How Permanent is CD-R Media?

DEPARTMENTS
How Permanent is CD-R Media?
Understanding CD-R's Variables

HOME
COMMENTARY
LATEST PRODUCTS
ABOUT US
SUBSCRIBE
WEBLINKS
2001 MEDIA KIT BPA Audited
email editor
email for ad info
email webmaster

We're Part Of:
United Entertainment Media, Inc.
United News & Media by Jacob Trock

(March 2001)-Many estimates on the permanence of CD-ROM and CD-R have been made by the world's CD-R suppliers over the past several years. However, a lack of standards for testing has allowed some companies to skew their own results, raising questions of the validity of such tests, especially when comparing competitors.

In cooperation with the School of Conservation of Copenhagen, Denmark and CD testing equipment company AudioDev (www.audiodev.com), I set about to conduct my own tests on CD-R permanence.

Variables
CD-R media consists of five elements, all of which influence the permanence of the media. Polycarbonate belongs to the most stable family of polymeric plastics and is rarely a problem for permanence of compact discs. The production speed of CDs is sometimes so high that the disc leaves the molding machine before complete curing, causing inhibited stress that will influence the laser beam's ability to read the pits and lands (figure 1). A disc is normally pressed within 10 seconds.

The dye is the most fragile part of the CD-R disc. Four different kinds of dye are used today: cyanine, metal-stabilized cyanine, phthalocyanine and azo. Cyanine is the oldest formulated dye, and is patented by Taiyo Yuden. When the Orange Book, part II for CD-R media was written, it was the only dye on the market. This has of course given some technical advantages for the cyanine dye with regards to compatibility.

Cyanine has unfortunately shown to be less stable than the newer dyes for CD-R. In 1996 TDK presented its metal-stabilized cyanine dye, which is demonstrably more stable than original cyanine. It is almost impossible to distinguish between metal-stabilized cyanine and cyanine (both look blue-green).

The phthalocyanine dye, patented by Mitsui, is regarded as the most stable dye, but also more sensitive to particular types of recorders. The color of phthalocyanine is pale blue. The last type of dye is the bright blue azo, patented by Verbatim/Mitsubishi. Results from the companies' own tests with azo-discs indicate compatibility problems with some recorders. All dyes, but cyanine dyes in particular, change in three ways during aging. First, contrast between the pits and lands changes: this has influence on the reflectivity, which in turn causes interpretation problems. Second, the shapes of pits and lands become smaller and longer causing higher jitter. Third, bleeding of the dye from neighboring grooves causes higher crosstalk.

For the reflective layer in CD-R, gold has been the most common metal used to date. While this inert metal does not degrade, the thin reflective layer is very vulnerable towards scratches and dust particles trapped between it and the dye layer. The new silver halide reflective layer type is now in many cases replacing the gold layer, but only for one reason: it is cheaper. Lastly, the protective lacquer and the label can perform damage to the data, due to contraction during drying after spin-coating/printing in production. Surface presentations of radial noise (RN) after aging clearly illustrate how the label affects the data. High RN can cause problems with tracking.

Analyzing the Disc
Testing CD-R media is not possible for everyone. Professional equipment for testing CD-R media is costly, and it takes time to become familiar with 80-plus parameters that are used to analyze the disc.

The standard parameter for testing compact discs is block error rate (BLER), the number of errors that occur per second at the first level of error correction. This parameter is very useful when analyzing changes on a disc, because it incorporates all kinds of defects. On the other hand, BLER is almost useless if you want to find out why errors occur. For that, you need to look closer at the numerous parameters themselves.

Test results from AudioDev's SA3 Advanced system give users the ability to get a very detailed picture of critical quality characteristics. Users are able to tell if the recorder is the problem, or if there are problems with the dye, the reflective layer or the polycarbonate. Testing compact discs gives one other important piece of information like the "disc manufacturing code," which reveals who made the disc, and optimal recording power (ORP), which specifies how many milliwatts (mW) the laser beam shall use for its initial optimum power calibration (OPC) before recording onto the disc. Some low-end CD-R recorders use solely the ORP data during the whole writing session, without completing an initial OPC or the following running OPC measurement.

Compatibility tests between CD-R media and CD-R recorders are not very common. OSTA (the Optical Storage Technology Association) has completed two compatibility tests in which no problems with compatibility were found. Doculabs arrived at almost the same results in its own compatibility tests, conducted in collaboration with National Media Lab. I performed my own tests, analyzing data from 30 discs (five different brands) that were recorded on 10 different recorders. Some tests were double tests, giving a total of 20 different combinations for the entire test.

In testing, BETA measurements show as a percentage the difference between pits and lands on the disc. In doing so, BETA tells immediately if the recorder has a functional running OPC. All discs I tested had a BETA variation over 2 percent-the limit in the Orange Book, part II specification (figure 2).

Too low or high a BETA can very well be the reason why a disc can be read in one disc drive and not in another. In the worst case, the disc won't be readable in any drive at all even if there are no block errors. This illustrates the complexity of analyzing permanence of compact discs and the reason why the conclusion of this test deviates from the other compatibility tests. The conclusions of a particular compatibility test depend on the parameters the test focuses on and what weight the results from each tested parameter are given.

While compatibility has improved since my first tests were performed back in the winter of 1996, the constant emergence of new recorders makes compatibility tests like this almost outdated for use as a tool for selecting the best recorder and media.

The BETA range for optimum recording is different for cyanine and phthalocyanine dye. In order to obtain the best "dialogue" between the recorder and disc, the recorder needs to know something about what kind of dye it's recording to. The latest revision of the Orange Book, part II (version 3.0) describes how information shall be placed in the disc's ATIP signal, so that the recorder can discern what kind of dye the disc features, and what BETA-range is accordingly recommended. This will surely reduce compatibility problems.

Does Writing Speed
Influence Data Integrity? In 1995, Mike Martin & Jack Hyon of the Jet Propulsion Laboratory in Pasadena, CA came to the conclusion that the phthalocyanine discs of the day were not compatible with 1X writing speed, and that cyanine discs were incompatible with 4-6X writing speed. A small test was made to see if phthalocyanine has since become 1X compatible, since almost all phthalocyanine discs are labeled as such. A top-brand disc was recorded in four sessions with alternately 2X and 1X speed. The disc was analyzed before and after two months in a climate chamber maintained at 85 percent relative humidity (RH) and 65 degrees Celsius. The result was very clear: phthalocyanine is not compatible with 1X writing speed.

It is not possible to give a detailed explanation of why writing speed at 1X gave many errors. That writing speed is a problem with phthalocyanine discs is illustrated by the increase of BLER from the inner to the outer part of the disc, where the rotation speed slows down from 600rpm to 230rpm.

Is one disc representative in aging tests? When you have limited time and resources to make accelerated aging tests, you have to find shortcuts and methods to make End of Life (EOL) estimates on a single disc. For me, four observations came in handy:

1. Measuring a disc does not give a single result for each parameter. Middle values from a tested disc represent about 4500 single measurements-a very clear, normal distribution with possibilities to use the data in statistical calculations, like standard deviations.

2. Many of my double tests in compatibility tests and accelerated aging tests were identical when comparing the middle values.

3. Analyzing the results from existing accelerated aging tests on CD-R media indicate that compact discs in many aspects act the same way as with other archival materials like paper, photographs, film etc. Assuming this is true gives the opportunity to use Time Weighted Preservation Index (TWPI) on CD-R media, described by James M. Reilly, Douglas W. Nishimura and Edward Zinn in their 1995 paper for the Commission on Preservation and Access in Washington, DC.

4. Comparing the change in the BLER average (avg) with the change in other parameters' middle value demonstrated that BLERavg is representative when analyzing the stability of compact discs.

All four observations contain plenty of reservations and exceptions. On the other hand, a qualified guess is far better than just a guess. With that in mind, I started to select discs for accelerated aging tests, following, as close as possible, the descriptions from ANSI/NAPM IT9.21. Eighteen discs representing 11 different disc labels (brand/type) were exposed to 65 degrees Celsius/85 percent RH for two months (such a timeframe at 22 degrees Celsius/55 percent RH is regarded to equal 45 years). The EOL limit for the tested discs was set to be the BLERavg as defined by the industry.

Four of eleven disc labels showed almost no change in BLERavg during aging. These discs employed phthalocyanine or metal stabilized cyanine dye. Four other discs changed during aging in a way that made linear regression of BLERavg possible. The last three disc labels exhibited changes in BLERavg that excluded them from use in linear regression analysis. Some of the tested discs were tentative discs, delivered directly from CD plants, and not yet ready for consumers. These tentative discs (products c and d in figure 3) were the most unstable. Both discs turned out to have dust particles between the polycarbonate and the reflective layer, causing pinholes in the reflective layer.

Conclusion
Phthalocyanine and metal-stabilized cyanine are very stable. Using a recorder that is compatible with a phthalocyanine/metal-stabilized cyanine disc and recording at a speed 2X or 4X gives the possibility to create discs that will last for more than a century. Compatibility tests are highly recommended if the recorded discs are intended for archival purpose. Using the many parameters that follow from testing gives rich possibilities to find compatibility problems and, in connection with accelerated aging testing, the possibility to find the causes for instability.

Jacob Trock has studied CD-R's permanence under the Royal Danish Academy of Fine Arts, School of Conservation. For a complete list of references, please e-mail Terence Keegan at tkeegan@uemedia.com.
BACK TO TOP

commentary | latest products
web links | search | subscribe
media kit | editor | webmaster | ad info
All pages copyright 1998 - 2001, United Entertainment Media, Inc.
All Rights Reserved.

DEPARTMENTS		How Permanent is CD-R Media? Understanding CD-R's Variables
HOME COMMENTARY LATEST PRODUCTS ABOUT US SUBSCRIBE WEBLINKS 2001 MEDIA KIT BPA Audited email editor email for ad info email webmaster We're Part Of: United Entertainment Media, Inc. United News & Media		by Jacob Trock (March 2001)-Many estimates on the permanence of CD-ROM and CD-R have been made by the world's CD-R suppliers over the past several years. However, a lack of standards for testing has allowed some companies to skew their own results, raising questions of the validity of such tests, especially when comparing competitors. In cooperation with the School of Conservation of Copenhagen, Denmark and CD testing equipment company AudioDev (www.audiodev.com), I set about to conduct my own tests on CD-R permanence. Variables CD-R media consists of five elements, all of which influence the permanence of the media. Polycarbonate belongs to the most stable family of polymeric plastics and is rarely a problem for permanence of compact discs. The production speed of CDs is sometimes so high that the disc leaves the molding machine before complete curing, causing inhibited stress that will influence the laser beam's ability to read the pits and lands (figure 1). A disc is normally pressed within 10 seconds. The dye is the most fragile part of the CD-R disc. Four different kinds of dye are used today: cyanine, metal-stabilized cyanine, phthalocyanine and azo. Cyanine is the oldest formulated dye, and is patented by Taiyo Yuden. When the Orange Book, part II for CD-R media was written, it was the only dye on the market. This has of course given some technical advantages for the cyanine dye with regards to compatibility. Cyanine has unfortunately shown to be less stable than the newer dyes for CD-R. In 1996 TDK presented its metal-stabilized cyanine dye, which is demonstrably more stable than original cyanine. It is almost impossible to distinguish between metal-stabilized cyanine and cyanine (both look blue-green). The phthalocyanine dye, patented by Mitsui, is regarded as the most stable dye, but also more sensitive to particular types of recorders. The color of phthalocyanine is pale blue. The last type of dye is the bright blue azo, patented by Verbatim/Mitsubishi. Results from the companies' own tests with azo-discs indicate compatibility problems with some recorders. All dyes, but cyanine dyes in particular, change in three ways during aging. First, contrast between the pits and lands changes: this has influence on the reflectivity, which in turn causes interpretation problems. Second, the shapes of pits and lands become smaller and longer causing higher jitter. Third, bleeding of the dye from neighboring grooves causes higher crosstalk. For the reflective layer in CD-R, gold has been the most common metal used to date. While this inert metal does not degrade, the thin reflective layer is very vulnerable towards scratches and dust particles trapped between it and the dye layer. The new silver halide reflective layer type is now in many cases replacing the gold layer, but only for one reason: it is cheaper. Lastly, the protective lacquer and the label can perform damage to the data, due to contraction during drying after spin-coating/printing in production. Surface presentations of radial noise (RN) after aging clearly illustrate how the label affects the data. High RN can cause problems with tracking. Analyzing the Disc Testing CD-R media is not possible for everyone. Professional equipment for testing CD-R media is costly, and it takes time to become familiar with 80-plus parameters that are used to analyze the disc. The standard parameter for testing compact discs is block error rate (BLER), the number of errors that occur per second at the first level of error correction. This parameter is very useful when analyzing changes on a disc, because it incorporates all kinds of defects. On the other hand, BLER is almost useless if you want to find out why errors occur. For that, you need to look closer at the numerous parameters themselves. Test results from AudioDev's SA3 Advanced system give users the ability to get a very detailed picture of critical quality characteristics. Users are able to tell if the recorder is the problem, or if there are problems with the dye, the reflective layer or the polycarbonate. Testing compact discs gives one other important piece of information like the "disc manufacturing code," which reveals who made the disc, and optimal recording power (ORP), which specifies how many milliwatts (mW) the laser beam shall use for its initial optimum power calibration (OPC) before recording onto the disc. Some low-end CD-R recorders use solely the ORP data during the whole writing session, without completing an initial OPC or the following running OPC measurement. Compatibility tests between CD-R media and CD-R recorders are not very common. OSTA (the Optical Storage Technology Association) has completed two compatibility tests in which no problems with compatibility were found. Doculabs arrived at almost the same results in its own compatibility tests, conducted in collaboration with National Media Lab. I performed my own tests, analyzing data from 30 discs (five different brands) that were recorded on 10 different recorders. Some tests were double tests, giving a total of 20 different combinations for the entire test. In testing, BETA measurements show as a percentage the difference between pits and lands on the disc. In doing so, BETA tells immediately if the recorder has a functional running OPC. All discs I tested had a BETA variation over 2 percent-the limit in the Orange Book, part II specification (figure 2). Too low or high a BETA can very well be the reason why a disc can be read in one disc drive and not in another. In the worst case, the disc won't be readable in any drive at all even if there are no block errors. This illustrates the complexity of analyzing permanence of compact discs and the reason why the conclusion of this test deviates from the other compatibility tests. The conclusions of a particular compatibility test depend on the parameters the test focuses on and what weight the results from each tested parameter are given. While compatibility has improved since my first tests were performed back in the winter of 1996, the constant emergence of new recorders makes compatibility tests like this almost outdated for use as a tool for selecting the best recorder and media. The BETA range for optimum recording is different for cyanine and phthalocyanine dye. In order to obtain the best "dialogue" between the recorder and disc, the recorder needs to know something about what kind of dye it's recording to. The latest revision of the Orange Book, part II (version 3.0) describes how information shall be placed in the disc's ATIP signal, so that the recorder can discern what kind of dye the disc features, and what BETA-range is accordingly recommended. This will surely reduce compatibility problems. Does Writing Speed Influence Data Integrity? In 1995, Mike Martin & Jack Hyon of the Jet Propulsion Laboratory in Pasadena, CA came to the conclusion that the phthalocyanine discs of the day were not compatible with 1X writing speed, and that cyanine discs were incompatible with 4-6X writing speed. A small test was made to see if phthalocyanine has since become 1X compatible, since almost all phthalocyanine discs are labeled as such. A top-brand disc was recorded in four sessions with alternately 2X and 1X speed. The disc was analyzed before and after two months in a climate chamber maintained at 85 percent relative humidity (RH) and 65 degrees Celsius. The result was very clear: phthalocyanine is not compatible with 1X writing speed. It is not possible to give a detailed explanation of why writing speed at 1X gave many errors. That writing speed is a problem with phthalocyanine discs is illustrated by the increase of BLER from the inner to the outer part of the disc, where the rotation speed slows down from 600rpm to 230rpm. Is one disc representative in aging tests? When you have limited time and resources to make accelerated aging tests, you have to find shortcuts and methods to make End of Life (EOL) estimates on a single disc. For me, four observations came in handy: 1. Measuring a disc does not give a single result for each parameter. Middle values from a tested disc represent about 4500 single measurements-a very clear, normal distribution with possibilities to use the data in statistical calculations, like standard deviations. 2. Many of my double tests in compatibility tests and accelerated aging tests were identical when comparing the middle values. 3. Analyzing the results from existing accelerated aging tests on CD-R media indicate that compact discs in many aspects act the same way as with other archival materials like paper, photographs, film etc. Assuming this is true gives the opportunity to use Time Weighted Preservation Index (TWPI) on CD-R media, described by James M. Reilly, Douglas W. Nishimura and Edward Zinn in their 1995 paper for the Commission on Preservation and Access in Washington, DC. 4. Comparing the change in the BLER average (avg) with the change in other parameters' middle value demonstrated that BLERavg is representative when analyzing the stability of compact discs. All four observations contain plenty of reservations and exceptions. On the other hand, a qualified guess is far better than just a guess. With that in mind, I started to select discs for accelerated aging tests, following, as close as possible, the descriptions from ANSI/NAPM IT9.21. Eighteen discs representing 11 different disc labels (brand/type) were exposed to 65 degrees Celsius/85 percent RH for two months (such a timeframe at 22 degrees Celsius/55 percent RH is regarded to equal 45 years). The EOL limit for the tested discs was set to be the BLERavg as defined by the industry. Four of eleven disc labels showed almost no change in BLERavg during aging. These discs employed phthalocyanine or metal stabilized cyanine dye. Four other discs changed during aging in a way that made linear regression of BLERavg possible. The last three disc labels exhibited changes in BLERavg that excluded them from use in linear regression analysis. Some of the tested discs were tentative discs, delivered directly from CD plants, and not yet ready for consumers. These tentative discs (products c and d in figure 3) were the most unstable. Both discs turned out to have dust particles between the polycarbonate and the reflective layer, causing pinholes in the reflective layer. Conclusion Phthalocyanine and metal-stabilized cyanine are very stable. Using a recorder that is compatible with a phthalocyanine/metal-stabilized cyanine disc and recording at a speed 2X or 4X gives the possibility to create discs that will last for more than a century. Compatibility tests are highly recommended if the recorded discs are intended for archival purpose. Using the many parameters that follow from testing gives rich possibilities to find compatibility problems and, in connection with accelerated aging testing, the possibility to find the causes for instability. Jacob Trock has studied CD-R's permanence under the Royal Danish Academy of Fine Arts, School of Conservation. For a complete list of references, please e-mail Terence Keegan at tkeegan@uemedia.com. BACK TO TOP commentary \| latest products web links \| search \| subscribe media kit \| editor \| webmaster \| ad info All pages copyright 1998 - 2001, United Entertainment Media, Inc. All Rights Reserved.

How Permanent is CD-R Media? Understanding CD-R's Variables

How Permanent is CD-R Media?
Understanding CD-R's Variables