Tested: DVD±RW Endurance Round 2 – Part 2: Results & Failure Modes

In the previous post, I covered the contenders that would be taking part in this second round of DVD rewriting endurance tests. In this video-heavy post, I’ll be covering the results of a total of 4285 cycles of rewriting in total, although not all are successful. After all, I left the drives to keep rewriting and only checked for the first verify failure after-the-fact – most of the time, terminating the tests either after reaching over 1000 cycles or in case of the drive throwing sustained write errors.

Results

The following sections will detail the results and examine any visually apparent failure modes. A summary of the results follow.

Fujifilm 2x DVD-RW [MCC 01RW11n9]

This disc writes stably run-after-run at its rated 2x.

The first verify error occurs at run #66.

That coincides with the first dip in the TRT. After that, there are a few good runs and then another error occurs. There are long stretches of good writes between errors.

The quality scan suggests the disc was not in good health overall – the PIFs were high throughout even initially, but the quality of the burns tends to swing wildly the more rewrites that occur.

A look under the microscope seems to show there are regions of the disc, albeit infrequently compared to some of the other failure samples, where there is a darkening of the recording layer.

These spots may indicate the alloy in the recording layer is not stable in its composition and that might be the cause of the “carpet” of PIFs. But aside from this observation, the recording layer still looks visually normal.

Imation 2x DVD-RW [JVC_VictorW7]

The disc was written at the rated 2x speed.

While this is a Japanese-made disc, it’s not performing well. Only the first write seems to be taken well. The quality degrades rapidly, throwing a verify error by the third write.

The inner-diameter seems to be the area that suffers most.

Only the first write would be considered decent. Unfortunately, after the 6th run, the drive simply would not write to the disc anymore claiming of an Illegal Medium Format.

Interestingly, this disc exhibited a failure that I’ve heard rumoured about but never experienced – the claim that the embossed section of the DVD-RW disc would somehow fail after rewriting. We can see that there is some sort of dark areas developing around the embossed data.

But to blame the DVD-RW format or the embossing for the failure would be premature. Instead, this disc is suffering some form of disc rot despite the recording layer being inorganic.

This is apparent, as even away from the embossed section, there are black spots scattered about. This is a subtle form of rot, but the precipitating cause is not known. The reflective layer does not show pinprick holes on examination with light, suggesting there may be some sort of chemical issue with the recording alloy and reflective layer interacting.

Imation 4x DVD+RW [MBIPG101-W04-000]

This disc wrote consistently at 4x.

No verification errors in 1005 cycles, the test was terminated as the disc achieved the 1000 cycle lifetime expectation.

The read speed graph was consistent throughout.

The quality scan did show some yo-yoing of the write quality especially as the disc ages, with the outer diameter seemingly suffering some slow failure that increased the error rate in that area. I feel that my previously hypothesised ageing of the PCA impacting on burn quality may be supported by such observations, as the PCA is used up in a “random” order and is what varies on a burn-to-burn basis. As a result, the laser power will vary write-by-write especially if there are defects in the media’s PCA.

Imation 4x DVD+RW [RITEK-004-048]

This 4x disc wrote consistently at 4x.

The first verification error occurred at cycle 259.

The use of TRT does not detect errors until the burn is noticeably bad.

The errors seem to be at an intermediate diameter and at the outer diameter, rather severe oscillations in burn-to-burn quality was exhibited.

Memorex 2x DVD-RW [CMCW02]

This 2x speed disc burned at 2x.

At cycle #88, the first verification failure occurred.

Once again, the TRT was not very sensitive to disc failure until very severe errors occurred.

The disc seemed to develop spiky errors which increased in density as the disc aged. This suggests some sort of material degradation, but this was not apparent on a visual inspection of the disc.

Memorex 4x DVD+RW [INFODISC-A10-001]

I did not have high hopes for this DVD+RW, but it seemed to work okay at 4x.

Surprisingly, it survived 841 cycles, which is not quite 1000 but a lot more than I expected.

There are only minor jitters in the middle of the TRT prior to the disc’s first verify failure.

For some reason, this disc had a “bad patch” in the middle of the disc, even from fresh, with a carpet level of PIFs which would not have me consider the burn as being good. But the rate of degradation was low, which is ultimately what allowed the disc to survive.

Omega 8x DVD+RW [PHILIPS-RW8-000]

This disc is a high-speed disc rated at 8x, but was already used. The drive consistently wrote with a Z-CLV 8x strategy.

A total of 398 cycles were completed prior to the first verification error.

While a TRT is not highly sensitive, it does pick up on the worsening conditions towards the end of the 6x zone at about cycle 140 onward.

The quality scan result shows the burner likely was not working well with the media, with very high errors and an obvious Z-CLV jump in error and jitter. Despite the horrible looking quality scans, the disc was still smoothly readable.

Philips 4x DVD-RW [INFOMEDIAA20]

There seems to be something unusual about this media that means that the 4x disc is detected as a 6x disc as the firmware has an entry for this MID at 6x. It seemed unwise to try writing a rewritable disc at a higher speed, based on my overspeeding experiments, so I requested a 4x burn. Instead, the drive ignored the request and ploughed on through at 6x.

The first cycle resulted in a read error, so basically it was not possible to write this disc properly.

The TRT shows the outer edge to be a major issue. This is not unexpected as this disc appeared sun damaged from the start.

The quality scan implies this disc is just impossibly-badly burned.

The final write failed part-way during the write phase. It seems the disc’s formatting is permanently damaged and the disc was murdered, but less severely compared to the Prodisc DW06 (see later in this post). I wonder if the iHAS120-6 has a firmware bug or perhaps some vulnerability where incorrectly read disc metadata could result in incorrect burning of media.

Under the microscope, there are fractal/Lichtenberg figures from the outer edge where the sun damage appears to have occurred.

This suggests to me that the disc adhesive or lacquer may have failed in some way, resulting in moisture migrating into the recording layer, causing it to fail.

Pioneer 2x DVD-RW [PVCW00V00245]

The disc burned at the rated 2x and the Mediatek chipset in this drive didn’t murder the disc right away.

The disc survived 346 cycles to the first verification failure.

The disc appears to have degraded throughout the experiment, with dips starting to appear at 110 cycles near about 400MB then later towards the outer diameters.

The initial burn quality is not bad although the peaks of the PIF are still higher than I’d like. The PIE/PIF rate increases steadily as the number of rewrites increases. The “yo-yoing” really only begins at about 110 cycles.

The failure is a form of disc rot as well, resulting in visible “clearings” in spots across the disc.

Backlighting the disc suggests that this failure is precipitated by pin-prick holes in the reflective layer. Perhaps the rewritable phase change layer interacting directly with the lacquer or adhesive results in outgassing during a write (as that causes high localised heating) and that’s “blowing” up a small pocket every time the laser crosses over.

Under the microscope, those pin-prick holes create darkened spots in the recording layer. Once the hole is large enough, it seems to “inflate” an oval around the hole. Some of these defects are in the PCA, which would likely affect the consistency of recording.

This sort of defect is scattered across the whole disc, so it’s surprising the disc even survived so long. Again, this is somewhat disappointing because this is a Japanese-made disc, but it didn’t survive very well.

At maximum zoom on my digital microscope, we can see a little more detail on the pin-prick hole. It appears black – perhaps burned by the laser. The concentric rings are likely an interference effect caused by the pitch of the spiral tracks on the disc.

Smartbuy 4x DVD-RW [Prodisc DW06]

This 4x disc wrote at the rated 4x speed.

The disc was a bit disappointing, managing 110 cycles before the first verify failure.

TRT disruptions happen around the 108 cycle mark, with increasing errors towards the outer diameter.

The disc burned very well at first, then started to get really spiky towards the end. I wonder if the spikiness is related to the disc’s construction in any way.

On the final burn that failed, the drive appears to have murdered the disc, causing the disc to have an unrecognised format. Visually examining the disc, the inner tracks appear to be written in such a way to be unusually bright compared to the rest of the disc. It seems the wrong power has been used and may have damaged the recording layer permanently.

A closer look at the recording layer shows there are a few very tiny dark spots.

These spots are distributed across the media but are not uniformly dense.

Interestingly, looking at the CPRM code area, it seems there is some sort of “bloom” around the written barcode. Compare this with the disc below and it is somewhat unique. It’s not likely related to the cause of failure and could be related to viewing it through a microscope, but is still an interesting observation.

Smedia 2x DVD-RW [OPTODISCW002]

The disc wrote at the rated 2x.

The disc survived 86 cycles before the first verification failure.

The TRT is stable until the very end.

The first 56 cycles seemed to show good burn quality, but then the outer diameter becomes spiky out of nowhere.

The CPRM barcode on this disc appears a lot more “regular” to me, without the blooming around the code. Compare this with the Prodisc DW06 example above.

SWTechnology 2.4x DVD+RW [OPTODISC-OP1-001]

This disc was a big problem. It’s so problematic it didn’t even deserve any images – it burned at 2.4x just fine, but then, the disc was completely unrecognisable by the drive afterwards. The first burn killed the disc and there was nothing visual about the disc that would explain why. I suspect it’s just that the disc is either of poor quality or the strategy in the drive’s firmware is completely inappropriate.

TDK 2.4x DVD+RW [RICOHJPN-W01-001]

The predecessor to my former favourite RICOHJPN-W11, this disc runs at 2.4x.

Unfortunately, even the first burn did not pass verification.

The TRT is all over the place …

… and the quality scan implies the disc is horrible.

From a wider view, it doesn’t look like there’s anything wrong with the disc. The slightly uneven herringbone pattern is very common, even amongst good discs.

Under the microscope, microscopic dark spots can be seen although there are no visible pin-prick holes through the reflective layer. It seems this is another “rotted” disc – even brand-new discs stored well may not be usable.

TDK 4x DVD+RW [PHILIPS-041-000]

The disc burned consistently at 4x.

This disc did fairly well, reaching 551 cycles before verification failed.

The TRT is clean, with only slight dips on the outer diameter towards the end.

Burn quality is excellent, until about 190 cycles when the outer diameter starts to show some signs of distress. The outer diameter error mound seems to oscillate on and off, but at about 490 cycles, it starts to grow almost monotonically and the errored area starts to increase. There seems to be no visually discernible change to the recording layer though.

Summary

The discs of this round are highlighted in light-yellow. Of all the discs, the Imation 4x DVD-RW from Moser Baer India is the stand-out, being the second disc across both rounds to meet and exceed 1000 cycles. The Memorex 4x DVD+RW from Infodisc was also a surprise, managing 841 cycles which is pretty good. I wouldn’t have expected this from a smaller manufacturer. The Imation 4x DVD+RW from Ritek was pretty average, being fairly similar to the other Ritek discs. The Pioneer 2x DVD-RW seemed to hang in there with 346 cycles, as did the used Omega 8x DVD+RW from Philips at 398 cycles. The Smartbuy-branded Prodisc DW06 was a bit poor at 110 cycles.

Unfortunately, there were plenty of “sick” discs – the SWTech 2.4x DVD+RW (OPTODISC-OP1-001), the TDK 2.4x DVD+RW (RICOHJPN-W01-001), the Philips 4x DVD-RW (INFOMEDIAA20) and the Imation 2x DVD-RW (JVC_VictorW7) which either didn’t manage a single verified burn or only a handful of burns in the latter case. The Fujifilm 2x DVD-RW (MCC 01RW11n9), Memorex 2x DVD-RW (CMCW02) and S-media 2x DVD-RW (OPTODISCW002) all didn’t make 100 cycles and may also suggest some degradation of the recording layer.

To compare between rounds, the TDK 4x DVD+RW (PHILIPS-041-000) of this run managed 551 cycles while the previous one only managed 218 cycles. This is quite the gap, although both discs come from different batches and the tests were done on separate drives. This suggests that there might be a significant difference on individual discs and drives – but a disc that does <100 cycles doesn’t seem likely to do 1000 cycles on another drive.

Conclusion

Unfortunately, many of the discs in this second round did not fare all that well, in part because it seems they have not aged well in storage. Subtle forms of disc rot were spotted that resulted in dark spots in the recording layers which increased error rates, to those which created larger cleared regions which seem to coincide with pin-prick holes in the metallisation. A failure of a DVD-RW with visible darkening of areas around the embossed region was observed, although the failure was distributed across the surface and not exclusively at the embossed area. A good number of discs were unable to complete a single successful verification, and another managed a single-digit number.

Of the tested discs, the Imation 4x DVD+RW from Moser Baer India (MBIPG101-W04-000) managed to make it to a full 1005 cycles before the test was aborted with no failures. Honourable mention goes out to the Memorex 4x DVD+RW from Infodisc (INFODISC-A10-001) which managed 841 cycles – not quite 1000 but a lot more than I’d expect from a lesser-known manufacturer. The TDK 4x DVD+RW from Philips (PHILIPS-041-000) managed 551 cycles, which was more than the 218 cycles of the same type but from a different batch in the last round. This suggests there is some potentially significant variation from disc-to-disc and drive-to-drive.

It would seem that having an inorganic recording layer (as phase change layers are metal-oxide-based) is not, in itself, sufficient to ensure long disc lifetimes. A breach of the adhesive bonding layer may allow atmospheric attack of the protective lacquer, reflective layer and recording layer. Similarly, construction defects in the form of pin-prick holes in the reflective layer may enable contamination of the recording layer by the lacquer or adhesive layer especially when super-heated by the laser. It cannot be ruled out that some recording layers may not be stable over time, so the recording characteristics may drift without visually observable defects. As a result, brand-new discs used soon after manufacturing may have exhibited better cycle characteristics than these tests have demonstrated.

This exploration answers a lot of questions I had about the rewritability of DVDs and gave me the opportunity to test even more media. But this second round of testing isn’t over … there’s a third part which is somewhat tangentially related.