Okay, thanks for the input. May I bother you with one more question? The data that's read into the buffer must be clocked out, and the disc kept in sync with the data rate. Is the disc drive's motor kept in sync by the word clock? I thought the word clock was embedded in the data on the disc. Thanks again for your time, Ethan. Sorry if I briefly hijacked this thread, everyone.

Quote:
The data that's read into the buffer must be clocked out, and the disc kept in sync with the data rate. Is the disc drive's motor kept in sync by the word clock?

I'm not an expert with the low-level details, just the basics. But I'm pretty sure the motor and data clock are not tied directly like that. With disk drives generally, there are a few layers involved. With an operating system like DOS or Windows, the OS tells the drive to read a block of data starting at a particular sector, and put the data into memory starting at a specific address. The drive finds that sector, reads the data, and puts it in memory. If a disk is scratched or whatever and the drive has to read the data again or reconstruct parts of it, that is done "behind the scenes" from the software's perspective. I assume the same happens with CD players, where the player's firmware requests data from the drive at a specific sector. Once the data is in memory, then the player sends it out at precisely 44.1 KHz. At that point, the drive is already reading the next block and getting that ready so the blocks are all sent without interruption.

Quote:
I thought the word clock was embedded in the data on the disc.

No, the word clock is a separate crystal oscillator that sends the data from memory out to the D/A converter.

--Ethan

Quote:
Okay, thanks for the input. May I bother you with one more question? The data that's read into the buffer must be clocked out, and the disc kept in sync with the data rate. Is the disc drive's motor kept in sync by the word clock? I thought the word clock was embedded in the data on the disc. Thanks again for your time, Ethan. Sorry if I briefly hijacked this thread, everyone.

The data is not clocked until it's fed into the DAC. It's fed from the buffer and then it's buffered by the DAC before it's clocked and converted to analog signal.

If it was so high end, it should have a better drive setup, and it wouldn't need to be so necessary to make it level...any other nonsense related to high end? If the "high end" units need to be leveled just right, I'll stick to the medium stuff since I've never had a problem with being level or not, on a $1200 Tascam, but then that's not audiophile just pro , so what do they know? My $300 Philips don't care either...maybe you need to clean or lube the optical assy, and make it track properly, the sleds may have dried out grease, making it not respond to signals to it properly,. how does the old Cd changer in my trunk keep playing over bumps and everything, the cartridge goes in vertically? any other high end helpful hints, another myth debunked, next!!!!

I believe a lot of CD drive mechanisms are designed to work even upside down and sideways. I know the Philips VAM 12 series were. And some high end manufacturers even place the drive on an angle. Kind of defeats the level arguement.

Quote:

I realize some people believe jitter is audible, but they would be wrong. At least for modern gear made in the last 10+ years. The article you linked to is more than 16 years old. But I doubt jitter was a "serious" problem even back in 1992. And certainly not when compared to the amount of "jitter" inherent in LP records and analog tape. I'll never understand why jitter with vinyl and tape is cool, but for some reason it's a huge problem with CDs.

--Ethan

Your incapacity to understand jitter as affecting digital audio spans multiple forums and is unfounded and beyond reason. Just like the title of the Elvis album, '5 million audiophiles say you are wrong'.

I'm not here to get into an argument with you about it, you've already shown your incapacity to deal with the subject-across multiple forums.

I'm simply offering the casual reader balance that you are not providing.

Incapactiy? Is that a real wurd? Incapacitated....Inabilty.....Incapacity? Jitter is a non issue now, both my standa alone DAC units deal with anything as an input the VanAlstine OmegaStar EX Hybrid DAC or the Benchmark unit...jitter is so 20th century.....we ain't got no stinkin Jitter. Stop reading outdated articles, that mean nothing anymore..have you noticed how the IC business has progressed? There is an incapactiy to have outdated information being applied to modern products.

Quote:
Your incapacity to understand jitter as affecting digital audio spans multiple forums and is unfounded and beyond reason ... I'm simply offering the casual reader balance that you are not providing.

You are offering absolutely nothing at all m'man. Any bozo can say "You're wrong," but it's quite another thing to explain how I'm wrong and why. You have done neither.

Now, if you'd like to explain how artifacts 100+ dB below the music can possibly be audible, I'm all ears (eyes). But I suggest you first try my Artifact Audibility test files linked above, so you'll have at least a basic understanding of what you're talking about.

--Ethan

Don't bother. Trying to change Ethan's mind is like trying to change dup's.

Quote:
With disk drives generally, there are a few layers involved. With an operating system like DOS or Windows, the OS tells the drive to read a block of data starting at a particular sector, and put the data into memory starting at a specific address. The drive finds that sector, reads the data, and puts it in memory.

That is correct for data (hard drives, CD-Rs, etc.).

Quote:
I assume the same happens with CD players, where the player's firmware requests data from the drive at a specific sector.

No, that can't be done easily with audio CDs, because there is no block addressing. Instead, you have one single stream of data, marked with time code.

John Atkinson
Editor, Stereophile

Jitter does matter and is an issue. Look no further than the latest Mac Music Server. There is also a great article on jitter in the Phile archives, but YOU'LL have to look it up. Jitter is real and can be a problem if a circuit designer makes poor choices. JA has continually made this clear.

Quote:
No, that can't be done easily with audio CDs, because there is no block addressing. Instead, you have one single stream of data, marked with time code.

Thanks for clarifying John. But if this is the case, how can a CD player start at song #12 or whatever?

--Ethan

Quote:
Jitter does matter and is an issue.

I've been round and round on this in a few pro audio forums lately. So far not one person has been able to 1) come up with a concrete audio example that shows normal amounts of jitter as audible, or 2) explain how artifacts that are 100+ dB below the music could be audible even in theory let alone practice. The best one fellow was able to do was contrive a Wave file where an artifact 80 dB down was audible. That's a good 20 dB louder than any jitter I can possibly imagine.

I do understand that most people believe jitter is a real problem, but as best I can tell it's yet another myth that may have been relevant 20 years ago. Say it enough times and next thing you know it's accepted as fact by everyone. But if you have anything for either 1 or 2 above, I'd love to hear it and be educated.

--Ethan

If audio CD's are in a different format than data CD-R etc. What about when you record audio on a CD-R? Is a CD-R different data format than a pre recorded data CD, that say is newly purchased software. What's the difference in data format if I record an Audio CD on a Cd recorder, whre you need teh "audio" CD blanks, BUT if i use a CD-R on the CD recorder like the HHB or Tascam which will record either type. Or record the CD-R on a computer drive? When it's played back on a different machine when does jitter matter, or not matter. What if I record in DSD on the Kord? Is jitter a problem in DSD?

And is ANY of this stuff audible anyway? If i move my head while listening to something ain't I changing the listening enviorement, and then there is a new pile of variables? And you made a bigger change in something else, that makes any of this jitter stuff a wash? Bigger SPEAKERS less JITTER!!! that's what I say

Quote:
No, that can't be done easily with audio CDs, because there is no block addressing. Instead, you have one single stream of data, marked with time code.

I'm having a hard time confirming this. So far everything I've found about the structure of audio CDs refers to sectors, which to me implies random access. There is a TOC on audio disks which holds the starting sector for each song, and the CD player can ask the firmware to start at a specific song number to avoid dealing with sectors. But as best I can tell the data still has to be read in blocks, put into RAM, and sent to the D/A converter.

Indeed, if this were not the case, those jogger CD Walkmans I mentioned earlier with 8 MB buffers would not work to buffer against physical shock. And when the player gets jostled and the drive can't confirm the data, the player must seek to the same sector and try again before resorting to reconstruction.

Also, these days all CD players are computer drives anyway and can play CD-Rs reliably, and most also play MP3 files. So at the lowest level of disk I/O in current players, a stream-only scenario doesn't make sense to me.

BTW, I searched for half an hour and found very little. Do you have any links to such details? I called my expert friend Bill Eppler because I'm sure he knows how audio CDs are read, but alas I got his answering machine.

--Ethan

Quote:

Quote:
No, that can't be done easily with audio CDs, because there is no block addressing. Instead, you have one single stream of data, marked with time code.

Thanks for clarifying John. But if this is the case, how can a CD player start at song #12 or whatever?

The PQ subcode on the CD tells the player the time code for the subframe where the song starts. Actually, when a CD is mastered, it is good practice to put the track start marker a little bit before the exact subframe. This allows the player time to unmute its output in order that the beginning of the music isn't clipped. For example, when I master Stereophile CDs I place the marker 1/10s ahead.

John Atkinson
Editor, Stereophile

http://en.wikipedia.org/wiki/Compact_Disc#Data_structure

PQ http://en.wikipedia.org/wiki/Compact_disc_subcode

The term "stream" is a group of FRAMES data, so is it really a continuos "strea" as it would imply a continuos like a river, FRAMES implys a coridnated section group of data http://en.wikipedia.org/wiki/Compact_disc#Data_structure

Quote:
The term "stream" is a group of FRAMES data, so is it really a continuous "stream" as it would imply a continuous like a river?

Yes, as the smallest amount of data that can be addressed is 1/75th of a second's worth, what I referred to as a "subframe" in my earlier postings.

Thanks for the Wikipedia references, Carl. The subject is also completely discussed in Ken Pohlmann's excellent book on digital audio.

John Atkinson
Editor, Stereophile

Yupper, I have that book. Lotsa "data" on data.

Quote:
http://en.wikipedia.org/wiki/Compact_Disc#Data_structure

Yes, I found those pages, but they didn't address what John and I are discussing. Or maybe I couldn't figured out how to read that page.

Quote:
The subject is also completely discussed in Ken Pohlmann's excellent book on digital audio.

D'oh - I have that book! Okay, busy reading, back soon.

--Ethan

I just got off the phone with my expert friend Bill Eppler and got the low-down. CDs do not use the same sector addressing structure as hard drives, but it is possible for a drive to seek to a specific location. So my original scenario is correct in principle even if I used the wrong terminology. If a CD player encounters a problem while playing a track, it can indeed try reading from the same place on the CD again. This is done inside the drive's firmware and so is transparent to the rest of the player. So John and I are both correct - addressing is not by blocks, but a drive still seeks to specific locations on the CD.

--Ethan

Quote:
So my original scenario is correct in principle even if I used the wrong terminology.

Hey, I called my car a cake one time. Pretty simple mistake.

No, Jan... you called a single driver speaker hi fidelity....

Quote:

Quote:
The term "stream" is a group of FRAMES data, so is it really a continuous "stream" as it would imply a continuous like a river?

Yes, as the smallest amount of data that can be addressed is 1/75th of a second's worth, what I referred to as a "subframe" in my earlier postings.

Thanks for the Wikipedia references, Carl. The subject is also completely discussed in Ken Pohlmann's excellent book on digital audio.

John Atkinson
Editor, Stereophile

Interestingly enough, the story goes that the finest jazz (or more 'aware' players) players play with enough subtlety that the 'timing' considerations for MIDI (at 1/64th of a second) ---just don't work for them. The subtle 'work' is lost.

MIDI timing resolution is much better than 1/64 of a second. Fortunately.

Real time resolution of a good sequencer is about one millisecond per note.

Better than I can hear.

Then there are the high-end specialty products such as that used by Zenph Studios to recreate the Gould performance of the Goldbergs which have even more resolution.

Quote:
Then there are the high-end specialty products such as that used by Zenph Studios to recreate the Gould performance of the Goldbergs which have even more resolution.

This is totally tangential, but was anyone else surprised by how un-Gouldlike the Zenph recreation of the Goldberg sound? Maybe it's b/c they used a different piano (I think it was a Yamaha?)...

The reactions to this recording are fascinatingly diverse.

Of course it doesn't sound like Gould; no humming, grunting or chair squeaks.

The particular Yamaha piano used was carefully voiced for the project. While Gould recorded the 1955 Goldberg's on a Steinway, he later played Yamaha and was a "Yamaha" artist.

Yamahas tend toward huge dynamic range and a sparkling sound, especially in the higher registers. (This sparkle and clarity is why many jazz musicians prefer a Yamaha.) Steinways tend toward a darker and somewhat murky sound, although harmonically rich. Would Gould have played the Goldbergs differently on this particular Yamaha? Perhaps.

Some also claim that the acoustic space is different and that Gould would have played differently for the different space. Again, perhaps.

Another Yamaha/Steinway difference: Yamahas have spectacular quality control and are ready to go as delivered by the factory - even the tuning is usually quite good yet even after shipping. Steinways almost always need some work by a good technician.

Amusingly, being Japanese, Yamaha dampers are incredibly effective - making them almost too good. When a key is released, the sound STOPS. It takes a talented player to play a truly legato phrase on a Yamaha. I am not one of those pianists.

As a side note, it really makes a difference if a recording is played back in the same space in which it is recorded (or, even better, the space is carefully acoustically treated for accurate sound reproduction.)

I have a Yamaha conservatory grand in the same space as my main system. The room is about 35'x30' with a 32' ceiling. It has a characteristic sound. I recently recorded a CD of piano four hands on this piano with large diaphragm condensers about two feet away from the fully open lid half-way between the body and lid.

Play back of this recording on my system is the only recording I have heard that makes me believe there is a real acoustic instrument in a real space. It is jaw-droppingly realistic. This quality disappears and it becomes a normal recording once it is played back anywhere else.

Thanks for the explanation, Elk! I have no doubt that Gould would have played the Variations differently on a Yamaha, in a different place. I mean, his less well-known version of the Goldberg recorded later in life is SO different from the earlier one... but the Zenph's marketing led me to believe that it was the '55 Goldberg somehow re-created, and that wasn't what I heard... but I think I should now listen to it with an open mind. At the time, I was kind of taken aback.

On Yamaha - I love their pianos. I used to play pretty seriously when I was much younger... I was in the competition circuit, and was training for the Van Cliburn competition & Juilliard prep at the same time when I was in 9-10th grade. My parents had decided to buy me a grand piano, and I ended up buying the Kawai 8-ft, mainly b/c it was cheaper. The action was easier/lighter on the Kawai, but I always found Yamaha more pleasurable to play... I had a little bit of a different experience from you; I could manage legato much easier on the Yamaha, and it sounded much more beautiful than the Kawai for sure. Of course I loved the Steinway, but way out of our family's budget. I found that the Steinway was the piano that responded exactly to my playing, laid bare all the flaws & strengths. Loved playing the opening chords to Rach 2 on it. Made me feel like a pro. I also liked the Bosendorfer a lot. Yamaha, to me, sounded the softest among them all, in good & bad ways... it sounded the best on quiet pieces, like the late Brahms intermezzi, but not as good on concerti & bigger material. But I've no doubt that it was probably b/c I wasn't quite developed as a pianist yet (not saying that I am now, by any means)

I always walk on the 58th St piano shops in Manhattan & drool over the Faziolis, Erards, etc... I think a lot of myself died when I quit playing seriously. Maybe my recent fixation on audio is that lost desire sublimated, who knows...

Gould's later recording is very different, I suspect as a result of leaving with the music for multiple decades and a desire to say something different than before. He brings out different lines, employs different tempi, etc.

Your comments made me consider how a pianist likely relates differently to his instrument than I do as a trumpet player. I play, and choose the instrument upon which to play a piece, based upon how I want to sound. That is, the trumpet does not dictate the playing, the music does.

A piano is often a solo instrument however. Thus it makes sense that the formula works the other way around, at least to some degree. Of course the player always serves the music, but that big instrument that most often is supplied by the venue, will control some degree how the music is played.

I have never even touched a Fazioli, let alone played one. I rarely even see one.

Dbowker, my woodworking experience is primarily as a luthier and harpsichord maker - all avocational. Thus, I ahve worked with quite a few fun woods. Of course, I have made a few bookcases and the like. I haven't made anything in a good while, no longer having much of a shop (the typical moves, space issues, etc.)

I love your MyPingo experiment. It's been fun to travel along on the trip.

Elk,
luthier as in guitar maker or as in violin?

Guitar, fretted dulcimers and hammered dulcimers.

I made only "OK" guitars - perhaps partially as I don't play guitar well.

I can't imagine making a violin without lots of training.

Leo Fender didn't play at all. He just knew what he needed to do.

Yes, the guy that screwed up the meaning of both the terms tremolo and vibrato.

Keep in mind that he was dealing primarily with electronics, not with the acoustic properties of an unamplified guitar. His instruments were solid bodies - not much art in cutting a slab-o-wood. He did, however do a magnificent job at what he did and was apparently a delightful man as well.

Quote:
not much art in cutting a slab-o-wood.

Agreed, but not having the usual body resonances is an important part of solid body electric guitars. All notes sound and sustain more or less the same, which in this context is an important feature.

--Ethan

Absolutely - it must work. But one is essentially making sure that nothing resonates, rather than that everything resonates equally. The pickups do the work. Pickup placement and type are the primary voicing variables.

I am taking nothing away from a good electronic instrument maker. There is art here. The art however is not in the woodworking.

Then there is the Les Paul from Gibson, they do come in different woods, and yes they are very artful....expert guitarists know the sound, and which does what. Over the years Les Pauls have come in many different blends of wood. If it was just electronics a Fender would sound like a Les Paul, they don't, differetn pickups sure, but entirely different bodys, differetn wood. Then the Taylor versus Martin....Even the coatings on old guitars sound differetn than new ones, why Gibson VOS has the old finish to sound like teh old models, electric guitars are more than just what pickup is put on them.

There are, of course, other variables - but these are much less influential.

Again, I am taking nothing away from those that make good electric instruments - it is simply a very different craft than making an acoustic instrument that relies only on the guitar itself to produce the sound.