater data capabilities. For the last time, the point was that; due to the potential for errors in the mechanical, electrical, and optical domains and the higher data rate present in the pickup, that a transport would have more potential to effect the overall sound quality than the player's DAC. I don't know that there's much to be gained in exploring it any further.

Personally, I would much rather resume the conversation about Redbook vs. Hi-Res. Along those lines, I purchased some tracks from AIX Records (iTrax) yesterday and was very impressed with the quality of sound of the Jazz and Folk tracks. I also got 6 tracks from their Classical catalog, but I was somewhat less than impressed by these.
All of the tracks were 24 bit 96kHz Hi-Res. LPCM. I bought them as WMA Lossless files, but its also possible to buy the entire CD Audio files. You can get them in 2 channel or 5.1 surround from a Stage or Audience perspective. I only concern myself with 2 channel. I converted the WMA's to FLAC to be consistent with the rest of my Library. Prior to deleting the WMA's, I listened for differences in the FLAC files and heard absolute nothing. I also burned a CD copy which effectively "downsamples" them to Redbook. In order to do a comparison, I have to listen to the Hi-Res. from my computer, and compare that to the 16/44 from my transport through the same DAC. I could downsample in JRMC or rip the CD's back into the Library for a closer comparison, but I think that the above method is sufficient.

You could argue that the JRMC software is imposing a sonic characteristic during encoding, decoding, importing, burning. Using my CD player as a transport would certainly allow for some differences. I use a Sonicweld Diverter 24/96 USB>SDIF converter for computer playback. This is a high quality device that may well yeild a better (less jittery) SPDIF output than my Marantz R15S2. That notwithstanding, the sound is quite good on the Jazz and Folk tracks through CD or JRMC. I hear the same drawbacks to the Classical tracks through either medium, but the Clarinet's texture becomes a bit harder on CD. The soundstage is a bit better defined on the Hi-Res. material and perhaps a bit deeper. Comparing from memory, in a broad sense, I would say that the sound offered by the best of the AIX tracks exceeds the best I've heard from CD.

{Moderator's comment: noted about the hires tracks and Alan's demonstration recently of 'spaciousness'. What did you make of that demo?}

[EricW]

In the wikipedia (Redbook standards) it is stated that the raw bitrate at optical pick up is 4.3218Mbps but it is not exposed to the application reading the disc. The actual data read is 2048 bytes per sector x 75 sectors per second = 150KiBytes/s (1228.8 kbit/s).

ST

Actually, according to the Wikipedia article you cite, the 2048 bytes per sector figure is for a data CD (CR-ROM). The figure for an audio CD - which I think is what this discussion is about - is 2352 bytes per sector. The 75 sectors per second figure is the same.

[STHLS5]

Hi Diminish, thanks for clearing up the confusion. Please carry on with the discussion about Hi-Rez but I have reached a point that my ears only good enough for CDs only. Though, I prefer to buy SACD, XRCD, HQ etc etc I believe their superiority is in recording or mastering techniques and not in the format. I am even contended with MP3 nowadays after realizing some oldies that I was listening were actually in MP3 and I didn't even realize that.

I just learned how to use my laptop as a music server using the ONkyo NDS-1 as an interface to my legacy DAC Theta Digital GenIII and my young children didn't even notice any difference until I began side by side comparison. And speaking of hi-Rez, it is only recently I found out my SACD Marantz Reference SA11S2 is using the the good old 44.1kHz 16 bit to play Cd. Even though, the industry standard for most high end player is 24bit and 196khz DAC, Marantz engineers thought it was okay to stick to the old 16bit/44.1Khz and it was positively reviewed by Stereophile.

Actually, according to the Wikipedia article you cite, the 2048 bytes per sector figure is for a data CD (CR-ROM). The figure for an audio CD - which I think is what this discussion is about - is 2352 bytes per sector. The 75 sectors per second figure is the same.

Yes, you are correct but the 256 bytes of CRC and another extra 48 bytes do not contain any data. So technically only 2048 is important for the discussion above in regards to data transferred.

I am no expert in digital audio so any further arguments by us about what is the actual data stored, read or transferred would be like the story of the blind men and the elephant.

ST.

[Kumar Kane]

When one talks of high resolution, the impact is very marked in video, with the difference between hidef and standard very clearly visible to anyone that sees it, trained eyes or not, be it Bluray DVD or HDTV via satellite. And video seems to be making further leaps, with 3D. Although even the current solutions are clunky ( glasses required ), I guess progress will continue, to being able to see realistic 3D as easily as one sees 2D today.

That certainly does not seem to be the case with audio for music listening with no video involved, which seems to be at the end of the road. What, I wonder would be an equivalent leap? There has been much progress from a convenience standpoint with digital audio, music servers, internet music services, multi room playback and the like, but my question is towards the quality side of things, where there doesn't seem to have been any real progress for many years now.

[Diminish]

"{Moderator's comment: noted about the hires tracks and Alan's demonstration recently of 'spaciousness'. What did you make of that demo?}"

Hi, since you asked, I went ahead and listened to the Rach. tracks. There were some notable differences that I picked up on: the SD track sounded more distant, or another way of saying this, would be that it was more "laid back". At times, it sounded flat, but when the full range of the piano was invoked, I had to question that. The wind instrument (flute?) that comes in during the final seconds seemed further from the piano than in the HD.

In the HD track, I thought the piano sounded closer; a more "forward" perspective. The most significant difference was longer reverb. trails ("note decay") in the HD recording. The wind instrument appeared closer to the piano. If you take that along with the observation that the flute sounded farther away in the SD, and the forward perspective noted on the HD compared to the laid back SD track; it would certainly seem that I'm saying that the SD version conveyed more spaciousness.

Well, that's not exactly what I mean to say. I heard more "air" in the HD recording although it was closer in perspective. I also heard more depth, or 3 dimensionality, in the HD although the SD seemed placed further back. Whatever the case, I was able to readily distinguish between the two during blind trials.

Am I correct that what we are listening to two 192kbs tracks; one taken from a 24/88 Hi Res. and the other a standard Redbook? Honestly, I would have expected the lossy conversion to mask most of the details that were preserved in the 192kbs tracks.

When I do these comparisions at home, I'm able to directly compare Hi Res. material with Redbook by routing both through the same software, USB>SPDIF converter, coax cable, and DAC. Most of the time, I hear less disparity than what was revealed by the 2 examples here. With one notable exception, I have yet to hear a true "Night and Day" difference.

I recently purchased YES's "Fragile" from HD Tracks. If I compare one of the songs on that album with the same song from another album, Classic YES has some of the same songs. There is a huge difference. I never thought that Classic YES was very well recorded to begin with, but they are a hard band to capture. Comparing Long Distance Runaround, there simply is no comparison. Its impossible to focus on one specific area where the 24/96 is better; it should be obvious to anyone; even audiophobes and music haters. I'm not sure that these are the same recording, and therefore that the only difference is the resolution. Long Distance Runaround is the same duration on both albums, but the recording volume is lower on Classic YES. This could be due to the 8 extra bits.

Generally speaking, what I normally hear in terms of differences is most notable on classical recordings. Most CD's portray massed strings as a single big instrument rather than a group of individual instruments playing together. For some analog buffs, any digital reproduction of stringed instruments is anathema. While still not to the level of good analog, Hi Res does a better job than Redbook for sure. SACD is even better than the PCM formats at this particular task. In general treble has less offensive glare or hardness, bass is better defined in terms of pitch, and the midrange somewhat more compelling.

There are many experts who feel that DSD (the SACD coding) masters produce better Redbook sound. I also think that record labels who produce Hi Res. digital care more about sound reproduction. Much less, if any, dynamic compression takes place.

Significant thought is given to microphone placement and less mixing, and for the most part no overdubbing takes place. A big advantage of higher bit word lengths is a greater dynamic range and more precision within this range. Higher sampling rates capture more sonic information and allow for more low pass gradual filters.

[Pluto]

That certainly does not seem to be the case with audio for music listening with no video involved, which seems to be at the end of the road. What, I wonder would be an equivalent leap?

Ambisonics starts to bridge the gap between recorded sound being a mere photograph, and reality.

Loudspeaker manufacturers would love it too :-)

Sadly, the required positioning of the loudspeakers is prohibitive for all but the most devoted.

[kittykat]

It's hardly worthwhile to argue about the merits of the CD format any more since it won't be with us for much longer...

That’s quite a big call.

Looking at Billboard’s market watch stats (24th Sept edition), 212 million albums were sold in the last 12 months, of which 140 million (66%) were in CD and 32% (68 million) in digital. Digital album sales climbed 20% y-o-y. At this rate, digital album sales would be greater than CD album sales sometime between 2014 and 2017. At the current trajectory of 4% CD album sales decline y-o-y, CD sales could still top a hundred million per year in 2020 (and that’s almost 40 years after its introduction).

Imo, what makes most people think the “CD is dead” is due to digital track sales. 889 million tracks (or 301 million equivalent albums) were sold last year. If we add digital tracks to digital album sales, digital sells 2.5 times more than the physical. The demand for albums (and CD) should however not just drop off, especially if 47% of album sales are for catalog and approximately 53% for current music. What the figures don’t report is the proportion of catalog and deep catolog by format. I suspect most catalog sales is in CD format.

As long as there are still people who enjoy music and are pupils to it, enjoy artwork, the smell of print and the tactility of touching an album, CD’s (or any physicals) could still be around for a while.

Your Myanmar taxi driver could be in this group and if he is, I’m very sure he kept his CD copy safely at home.

[HUG-1]

"{Moderator's comment: noted about the hires tracks and Alan's demonstration recently of 'spaciousness'. What did you make of that demo?}"

Hi, since you asked, I went ahead and listened to the Rach. tracks. There were some notable differences that I picked up on: the SD track sounded more distant, or another way of saying this, would be that it was more "laid back". At times, it sounded flat...

Last chance to comment on these test tracks before we look at them technically ..................

Just as with all matters concerning audio there are no "right" answers and no "wrong" answers. Everything is in the grey area between and a matter of personal perception.

[Pluto]

Last chance to comment

Listening only on naff computer speakers -

HD has more stereo width, possibly mainly at the lower signal levels. Technically speaking, only the louder elements of the 'S' component ('S' = width for these purposes) are making it through the codec. As the hall acoustic exists mainly within the 'S' component, the narrower (SD) example exhibits less of the hall acoustic.

[A.S.]

Listening only on naff computer speakers -

HD has more stereo width, possibly mainly at the lower signal levels. Technically speaking, only the louder elements of the 'S' component ('S' = width for these purposes) are making it through the codec. As the hall acoustic exists mainly within the 'S' component, the narrower (SD) example exhibits less of the hall acoustic.

... the professional (sound engineer) listener comments!

One point I should have made long ago is that I only generate and test these audio clips on my own $30 plastic pc speakers. I do not play them through quality speakers or hi-fi system because I am not intent on revealing some extremely subtle nuance but a general point. And I do that by allowing a side-by-side comparison to be made by the listener And for that, your plastic PC speakers should be perfectly adequate. For about $30, Logitec (a brand that makes a wide range of well engineered pc accessories) offer several speakers: these here used to make the clips are their LS11.*

Turning to these two clips first presented in post #32 ...

Loading the player ...

SD example

Loading the player ...

HD example

The listener's feedback was as I expected and is neatly summarised by Pluto above. 'The HD recording has more stereo width'. Assuming that we agree on that (listen again to the repeated clips above) we can pursue why this is. Without any further background information, you and I casually listening could draw the conclusion that the HD recording brought some additional realism, a greater 'being there' experience. And probably we would cheerfully cough-up an extra few (or many?) dollars for that masterpiece recording even if we already owned the 'SD' recording. That is how the movie and recording industry works; they are very creative in selling you (again) what you already have. Be it VHS>DVD or DVD>BluRay, or LP>CD then CD>Remastered CD. It's a game.

In my post #32 introducing these clips and running with the flow of the discussion I said -

... neither you nor I, as ordinary members of the public, have (any) awareness of the mysterious process of 'mastering' these recordings. We would be barred from the CD mastering facility. Their tricks and techniques are closely guarded commercial secrets. Their secret business is of increasing the appeal of recorded music. The more they sell the better so 'mastering' is an absolutely crucial step in the commercial process of bringing music from the recording studio to our homes. 'Whatever it takes' is their motto.

Now the sad fact is that every commercial recording once 'in the can' is frozen in time at the day it was captured, with those performers, those microphones and with the limitations of recording skill and technical equipment on the day. Sounds not 'on' the recording are not on the recording - period. The recording is as complete and sealed as an oil painting. Mono cannot be regenerated as surround sound in post-production. But for marketing purposes, ways have to be found to 'leverage' the significant cost invested in the recording to find new consumers for it. And since music is a highly personal subjective matter, the most likely new market for a re-release is those that have 'already expressed a preference for' (marketing speak) the artist. And so, the crucial importance of remastering what was sealed at a point in time into something new, fresh and sellable.

The truth is that both of my recordings clips started out as exactly the same. I simply took the original high quality recording and duplicated it. The original I labelled 'SD' and the copy I manipulated - I remastered - and labelled it 'HD'. With a little appreciation of listener preference behaviour (i.e. the sort of things that most people would find audibly attractive most of the time) it took just a few minutes to remaster the clip. I'm sure you have the tools to make the same transformation on your PC: you can remaster your entire audio collection for free! More spaciousness, more open, greater dynamics - more revealing, a greater sense of scale etc. etc. etc. .... all there at the click of a few keys.

What did I do? Can you do it yourself? What can we learn from this?



*The sound quality of active plastic pc speakers can be fairly good. The usual problem though is hum, both mechanical and electrical.

[Kumar Kane]

The truth is that both of my recordings clips started out as exactly the same. I simply took the original high quality recording and duplicated it. The original I labelled 'SD' and the copy I manipulated - I remastered - and labelled it 'HD'. With a little appreciation of listener preference behaviour (i.e. the sort of things that most people would find audibly attractive most of the time) it took just a few minutes to remaster the clip. I'm sure you have the tools to make the same transformation on your PC: you can remaster your entire audio collection for free!

Why am I not surprised?!:-)) I was expecting something on these lines, and I now look forward to reactions....all of this is like photoshop for audio.

[Pluto]

What can we learn from this?

We can learn to be more forthright! I has assumed that we were comparing the Radio 3 "hi-def" AAC feed with the lower quality WMA feed.

When I played these (quite late at night), I had a feeling that the "wider" one felt a bit too wide - width added by increasing the proportion of the 'S' component, i.e. artificial!

Had I spoken my mind (instead of thinking "time for bed") I would have had our evil moderator bang to rights :-)

That said, on the whole topic of remastering, I have a few "remastered" CDs that are vastly better than the original releases, the quality of the converters having come on in leaps and bounds between then and now.

I have also heard several remasters that suffer badly, in the latter releases, from excess limiting. "Loudness war" stuff.

[A. E.]

*The sound quality of active plastic pc speakers can be fairly good. The usual problem though is hum, both mechanical and electrical.

I do understand now, why some ladies are happy with kitchen radio

[A. E.]

What did I do? Can you do it yourself? What can we learn from this?

Dear Alan,

This shows me I can enjoy all my digital archive after a little touch, regardless of the other components like speaker stands, cables, pre and power amplifiers...

I could photoshop my raw photographs in the very first day, but it took me about a year to learn to get what I want exactly.

I don't understand how you did what you did, at the moment; but I want to know how to do it by myself...

[Kumar Kane]

What did I do? Can you do it yourself? What can we learn from this?

Hi fi = high fidelity. Doesn't all of this now beg the question, fidelity to what? How is one to know just what is the reference any more?

[A. E.]

Hi fi = high fidelity. Doesn't all of this now beg the question, fidelity to what? How is one to know just what is the reference any more?

who will choose the "reference" while everybody hears and perceives differently... I'm not sure even my L and R ears hear the very same... There should be more than one reference fidelity, tailor made for everyone...

On the other hand, time changes everything... When I was new at Photoshoping, all of my images were sharpened extremely and the colors were oversaturated, but just the opposite now. So even my own reference changes for getting further by time...

[jplaurel]

No surprise here. As I said in my earlier response:

"I still believe that it would be possible to achieve the "larger room" sound of the HD file during the mastering process with a lower bitrate. For all we know, Alan's two recordings may be identical except for some DSP effects."

What have we learned? That much of the music from HDTracks sounds better than the original CD versions because of the remastering rather than the increased bit depth or higher sampling rates. Maybe it's possible for some very young people to hear differences in bit depth and sampling rates beyond Redbook, but at 50 years old, I can't.

Quick aside about the acute hearing of young people: Years ago, when my son was still in high (secondary) school, we were talking about kids having mobile phones in class as his school had recently implemented a policy requiring that they be turned off. He laughed off the policy, stating that phones are ringing in class "all the time" and the teachers have no clue. The kids' solution to the mobile policy was novel. He hit a few keys on his phone and said "can you hear that"? I couldn't. It turns out that the kids didn't turn their ringers off at all. They merely switched to high frequency ringtones above the threshold of the teachers' hearing. A very simple and clever solution.

The truth is that both of my recordings clips started out as exactly the same. I simply took the original high quality recording and duplicated it. The original I labelled 'SD' and the copy I manipulated - I remastered - and labelled it 'HD'. With a little appreciation of listener preference behaviour (i.e. the sort of things that most people would find audibly attractive most of the time) it took just a few minutes to remaster the clip. I'm sure you have the tools to make the same transformation on your PC: you can remaster your entire audio collection for free! More spaciousness, more open, greater dynamics - more revealing, a greater sense of scale etc. etc. etc. .... all there at the click of a few keys.

What did I do? Can you do it yourself? What can we learn from this?

*The sound quality of active plastic pc speakers can be fairly good. The usual problem though is hum, both mechanical and electrical.

[jplaurel]

*The sound quality of active plastic pc speakers can be fairly good. The usual problem though is hum, both mechanical and electrical.

@Alan: I hear what you're saying about plastic desktop speakers. I once had a pair of the Harmon Kardon Soundsticks with their integrated DAC and miniature 4-driver line array and they sounded astonishingly good. And back when my work environment dictated that I not play music too loudly in consideration of my neighbor in the next office, I had a pair of Bose Computer MusicMonitors. These were active speakers with these teeny-tiny drivers and some sort of passive radiator on the back. They were actually pretty good. Imaging in particular was surprising.

{Moderator's comment: did they like most PC speakers emit a constant hummmmmmmm?}