Just when you thought it was safe to put green paint around the edges of your CDs without ridicule, there's yet another CD tweak that's sure to bring howls of laughter from the skeptics: cryogenically freezing CDs. They won't be laughing for long, however, when they hear for themselves the sonic results of this process.
Ed Meitner, designer of the Museatex line of electronics, has discovered that cryogenically freezing a CD changes the physical structure of polycarbonate, the plastic material from which CDs are made. The result is reportedly an audible improvement in sound quality. In this process, CDs are placed in a cryogenic freezing chamber and the temperature is slowly reduced over eight hours to 75 Kelvins, or about -300 degrees Fahrenheit. This is approximately the temperature of liquid nitrogen, the chamber's cooling agent. The temperature is then slowly brought back to room temperature over another eight hours.
This technique reportedly relaxes the lattice structure of a material (polycarbonate in the case of CD) that has been previously distorted by heat or pressure, both of which are present during CD injection molding. By reducing the molecular bonds holding the material together, the internal stress in the material is reduced, thus changing its resonant characteristics. Indeed, a treated disc feels slightly more flexible than an untreated disc.
But how could freezing a CD possibly affect its sound quality? So what if the polycarbonate has a different structure? The data are all ones and zeros. Furthermore, uncorrected data errors are almost nonexistent in most discsci without treatment, ruling out improved data integrity as an answer. I posed these questions to Ed Meitner and got the following explanation (footnote 1).
Mechanical vibration of the disc causes the HF signal to become noisy and have excessive jitter. The HF signal is the raw signal output from the CD player's photodetector (footnote 2). By freezing a CD, the disc's mechanical resonance is lowered, improving the quality of the HF signal retrieved from the disc. Although theory states that noise and jitter in the HF signal will have no effect on sound quality—the HF signal is squared, buffered, decoded, filtered, and clocked out of another buffer with quartz-crystal accuracy—many digital designers maintain that HF signal quality does affect the sound.
Ed Meitner claims that the HF signal improvement from a cryogenically treated disc is easily measurable. I looked at the HF signal on an oscilloscope from the Esoteric P2 transport with treated and untreated discs. I could see no difference in the signal quality. However, it is very difficult to make comparisons without seeing the two HF signals side by side.
Meitner is talking to some audiophile labels about mass-treating their releases. Apparently, the process is efficient and economical, with the ability to treat thousands of discs at once. Liquid nitrogen, which doesn't come in direct contact with the CDs, is inexpensive and readily available. Interestingly, this process is said to yield similar sonic improvements with a vinyl phonograph record. In addition to CDs and LPs, the process has been used on LaserVision-format video discs, speaker cable, interconnects, integrated circuits, and musical instrument strings.
Cryogenic freezing is also used to treat machine tools like drill bits, copper welding tools, and saw blades. The process reportedly improves their wear characteristics, thus extending the tool's useful life. The treatment doesn't always work, however, and there is no consensus among metallurgists that the process is always beneficial. In fact, the effects of cryogenically freezing materials is not well understood; little scientific research has been done to explain the phenomenon (footnote 3).
Another tweak developed by Ed Meitner is painting a CD's top surface black. This reportedly improves sound quality by improving the signal at the CD player's photodetector. Before describing how this works, let's look at the playback laser beam's path through the disc.
The playback beam enters the disc through the surface without the label. It travels through the 1.2mm disc thickness where it encounters pits impressed in the polycarbonate. To reflect the beam back through the disc and to the photodetector, a thin layer of aluminum is deposited on the disc surface, which conforms to the pit structure. A protective coating of varnish seals in the aluminum and prevents it from oxidizing. The label is then silk-screened on top of the protective coating.
Ed Meitner contends that several mechanisms are at work that degrade the HF signal picked up by the photodetector. One phenomenon is distortion of the aluminum layer by the laser beam's heat. Even though the beam is very low-power—about half a milliwatt—it is focused on such a small area (1.5µm) that the aluminum molecules bend, causing the aluminum layer to flex. This introduces jitter in the HF signal as well as noise in the focus signal.
This phenomenon has reportedly been measured by painting black bars on a CD's top surface (the label side) and looking at various signals. The bar pattern is readily apparent in both the focus servo and HF signals. Painting the CD black reportedly improves the thermal conditions by reducing the contrast in the aluminum molecules caused by laser-induced heat. Another mechanism that is also affected by black paint is the secondary reflection from the disc label. Some laser light passes through the aluminum layer and is reflected to the photodetector by the label. This reportedly causes noise in the HF signal which is manifested as uncertainty in the digital code transitions. Note that the above descriptions are those of Ed Meitner, and have not been independently verified.
How plausible are these explanations? I find some of them hard to believe, especially this last phenomenon. However, there is so much going on in digital audio that we don't know about—especially the optical considerations in data retrieval from CD—that I hesitate to rule out anything (footnote 4).
What really matters is if these treatments work. Since I believe that the ear is the highest-resolution instrument available to explore these phenomena, I gave Ed Meitner three copies of the Stereophile Test CD for treatment. One disc was cryogenically frozen, another was painted black, and the third was both painted and frozen. The Stereophile Test CD is ideal for this purpose: it has a wide variety of music, all recorded by Stereophile contributors. In addition, I know with absolute certainty that the three treated discs as well as my untreated control disc were made by JVC from the same master tape and CD stamper.
I began by listening to my guitar and bass recording from the untreated disc. After switching to the frozen and painted disc, the difference was immediate and obvious. First, the guitar appeared to become louder, with more clarity and detail. Subtle sounds like finger noises and minute instrumental detail jumped forward. The sonic picture became more vivid and immediate. The acoustic bass took on a more rounded character and its musical contribution seemed enhanced. There was a greater degree of air and life around the instruments; they suddenly became more palpable.
The degree to which these characteristics were apparent varied considerably with the type of music. During our annual Stereophile writers' conference in early August, I had an opportunity to play treated and untreated discs for some of the visiting writers. Arnis Balgalvis correctly identified the treated disc in a blind A/B/A comparison when he visited my listening room. He immediately knew that presentation B was different, and his description of the difference was remarkably similar to my impressions.
I repeated the blind test for Peter Mitchell in JA's listening room; Peter also immediately identified the treated disc. In fact, within seconds of hearing the treated disc with the guitar and bass recording, he let out a loud exclamation of surprise. His impressions were consistent with the differences I had heard, which I related to him after the test and his description to me of the differences.
A good point Peter raised was that although there was clearly a difference, he had doubts about which was "better" or more true to the original recording. The treated disc had a brighter, more detailed character that would exacerbate many of CD's problems.
The above listening comparisons were made between an untreated disc and one that was both cryogenically frozen and painted black. Further listening of frozen-only discs and painted-only discs revealed that most of the sonic difference was the result of freezing. The black paint, however, did add to the effect. A second frozen and painted disc sent to me by Museatex had similar differences. However, a look at the disc's inside ring, where the production number is written, revealed that it was a different pressing from my untreated control disc. I would therefore refrain from reaching any conclusions based on this disc. From my experience with the Stereophile Test CD, however, I am convinced that some unexplained phenomena are occurring in frozen and painted CDs.
http://www.stereophile.com/asweseeit/822/index1.html
