Illuminating CDs: A New Thread
I received a copy of an email this morning from Mark of CDIllumination, the company in Australia that is “vivifying” CDs by copying them to Gold CD-R. I wrote about the folly of this service for the obvious reasons that a “copy” from a Gold CD-R will output the same stream of bits as the replicated original. In fact, it might do it better than the copy because the disc is actually pressed or stamped instead of “burned” as the CD-Rs are. Anyway, Mark expresses confidence in his process:
Thank you for asking about the CD copying. I use several processes that I have verified to improve the sound of the copied CD, together they work synergistically. I am still trying other methods to try to further improve things.
These processes were derived empirically after I noticed that some copies of CDs improved others degraded. I kept the improving ones. Certainly decreasing jitter improves things as does reduced BLER on the CD, however, other processes improve things and I don’t know what to measure to make a correlation. But that doesn’t matter as long as there is an imrpovement.
I do make adjustments to the pits and lands of the burned cd as well as altering optics. I have also tried many types of discs from black to aluminium to silver to gold reflective layers and my personal preference is for the gold. The others are not to my personal taste but the sound is different.
I would group the differences in the rip, the computer hardware, the cd optics and finally the burning process. It takes me about half an hour to do this with a photocopy image of the original cd onto the new cd. I think the result is a bargain compared to an equipment upgrade that might cost many thousands and not return the same improvements.
But these are just words.
Here is a challenge to you: send me a cd in WAV with music that you know well as a reference (compilation is OK) and I will do the process free of charge. If your ears and equipment are good, I know you will hear the improvement first hand. If you do a compilation perhaps some of your friends would like to contribute a track or two.
Mark – CD Illumination”
Here’s a few questions or points that are worth making.
I’m a little skeptical of any company that uses the term “synergistically”. That term shows up too often in the high-end audio business. I would prefer to know what exact processes the gentleman from CDIllumination is using to “improve the sound of the copied CD”.
He mentions decreasing “jitter” but doesn’t say at what point in the process or that clock can be completely redone at the very final stage of the process…thus removing any artifacts associated with timing errors.
Reducing BLER or error rates on the CDs doesn’t really count for much either. Most errors are correctable and the original disc and copy will produce the same stream of bits.
Mark of CD Illumination says he does, “make adjustments to the pits and lands of the burned CD as well as altering optics”. What does that mean? How does he accomplish that since it’s a rigorous process defined by the specifications of the CD-R Orange Book.
Ultimately, this is going to come down to blind faith again. He says that everything in his process, “doesn’t matter as long as there is an improvement”.
Finally, he asks that the reader send him a cd in WAV file format. The process is taken from a replicated CD…that’s what should be sent. I wrote to Mark as well and look forward to hearing from him. I’ll keep you posted.
24 thoughts on “Illuminating CDs: A New Thread”
Interesting this one. I recall at a demonstration of Wilson Audio loudspeakers several years ago the salesman had some CDR’s of jazz, pop and classical he had obtained from Meridian UK. He stated they had discovered that recording a CD to a CDR was an improvement in sound over the original CD. Anyway he did not compare these at the time as he was busy trying to sell Wilson loudspeakers. I do recall a female jazz vocalist on CDR that was taken from a master being played that was stunning in sound and the standard CD’s of other music paled in significance in his demo.
The late Harry Pearson (The Absolute Sound Mag editor) at that time also listened and compared standard CD’s to CDR’s I believe from the same source Meridian UK. He did concur that he felt they did sound better but could offer no reason why this should be.
There’s just no way if the discs contain the same data…and that data is successfully retrieved.
Well having heard what others have produced with re writable gold discs I have to concur. They are indeed an improvement, it took me totally by surprise. Bits may indeed be bits but the laser reading them makes mistakes and has noise and jitter added.
I have burnt CD copies before with the inevitable sameness about them but when I was loaned some discs written on gold ‘master discs (Mobile fidelity sound lab Ultradisc CD-R)’ I was really surprised by the increase in overall sound quality.
I should wait until the disc arrives before passing comment. The CD player used for replay is a Marantz CD95 (not your favorite blue-ray player which pales in comparisons done locally) and the chip set and NOS IMV are far superior wrt a transport for getting the digital stream off the disc. Error correction may be working a lot less with the new gold discs due to pit spacing and the recorder creating the pits, a Yamaha I believe with interesting software controlling the pit spacing. I see no reason to doubt what Meridian have produced in the past as being an improvement as they are accepted as being a forerunner in the digital music world.
I have also tried the XRCD range of discs and these are usually a nice step up from the regular CD pressings attributable to their K2 laser and K2 Rubidium clock according to them.
A dedicated Audiophile friend will not buy any CD that has not been pressed in the country of recording, he has proved to me various ‘versions’ from around the globe all sounding worse, not to the MP3 extent but certainly a less engaging listen all round.
Just taking the data stream from a CD and writing it to a micro sdxc card and replying it from there tells us a lot about how poorly a lot of transports read the CD discs and passes it for D to A conversion, shocking how poor the CD sounds in comparison, yes really. So what are the gains in doing this, well immediately bass timing is much better and sound stage ambiance is improved, there is also a lack of glare which many disc spinners are guilty of.
So a CD with ‘bits’ should be the same but anyone with a decent system (>20K) will easily hear the difference. Much like a fine claret and a cheap bottle of plonk, they may both have the same deep red colour but they are far different once uncorked and aired in a fine crystal glass. So please keep an ‘open mind’ when it comes to others claims, you would be surprised at what lengths as audio types will go to to try and recreate what you recording engineers can totally miss using that damn pro tools. Yes that’s a pop at the types who masquerade around with a USB stick with all their settings on which I hope is not you!
Until I can do a rigorous test by playing an identical data stream from a GOLD CD-R vs. a replicated CD, I refuse to accept that there could possibly be any difference….given that everything else in the playback is identical. This is the very definition of digital. My plan is to let the gentleman from CD Illumination “copy” a replicated CD that I’ve mastered (probably the Chamber Music Palisades title…I made a CD of that project because of the targeted demographic of their audience). Then I will do the comparison.
XRCDs are the same thing…sorry but data is data. I seriously doubt that I will hear any difference. But I will give it a very fair shot…and share the experience with other engineers in my facility.
I don’t see a problem with accepting the test… if he is creating the CD-R from an audio CD (Not a data CD containing a WAV file), they should be bit identical. So first rip and compare (null) them to confirm bit identical, then play them in turn on your choice of CD player. There’s just one condition: The CD player itself must be out of view of the listeners, maybe in another room. Play the original, then the CD-R. Repeat until everyone who hears a difference is familiar with the differences. Now stop telling them which disc is in the player, and play either disc at random 10 to 20 times. Keep track of their choices, and later correlate to the list of actual discs played. If there is an audible difference, the results should show it.
I look forward to this whole business. He assures me that the discs will be bit for bit identical…otherwise I can’t move forward. I’m thinking the best way to compare the two will be to have them both playing in identical Oppo BDP-95 machines connected via S/P DIF to two inputs on a Benchmark DAC2 HGC. Then simply switch from one iput to another with the disc started at roughly the same time.
You could also try with the Benchmark DAC 1 – where the jitter reduction isn’t as good –, or single box CD player with very poor jitter immunity. If you only try it with one set of components, you can’t extrapolate your results to every system out there.
On the other hand, if someone wanted more than 80 of their CDs copied, they could, instead, pay for a DAC that would render such tweaks useless, and apply to the entirety of their music collection.
I’m not so sure about having to find a circumstance where the jitter or other issues can affect things. My point is that identical disc streams played through identical systems cannot sound or perform differently.
All that your proposed test would prove is that playing those two discs through that particular system sounds the same, or doesn’t. There is no reason, from that test, to expect the same findings to hold for other systems.
Look at it the other way: If I run the test with the two discs on a CD player with a poorly isolated transport and no jitter reduction, and I hear, and measure, a difference, would that mean the difference would exist on all systems? It’s still the same stream of ones and zeros taking an identical path.
To be clear, I agree with your position that the discs would sound the same, so long as the clock is completely regenerated and the electrical system of the transport is completely isolated from the DAC. I also agree that such is the case in many high quality systems today – and I wouldn’t buy one where it wasn’t. I’m only trying to get you to consider that the system context can give an entirely plausible reason for the discs not to sound the same.
After all, DAC designers wouldn’t waste money and effort on jitter reduction/elimination, if all that mattered was getting the right bits to the DAC chip.
“All that your proposed test would prove is that playing those two discs through that particular system sounds the same, or doesn’t.” That’s exactly what we’re trying to show…that’s the claim from the people pitching treatments, illuminators, green markers, disc cutters, and making Gold CD-R copies. The process doesn’t include replacing your whole system with better components. Of course, that would result in a different sound.
The central premise of these snake oil peddlers is that their process will result in “dramatically improved” fidelity by purchasing their products. In fact, the CD Illumination guy insists that you won’t have to purchase any expensive cables or other equipment to realize the improvement.
The system context in this specific comparison is not relevant. We can explore other aspects of the playback chain in another evaluation but at this juncture, the bits vs bits thing is the only thing under consideration.
Mark, we’re talking at cross purposes.
If you have a playback system with poor jitter immunity, it is *possible* that this tweak will improve the sound. If your system has good jitter immunity, it’s highly unlikely you will hear any change.
Benchmark has put forth a lot of effort to make a DAC that is not very susceptible to jitter. If this is your only test device, all your test proves is that there is no performance difference between these discs, when played through a system with good jitter immunity. It does not prove the general case that these discs played through the same system – any system – will sound identical.
We call this the inference space of the experiment. Consider a pharmaceutical company that tests a drug in a very small, homogeneous population – e.g. 25 to 45-year-old males who work out at least three times a week. There would be a significant question as to whether the drug would have the same effect in females over 80 years old who have a variety of health concerns, or in children under the age of three.
It’s not that your test is, per se, bad. It just can’t be used to draw as broad a conclusion as you would like.
You WILL hear a difference when he does his “thing”. Here’s how: he is going to rip the songs you send, run those files and apply equalization. A boost here, a cut there, then burn the files onto his gold disc.
Why is he only stopping with gold discs? Platinum disc will have zero jitter, allow 32 bits of dynamic range with a frequency response from 0 hz to 150 kHz .
I’m not going to let him make any changes to the original other than copy it. He insists that the discs will be identical and I will be able to verify that. Regardless of whether I can refute his claims or not, he’ll never acknowledge that his process isn’t creating a perceived improvement. Just won’t happen.
First, I think you should send him the CD since that is what his business replicates. Something may get lost in the WAV translation.
Second, in your whole bits is bits insistence, I believe that your being a teacher plus your engineering background, you know, where everything is laid out in a hand book, or series of equations, may preclude your being able to consider other dimensions to issues. Remember, everything that we do, including what, or the way that we measure, is closely tied to our senses. It’s fine to know how things work, I seek this knowledge constantly, myself, but sometimes the same things work differently for other people. Engineers use to insist that wire is wire (no qualification for the wire was made); you as a scientist, however, are not necessarily included.
In a related question I posed a few days ago, I asked about the role of timing data in in the d/a process. If it’s so simple to inject new unjittered timing data into the process, why was timing data there to begin with? It was my impression that timing data indicated in what portion of a sampling interval the data was captured. What am I missing?
I will definitely be sending a brand new, unopened CD to Mark for him to copy and “process” to a Gold CD-R.
I recognize that I’m a bits are bits guy. But I’m also a musician, composer, and artist. My reaction to wonderfully composed, performed and recorded music can be very emotional. It’s common for tears to appear simply by listening to Jennifer Warnes sing “So Sad” or at during the finale of Beethoven’s 9th Symphony. But there is no magical, mystical process that can change the sound quality of a digital stream. If there is a difference in the sound, there has to be a reason. Perhaps its human perception…some sort of ear/brain neurological reaction. But a sequence of numbers is a sequence of numbers and we better be able to count on them producing the same results when applied in the same context. Imagine if the computers aboard the moon missions were subject to some sort of “fidelity” modification…the astronauts would have missed their target.
As to your previous question about clocking. Virtually every piece of equipment in my studio requires a steady clock to run. The console, the recorders, the DAW, the converters etc. There are three ways to get digital equipment to clock correctly. The first is with an internal clock…a simple component that allows the raw digital data stream to flow. The second method is to externally clock a device with a “master clock”. This is how my sertup is arranged. My equipment ignores the internal clocks (which are not great quality) and replaces them with a single highly accurate clock. I use an Antelope Atomic Clock…very expensive and very good. The third way is to send the clock down the digital connection in a daisy chain configuration. This is a typical method and results in jitter and timing problems.
Good to hear that you are using an outboard clock to help rid the digital stream of jitter. This may allow reasonable results from your fav oppo player. Transports do make a difference IMO, as do clocks. Have you tried the 47 Laboratory 4704 PiTracer CD Transport? What about the Sony’s CDP-R10 or Denon DP-S1 CD Transport? All of these offer a better replay IMV compared to the vast majority of alsoran players. I would venture to suggest you compare you favorite player without external clock against any one of these to hear the obvious differences (IMO). You are already wise to the reclocking and what this brings to the party, now venture to hear the error correction and focusing issues of ‘cheaper units’, subtle it is not and your B&W’s should easily resolve the differences. Here is a link to a web site that covers such issues http://www.hifi-advice.com/CD-Transports-info.html
I had a right good laugh when I saw all those JBL’s in the same studio as the B&W’s which the film producers (?)preferred for repo of what was it, movie soundtrack music replayed at very high levels; no great surprise to me as different beast the two speaker types with totally different design objectives.
I must admit that I don’t have a lot of experience with alternate transports. I have had a CD-101, the original professional CD players from Sony and I have a Meridian CD/DVD-Audio player as well.
The statement from your linked article continues the myth that digital somehow suffers from the same vagaries as analog, “I’ll say it right here: there are quite large differences in sound amongst CD transports. Yes, even when you only use the digital output, still you can hear differences in speed, attack, fluidity. It seems incredible, especially since many enigeers swear that this is impossible.”
Here’s those audiophile adjectives again. Differences in “speed, attack, fluidity”…come on. It seems incredible because it is. This is the very test that I’m going to perform. I’ll take the same digital stream from a variety of machines and compare the output…in a computer, not in my ears. If they are the same, the sound will be the same.
If by comparing in a computer, you simply mean capturing the bits, you’re still entirely missing the point. The differences in transport have to do with jitter, and capturing the bits to disk again, strips out the timing information. Consider using a jitter analyzer instead.
You can insist all you want that a high-end DAC is impervious to jitter. Many are. However, if that could be taken for granted, why would companies like Benchmark bother talking about their jitter elimination solutions?
Following your link to the Benchmark website, from your post on the 28th, I found a link to another by Mr. Siau on jitter, wherein he states “Most D/A converters are adversely affected by interface jitter.” He also explains that interface jitter is completely eliminated when the digital stream is recorded to a hard drive.
This article supports the points I have repeatedly tried to make to you, in response to a number of your posts. Namely:
1. You should not take total jitter elimination for granted, and
2. Trying to disprove differences in transport quality by capturing the output to a hard drive is a fruitless endeavor.
Andrea, I have acknowledged that jitter presents real issues for digital playback systems…but in the audiophile world it has been conquered.
As for you second point, I don’t understand how you’ve managed to turn the essential question of matching data streams sounding the same into differences in transport quality. That’s not the issue at hand.
Whether the difference is between transports, between discs being read by the same transport, between digital cables, or whatever, capturing the stream to a hard drive eliminates the interface jitter that is an, entirely reasonable, explanation for differences in sound quality. It also ignores any electrical noise that could make it from the transport into the circuitry of the DAC. By looking at what you capture to the hard drive, you are only looking at one aspect of the digital stream being fed to the DAC. If you really want to compare the signals, use an oscilloscope.
In the end, I just want to encourage you to think more broadly, and, if you want to debunk some claim, do the right test.
Your blog is, generally, a good source of information. That’s why I make the effort to try to correct these missteps – not that I don’t have anything better to do.
You’re right and I want to make this as rigorous as is possible. This is just a first phase of the overall comparison. IF the data isn’t the same then there’s not point in continuing.
OK thanks for your answers and Andreas input too. It is now clear what you are setting out to do. At this juncture I would say that merely sending bits to a pc drive that you will perceive no change in the bits, only missing ones. It’s when you replay the disc from various transports into a decent DAC that you will hear the changes IMO.
Then we get onto streaming replay software. There is a far from fair difference to which software you replay the stored data to output to the DAC, problems just keep piling up and it appears (to me) that every time the data is re handled and passed on that it is somehow tainted or corrupted at each transfer. Ignoring Jitter why is this so, a 64k question!
Hearing the PS audio ‘memory player’ improve the results must indicate that the read / play performance of transports is far from perfect but you have already acknowledged that transports do make a difference. Thanks for your honest replies.
The challenge is that simply copying a disc to a Gold CD-R will result in an audible change when played back in exactly the same system. I don’t believe it.