Bits Are Bits, Right?


So I'm currently heading down the path of exploring which CD-Rs sound best in my CD player, along with what burn speeds sound best and what CD burners make the best CDs. I already know from my research that the more accurately the pits are placed on the CD (e.g. less jitter in the recorded data), the better chance I stand at getting the CD to sound good. There is a counter-argument to this idea that goes something like this: "Bits are bits and as long as the CD player can read them, the accuracy of the spacing doesn't matter because everything is thrown into a buffer which removes the effect of any jitter written into the data during burning." I know I don't agree with that logic, but for the life of me I can't remember the technical reasons. I know I used to know. Haha! 

So who here knows why buffers don't solve all of our problems in the digital realm? How come timing accuracy matters in the stages before the data buffer?
mkgus
From a post by member Kirkus (who has a vast amount of hands-on experience with the internal workings of CD players) in this Audiogon thread from 2011:

Two big conceptual errors I see very commonly are the assumption that any intrinsic jitter related to retrieval of information off of a CD actually occurs through the forward signal/data path, and that any sonic artifact associated with parts upstream of the DAC must be classifiable as jitter.

In reality, CD players, transports, and DACs are a menagerie of true mixed-signal design problems, and there are a lot of different noises sources living in close proximity with susceptible circuit nodes. One oft-overlooked source is crosstalk from the disc servomechanism into other parts of the machine . . . analog circuitry, S/PDIF transmitters, PLL clock, etc., which can be dependent on the condition of the disc.

One easy way of measuring this on the test bench is to have two versions of the same test-tone CD, one pristine, the other scratched. A conventional distortion analyzer is used to null out the the tone(s), and then an FFT (or visual 'scope analysis) is used to analyze the residual. One would be surprised at some of the nasty things that sometimes come up out of the noise floor when the focus and tracking servos suddenly have to work really hard to read the disc.

Regards,
-- Al

It all comes down to a couple things: the scattered laser light gets into the photodetector as noise, the CD player is susceptible to external and internal vibration, the CD itself flutters during play so much that the laser servo system can’t keep up. The Reed Solomon Error codes are practically worthless. There is no buffering in most CD players. All the King’s horses and all the King’s men couldn’t put Humpty Dumpty together again.
"Bits are bits and as long as the CD player can read them, the accuracy of the spacing doesn’t matter because everything is thrown into a buffer which removes the effect of any jitter written into the data during burning." I know I don’t agree with that logic, but for the life of me I can’t remember the technical reasons.

Could be what you forgot is its wrong. Bits are indeed bits, but the pits and land aren’t bits at all. Each is a word comprised of a whole string of bits, with the length of the pit and the breadth of the land being what determines the string of bits. So it is a time function and anything, including a lot of things GK clowns around about, can mess with the bits and thereby the sound. CD is in other words not digital in the sense almost everyone imagines, but analog. Only CD is analog with a whole lot of noisy circuits and whatnot in the path. So technically even worse than analog. If such a thing even is possible. Which CD proves every day that it is.

Perfect sound forever! Technophilia uber alles!
Yes, digital only exists as a mathematical concept. All of reality is analog (at least the reality we deal with - at the scale of Planck time and Planck lengths things may be different). A stream of “digital” data is an analog signal that a computer has to interpret as a 1 or a 0 by deciding when the value has changed enough and at what time to be interpreted as a different bit.

One of the ways I know that what happens before the data buffer matters is the difference in sound quality between streaming and reading a local file. I have always thought the local file sounds better than streaming even though it’s the exact same data. Just recently, I was driving down the road with a friend when they plugged the phone into the car and the sound quality was much better than usual. I asked what they did, and I found out they were playing songs off the phone’s “hard drive,” whereas I am usually streaming from Tidal. Same data, way different sound.
Almost all this goes away with the use of a modern DAC. :)
Just recently, I was driving down the road with a friend when they plugged the phone into the car and the sound quality was much better than usual. I asked what they did, and I found out they were playing songs off the phone’s “hard drive,” whereas I am usually streaming from Tidal.

Yeah. What I thought. Crap so bad you can tell its crap even in a car. And there’s guys come on here, what’s the best $5k speaker/amp/cable/whatever to pipe my Tidal crap through? So it doesn’t sound quite so crappy, I guess. Or hear exactly how  Then there’s guys try and pretend this can all be magically solved with a DAC or whatever. Right. As if.

Yep, I love Tidal but I don’t stream in my audio system. It’s nice to have when guests come over so they can play whatever song they want to hear but for me, it’s physical media or nothing. Now maybe things can be improved with a high quality reclocker or some other technology like asynchronous DACs, but I’ve long noticed that even when streaming through my DAC sounds good, it sounds even better from a CD.
In my system Quobuz is at the very least equal to my library stored on a Synology NAS (sometimes even better).

Oz



Not sure what you guys are streaming or using but not my experience at all!
Qobuz hires at 24/192 played back through my MHDT Orchid into BAT vk300se sounds spectacular.

Anyone who just disses streaming for the sake of it without actually trying a decent setup is not being fair to the media one little bit.
Glad you are liking the Orchid uber. I have mine in my computer system. It sounds better than my turntable in that system by a considerable margin and the turntable, cartridge and phono stage cost close to three times as much as the Orchid. In my main system, I'd say digital vs. analog is a toss up, with each doing some things better than the other. Again, much bigger investment in the vinyl rig. 

mkgus OP
Yes, digital only exists as a mathematical concept. All of reality is analog (at least the reality we deal with - at the scale of Planck time and Planck lengths things may be different). A stream of “digital” data is an analog signal that a computer has to interpret as a 1 or a 0 by deciding when the value has changed enough and at what time to be interpreted as a different bit.

>>>>No, actually reality is more like digital. At the quantum level, which is really where the rubber meets the road let me remind you, gentle readers, that everything can be described by its quantum state; electrons have certain quantum states and require a certain amount of energy to get to the next level/orbit. That is more like Digital than Analog, I.e., having non-continuous states. In an analog world the electrons would not have non-continuous orbits. Light also is described by quantum states. To whit,

“The photon model accounts for anomalous observations, including the properties of black-body radiation, that others (notably Max Planck) had tried to explain using semiclassical models. In that model, light is described by Maxwell’s equations, but material objects emit and absorb light in quantized amounts (i.e., they change energy only by certain particular discrete amounts). Although these semiclassical models contributed to the development of quantum mechanics, many further experiments[3][4] beginning with the phenomenon of Compton scattering of single photons by electrons, validated Einstein’s hypothesis that light itself is quantized.[5]
Greg
The Orchid took some burning in to warm up to my ears and a change of IC but now, oh yes it's good, nay it's very good.
And I just have stock tube in it as awaiting arrival of the WE tube.

It's in my main rig where it contends with a record rig at least 4 times cost.

In my second system I just use a $50 Google Chromecast Audio puck that produces amazing SQ for what it is.

Imagine what $50 of total record replay equipment would look and sound like....

Uberwaltz - you’re right. I’m not being fair to the media. I personally believe that a computer has the greatest potential to sound best of all the mediums, however, computers are notoriously noisy environments and very few computer parts are built with maximizing audio quality in mind. A CD player is a type of computer, albeit one that sounds better, dollar for dollar, than what we typical call computers because they are simpler in nature and purposely built to sound good. Note that I said “potential,” and not “in practice.”
"Bits are bits and as long as the CD player can read them

Not really, the error correction is a constantly working part, if the error correction is working hard and guesses whether to throw a 0 or a 1 into the bits it can’t read, it’s got a 50% chance of getting it right, because it always subs the one it can’t read with whatever came before it.

Here are three closeups of an of the shelf "retail CD" (left) and a "burnt one" (middle) and a "re-writable one" (right)
Which do you think is going to give the more errors.
https://ibb.co/5nXB6bC

Same happens with streaming and with H/D stored stuff, there are always read errors to be substituted

Cheers George

Mkgus
Agreed on computer being noisy for audio playback, one reason never really used just a computer.
I use a dedicated streamer whose sole purpose in life is to stream hires music from either an online streaming service or an attached hard drive of some description.

You will likely find that’s the way majority do it who are looking for top flight SQ.

Btw my comment was not necessarily aimed at yourself, just in general.
I hate to tread into the waters, and certainly dont want to design and teach the long class necessary to get to all this, so a few simple comments will do:
1. There are no bits on a CD. repeat after me "eight to fourteen modulation"2. A digital signal, primarily SPDIF which is source cloaked is a half-analog signal, with the amplitude digitized and the X-axis (timing) analog.3. Clock recovery or buffering/readout is impacted by many things including originating jitter and various types of noise4. That said, really good modern DACs using USB (asynch) reduce many of these problems by vast percentages.
Neither the hogwarts school of audio nor the "its all simple and really well understood" school (wrong twice!) are even close to right.
In general quit worrying about rippign speeds and CD media and get good files, on anything, transfer them over a DAC-clocked interface, galvanically isolated, and move along.
For the record i can get remarkable sounds from 320-384 kbps CBR MP3.  OK, FLAC is better, but not much.  HD - the jury is out, clearly good in the studio. I wonder if what  hear is simple better master transfer when the jury is occasionally in, to cripple a metaphor.
G
Unfortunately, our brains sample 10x more frequently than cds...vinyl is best.

mtdining
Unfortunately, our brains sample 10x more frequently than cds...vinyl is best.

>>>>>I’m from the future. When the CD is played on a CD player not (rpt not) encumbered with all the problems CD players have had since the very beginning, CD has much more detail than vinyl. It has always been right there on the CD. You just couldn’t hear it completely or accurately, that’s all. You don’t even have to play 20 bit or 24 bit CDs. 16 bit Redbook CD will do just fine. It’s the player, you can forget about everything else. In the future there is no more glare, no more congealed midrange, no more weak bass, no more two dimensional sound, no more thin paper mache sound. 
GK
I see that stand up comedy routine you were harping on about REALLY is working out for you.
Keep it up!
Did you forget to take your smart pill this morning, Uber? 😳
Pretty sure someone here dident……..
+2, @ozzy62.

Please don’t let this thread turn into digital vs. Vinyl never ending debate.  I see couple of vinyl fanboys already chimed in....they just can’t help themselves spewing their hatred for digital.

“I was driving down the road with a friend when they plugged the phone into the car and the sound quality was much better than usual. I asked what they did, and I found out they were playing songs off the phone’s “hard drive,” whereas I am usually streaming from Tidal. Same data, way different sound”.

@mkgus,

The SQ of a track greatly depends on source of master file, listen to ‘Hotel California’ from Hotel California 40th Anniversary Expanded Edition (2017) album first, then listen to the same song from Hell Freezes Over (Remaster 2018) album.

Looking forward to your feedback!
uber, are you looking in the mirror? Eat more fish.
GK
Sorry but no mirrors in my house......

Lalitk.

Totally agree, nice post!
I agree, the source file is all important. Different remasters no doubt sound different. In my car example, I’ve repeated the test several times. When I play the track from my phone’s memory, it sounds better than streaming even with the example same version of the song. The most noticeable part is the bass. From the phone, the bass is much more like a quick “punch,” and when streaming it’s more of a quick “shove,” if that makes sense. “Tighter, faster bass” might be a metaphor I could use.

Last night I copied one of my favorite CDs to a black CD-R and the difference in sound quality between the 2 discs was quite apparent. The treble was much more laid back and “tamed.” I could hear micro-details better. It’s too early to conclude which version is better as they both have their pros and cons. I’m interested in the “why.” Does one disc have more read errors than the other? If not, then I conclude that the way the data (the exact same data, that is) is arranged on the disc matters. The precision of the pit spacing, the width and depth of the pits, and the material of construction of the CDs may play a role. If it’s all the same data and the read errors are minimal, then what is happening before the buffer is having effect on the sound.
mkgus
Last night I copied one of my favorite CDs to a black CD-R and the difference in sound quality between the 2 discs was quite apparent. The treble was much more laid back and “tamed.” I could hear micro-details better. It’s too early to conclude which version is better as they both have their pros and cons. I’m interested in the “why.” Does one disc have more read errors than the other? If not, then I conclude that the way the data (the exact same data, that is) is arranged on the disc matters. The precision of the pit spacing, the width and depth of the pits, and the material of construction of the CDs may play a role. If it’s all the same data and the read errors are minimal, then what is happening before the buffer is having effect on the sound.

>>>>>There are a number of possibilities. One is that the copy routine was copy til perfect. Another is that the laser reads black CD-Rs better than standard silver CDs. It could be there’s less laser light scattering in the black CD-Rs. Just as 24 it Gold CDs have higher reflectivity than silver CDs and usually sound superior. One reason SHM Super High-Performance Material CDs from Japan generally sound superior is the clear layer of the SHM CD is more transparent to the laser than plain old polycarbonate which is only about 91% transparent. Less light scattering. Better optical signal to noise. It’s possible the clear SHM material is stiffer than ordinary polycarbonate and that the CDs are more perfectly round. Another more far out possibility is that copies just sound better, period. Copy an LP to tape, copy a CD to CD-R, copy a CD to tape. No one knows why.

The spacing between pits and lands varies, they represent a series of “words” of variable but specific meaning, the length of pits and lands themselves is also variable. the details are specified in the REDBOOK standard. The geometries involved with depth of pits is also specified in the REDBOOK. The system won’t work if the geometries are not absolutely correct as the laser light beam cancels itself out by wave interference when it strikes a pit. That’s why there is no return light signal for pits, only for lands.
@mkgus,

I hear noticeable differences between the sound quality when streaming through my iPhone over Bluetooth vs wired connection. The wired connection in my car always sounds more dynamic, less congested and smooth.

I have no experience with CD-R, I am either listening to Qobuz streaming or original CD / SACD’s through my CD and SACD players. To my ears, some of my favorite CD’s sounds better through my CD Player (with tube output stage) vs digital copy of same CD stored on my streamer internal hard drive.

To me, it’s all come down to the implementation. The devil is in the details.
@mkgus,

Bits are bits. Some DACs like MSB and PSAudio Directstream support bitperfect test files that you can stream from your source.  When the DAC detects the file contents, it lights up 'BitPerfect' ...which tells you the path from your source to the DAC is not disturbing the bits.  CD players are bitperfect sources - those bits from the CD get to the DAC with 100% fidelity.  
Then why do some CD's sound different?  Even pressed vs ripped CD's of the same album? Its because the semiconductor chip in the CD player is a clocked digital device that has to decode the error-correction of the physical media. So even though all the CDs contain the same music data, the physical substrate varies and reliability of the encoding on the disk varies and the chip works differently to extract the correct bits.  Hence the bits always are decoded correctly but the process of the chip doing so produces disturbances in the EM field generated ...and this (remarkably) affects the DACs analog electronics (stability of reference voltages).  This effect is true with a galvanic coax connection but even with a isolating Toslink, the EM field leaves the CD enclosure to travel meters to the DAC.
We're all dumbfounded by this because its seems impossible that such miniscule issues should be audible ...but they are.
Mkgus 2-21-2020
If it’s all the same data and the read errors are minimal, then what is happening before the buffer is having effect on the sound.

As I see it the reason is that what is happening before the buffer, which is dependent on the physical characteristics of the particular disc, is affecting circuitry that is after the buffer.  That is the basic point Kirkus was making in his 2011 post that I quoted above.  Those effects are not occurring via what he refers to as the "forward signal path," meaning the intended signal path that we tend to think of, but rather via coupling of noise via unintended pathways.  Those would potentially include grounds, power supplies, stray capacitances, or even the air.  Such effects will of course vary depending on the design of the specific component.  

Regards,
-- Al
 
Okay, which one of you submitted this question to Paul? The timing is eerie. 
CD or High Resolution Streaming?https://www.psaudio.com/askpaul/cd-or-high-resolution-streaming/
almarg
As I see it the reason is that what is happening before the buffer, which is dependent on the physical characteristics of the particular disc, is affecting circuitry that is after the buffer. That is the basic point Kirkus was making in his 2011 post that I quoted above. Those effects are not occurring via what he refers to as the "forward signal path," meaning the intended signal path that we tend to think of, but rather via coupling of noise via unintended pathways. Those would potentially include grounds, power supplies, stray capacitances, or even the air. Such effects will of course vary depending on the design of the specific component.

>>>>Is there a Babel Fish translator for that?

addendum: if it doesn’t make sense it’s probably not true. 
My only question is “what’s a CD player?”
Convenience over SQ, right?
Tidal through a good streamer and dac can be very pleasurable for sure. Don't be fooled otherwise.
Some years ago (maybe someone can help with the brand?) while I was working at a HiFi factory, we had tested AudioXsell audiophile grade CD-R
which not only sounded better than any other discs we’d heard (including HDCD and SACD) but they also played on my stubborn car player, which liked to only play purchased manufactured music CD not burned.

https://i.pinimg.com/originals/8d/e6/f5/8de6f5d6a5b54520b3bd213d587dd7b0.jpg

Yamaha did make an external audiophile SCSI CD grade burner [F1], which was pretty darned slow, but it surely was the absolute best burner, made specifically to be configurable to have less turns on the burn. Effectively broadening the width between data on the disc.

Here is a link to a list of drives tested for accuracy and using gold hospital archive grade CD- R media is the best bet.

https://forum.dbpoweramp.com/showthread.php?43786-CD-Drive-Accuracy-2019
A couple notes from my findings last night:

Black CD-Rs sound pretty good. With my burner and media, 4x burn speed sounds the best. 2x and 1x sound okay but they aren’t as musical as the original pressing.
The media I’m using says max 16x burn speed. Does anyone know if the max burn speed listed on the package matters? Is lower better or is higher?
+4 @ozzy62 
same exact experience!
 when streaming in main system Qobuz noticeably cleaner and more incisive than any format from Tidal. No complaints on Qobuz quality but hoping they add more over time!


Yes, Quobuz has a lot of holes in it's catalog. None of the services have everything covered, but they seem to be missing far more of the music I search for than Tidal or Amazon.

Oz