Optical disc rot: what public ageing studies measured, which error counters (BLER, PI sum 8) warn first, and what they cannot predict
Published 2026-10-12
Recordable discs fail mainly because the organic dye layer breaks down or, on discs with a silver reflective layer, the metal corrodes, and the public tell-tale is a rising correctable-error counter before data is lost: block error rate (BLER, counted at the first CD decoder) against a specified maximum of 220 per second on CD, and parity-inner errors summed over eight blocks (PI sum 8) against a maximum of 280 on DVD. In the NIST and Library of Congress study of discs bought in 2005, all 7 CD-R products reached an estimated life of more than 30 years at 25 C and 50% relative humidity (5 of 7 more than 45 years), while the 15 DVD products spread from under 15 years (2 products) to more than 45 years (7 products). Canada's conservation institute ranks a gold-layer phthalocyanine CD-R above 100 years and a dye or non-dye single-layer Blu-ray recordable at 5 to 10 years, and says discs recorded with poor technique can fail in as little as 2 to 10 years. These are accelerated-ageing estimates of temperature and humidity effects only, for disc products of 2005 or earlier, so they rank media and set expectations but are not a failure rate for any disc in use today.
- Failure mechanisms named
- Dye layer breakdown (NIST); corrosion of silver or silver-alloy reflective layers, called laser rot (CCI); possible adhesive separation in bonded DVD discs (CCI, anecdotal, and seen in elevated-temperature ageing).
- Warning counters
- CD: BLER, limit 220 per second, uncorrectable E32 allowed 0. DVD: PI sum 8, limit 280, uncorrectable PO errors allowed 0. Blu-ray: LDC and BIS (CCI).
- Largest sample
- NIST and Library of Congress: 7 CD-R and 15 DVD products, 90 discs per product, discs bought in 2005.
- Not covered
- Field failure rates, mass-produced read-only discs beyond the CCI ranking, M-DISC and other vendor claims, and any recovery of damaged discs.
Error counters and the limits the studies and the conservation note use
| Counter or parameter | Unit | Limit or target | What it is | Cross-check (second source) | Source |
|---|---|---|---|---|---|
| CD block error rate (BLER), maximum anywhere on disc | blocks per second | 220 maximum | Blocks with a detected error at the input of the first (C1) decoder, before correction | CCI also gives 220 as the specified maximum | NIST/LoC final report |
| DVD parity-inner errors, summed over 8 ECC blocks (PI sum 8) | errors per 8 blocks | 280 maximum | Rows with at least one byte in error before first-pass correction | CCI gives 280 for PI8 | NIST/LoC final report |
| Uncorrectable CD errors (E32) and DVD parity-outer errors (PO) | count | 0 allowed by the specifications | Data the decoder could not correct | The 2004 NIST paper states the same zero allowance | NIST 2004 paper |
| Starting error rate NIST accepted for ageing tests | BLER per second (CD); PI errors (DVD) | BLER under 40; PIE under 100 | Discs with higher initial errors were left out of the ageing study | CCI recommends under 50 for both at recording | NIST SP 500-263 |
| Recommended recorded error rate | BLER per second; PI8 errors | under 50 (maximum and average) | A low start leaves room for later errors before the limit is reached | NIST used stricter inclusion limits (40 and 100) | CCI Notes 19/1 |
| Blu-ray recordable: LDC and BIS | count (LDC, BIS) | average LDC under 13; average BIS under 0.26; maximum BIS under 15 | Long-distance-code and burst-indication-subcode counters | No second source for these Blu-ray limits was opened | CCI Notes 19/1 |
| Jitter limit | nanoseconds (CD); percent (DVD) | 35 ns maximum (CD); 9% maximum (DVD) | Timing spread of pit and land edges; the 2004 paper saw BLER jumps when CD jitter reached about 50 ns | NIST also reports jitter rising with PI errors on DVD-R | NIST 2004 paper |
NIST and Library of Congress 2005 discs: how many products fell in each life band
| Estimated life band | CD-R products (of 7) | DVD products (of 15) | DVD share (%, computed) | Cross-check (second source) | Source |
|---|---|---|---|---|---|
| More than 45 years | 5 | 7 | 46.7 | CCI ranks gold-layer phthalocyanine CD-R above 100 years | NIST/LoC final report |
| 30 to 45 years | 2 | 4 | 26.7 | 2004 NIST paper: stable samples expected to last 'several tens of years' | NIST/LoC final report |
| 15 to 30 years | 0 | 2 | 13.3 | CCI ranks silver-layer DVD+R at 20 to 50 years | NIST/LoC final report |
| Under 15 years | 0 | 2 | 13.3 | CCI ranks DVD-RW and dual-layer DVD+R at 5 to 10 years | NIST/LoC final report |
2004 NIST harsh-condition runs: how fast error counters rose in the least and most stable samples
| Sample and stress | Time | Counter reading | Unit | Cross-check (second source) | Source |
|---|---|---|---|---|---|
| CD-R S4 (phthalocyanine, silver-gold alloy), extreme temperature and humidity | 600 hours | BLER under 400 | blocks per second | CCI also ranks phthalocyanine CD-R as the most stable class | NIST 2004 paper |
| All other CD-R samples, same stress | within 400 hours | BLER over 600 | blocks per second | CCI: azo or cyanine dye with silver alloy is rated lower (20 to 50 years) | NIST 2004 paper |
| CD-R S2 (phthalocyanine, darkened polycarbonate), extreme temperature and humidity | within 150 hours | BLER over 1,000 | blocks per second | Same sample stayed stable beyond 1,400 h of light, so one dye type does not guarantee stability | NIST 2004 paper |
| DVD-R D1 and D3, extreme temperature and humidity | within 200 hours | PI sum 8 about 1,000 | errors per 8 blocks | CCI Notes: DVD-R stability ranges from very good to poor by metal layer | NIST 2004 paper |
| DVD-R D2, same stress | beyond 400 hours | PI sum 8 remained very low | errors per 8 blocks | NIST 2005: 7 of 15 DVD products over 45 years | NIST 2004 paper |
| DVD-R D1 and D3, metal-halide light | 800 hours | PI sum 8 about 1,500 | errors per 8 blocks | CCI: light damage is avoided if discs go back into cases promptly | NIST 2004 paper |
Canadian Conservation Institute ranking of average longevity by format and layer
| Disc format | Average longevity (years) | Metal layer or dye | Cross-check (second source) | Source |
|---|---|---|---|---|
| CD-R, phthalocyanine dye | more than 100 | gold | NIST 2005: 5 of 7 CD-R products over 45 years (a lower bound, not a contradiction) | CCI Notes 19/1 |
| CD-R, phthalocyanine dye | 50 to 100 | silver alloy | NIST 2004: S4 with a silver-gold alloy layer was the most stable CD-R | CCI Notes 19/1 |
| DVD-R | 50 to 100 (gold) or 10 to 20 (silver alloy) | gold or silver alloy | NIST 2005: 2 of 15 DVD products under 15 years | CCI Notes 19/1 |
| CD-RW and DVD+RW (erasable) | 20 to 50 | phase-change layer | NIST 2005 could not test CD-RW | CCI Notes 19/1 |
| BD-R (dye or non-dye, single or dual layer) | 5 to 10 | not applicable | No second source for Blu-ray was opened | CCI Notes 19/1 |
| Read-only DVD and Blu-ray movie discs | 10 to 20 | not stated | No second source opened | CCI Notes 19/1 |
Reading the numbers
The decoder counters are the warning signal, and their limits are public. A CD is within specification up to a maximum BLER of 220 per second and a DVD up to a PI sum 8 of 280, and both specifications allow no uncorrectable errors (E32 on CD, PO on DVD). NIST ran its ageing tests until those correctable counts approached the limits, because a rising BLER or PI count precedes data loss; in its 2004 runs BLER jumped when CD jitter reached about 50 ns and E32 errors appeared with the BLER rise.
Where the counters start matters as much as how fast they rise. NIST would only age discs recorded with BLER under 40 or PIE under 100, and the Canadian note recommends recording to under 50 on both; a disc recorded close to the limit has no margin, so a scratch or small amount of ageing is enough to push it over. The note names improper recording as one of the main causes of early failure, and says discs that ignore its guidelines can fail in as little as 2 to 10 years.
Product choice dominated the NIST results, not disc type. Of the 15 DVD products, 7 were estimated at more than 45 years and 2 at under 15, so a count of 'DVD' tells little. All 7 CD-R products were estimated above 30 years. The 2004 paper explains part of the spread: CD-R discs with phthalocyanine dye did best, azo dye did worse, and cyanine dye held up under light but failed quickly in heat and humidity. It also found one phthalocyanine disc that was stable in light but failed in 150 hours of extreme heat and humidity, so dye type alone is not a safe predictor.
The metal layer is the second failure point. The Canadian note says gold is not at risk of laser rot, while silver and silver alloys are susceptible to corrosion, especially when exposed to pollutants, and rates a silver-alloy DVD-R at 10 to 20 years against 50 to 100 for a gold one. NIST's 2004 paper adds that a faded polycarbonate layer may leave the data intact while a degraded dye layer cannot be read back, which is why the dye is called the most important layer.
Light and heat shorten life quickly, but are easy to avoid. The 2004 paper reports that some discs could fail from direct sunlight in as little as a few weeks, while CD-R samples S2 and S4 stayed stable beyond 1,400 hours of metal-halide light and others rose sharply within 800 hours. The Canadian note recommends 20% to 50% relative humidity (never below 10%), a temperature from minus 10 to 23 C (never above 32 C), and a test of representative discs every 5 to 10 years, more often if storage is poor.
What these sources do not give is a field failure rate. The NIST and Library of Congress report calls its results a snapshot of 2005 products and says model estimates may not match today's media. The longest lives are reported only as a lower bound. No natural-ageing retest of the 2005 discs was among the documents opened here. Blu-ray recordable discs appear in this evidence only through the Canadian note's own summary, which rates them lower than CD-R and DVD.
Related: what smartctl and smartd can and cannot see and what fails most often on HDDs, SSDs and links.
Method
Everything on this page was read on 2026-10-12 from documents opened that day: the final report of the NIST and Library of Congress optical disc longevity study (NIST SP 500-263 procedure and the September 2007 final report PDF hosted by the Library of Congress); the 2004 NIST paper 'Stability Comparison of Recordable Optical Discs' in the Journal of Research of NIST; the NIST SP 500-263 procedure text (the NIST-hosted PDF contains the test procedure and definitions but not the results); and Canadian Conservation Institute Notes 19/1 'Longevity of Recordable CDs, DVDs and Blu-rays' (Iraci, revised 2019). The two NIST documents and the Library of Congress report come from one research collaboration and are not independent of each other; the Canadian Conservation Institute note is a separate institution, but it summarises its own and earlier work. Values are copied as printed; counts of products in each life band are read from the printed text of the final report, not from its figures. No disc was read or tested for this page.
Limits
The NIST and Library of Congress life expectancies are model estimates, not observed failures: discs were aged at 60 to 80 C and 70 to 85% relative humidity and extrapolated with an Eyring model to 25 C and 50% relative humidity, with 95% survival at 95% confidence. Only temperature and humidity were modelled; light, corrosive gases, contaminants and handling were not. The media were bought in 2005, the report says results are a snapshot and may not reflect current products, brand names were withheld, and for the longest-lived products no exact figure was determined, only 'more than 45 years'. CD-RW could not be tested. The report notes that the link between measured BLER or PI errors and real data loss depends on the playback system as well as the media, and the 2004 paper says that with its small sample it could not identify a PI error value at which uncorrectable outer-parity errors begin. The 2004 paper's 'extreme' stress and light results are continuous exposure and do not represent normal storage, and its sample sets were too small for a life estimate. The Canadian ranking table is stated to be based on temperature and humidity only and would be lower for silver-layer discs in polluted air; its Blu-ray statement rests on a study (Iraci 2018) that this page did not open, so the Blu-ray rows are the note's own summary. No source opened here gives a field failure rate for discs in real archives, a count of how many drives log or expose C1/C2 or PI/PO counters, or a threshold at which a consumer drive's own reading becomes unreliable beyond the Canadian note's remark that free software results are specific to the drive used and report mainly uncorrectable errors. The page covers public ageing measurements and error-counter definitions only: nothing here concerns protected discs, bypassing copy or access protection, or recovering anyone else's media.
Sources
- 01NIST / Library of Congress Optical Disc Longevity Study, final report (September 2007 PDF, Library of Congress) · accessed 2026-10-12
- 02NIST Special Publication 500-263: NIST/Library of Congress Optical Media Longevity Study (procedure, November 2005) · accessed 2026-10-12
- 03Slattery et al., Stability Comparison of Recordable Optical Discs: A Study of Error Rates in Harsh Conditions, J. Res. NIST 109(5), 2004 · accessed 2026-10-12
- 04Canadian Conservation Institute Notes 19/1: Longevity of Recordable CDs, DVDs and Blu-rays (Iraci, revised 2019) · accessed 2026-10-12