Recordable Blu-Ray Disc Longevity – Periodic Survey

A while back, I posted a thread on CDFreaks (now known as myCE) about my Ritek BD-R’s which had completely lost data. That thread received in excess of 20,000 views and drew along a few skeptics – those who had used Ritek but suffered no failures. That being said, one of them (Kerry56) has reported failures just 2 years after the discs were originally checked as okay. I go into my own scan results more at my legacy page – this was before I owned anything capable of LDC/BIS scanning.

The recent revival of the thread prompted deanwitty to post:

“Who can ya trust?!!”

I consider trying to answer that question as our sacred duty at this forum. I think its great that we are seeing more and more discussions about BD-R longevity as the technology and discs are increasingly reaching the age of real world longevity testing.

So I decided, maybe it’s time for me to grab out a random sample of my BD-R’s in my 200-disc collection to give them a test on my LiteOn iHBS212 to see if they were indeed still readable, or in fact, there were more failures than I expected. All discs were kept in a light-tight box, under my bed, mostly temperature controlled below 30 degrees C (mostly around 20-24 degrees C). The results are surprising. The results will be grouped by Media Code, dates of burns provided where I can ascertain them (rarely). Most-to-all of the burns would have been done by my LG GGW-H20L Supermultiblue 6x Bluray burner – not the best burner if other quality scans are to be believed, but not always bad either. Unfortunately, as we don’t have “at burn time” scans, we don’t have a true “degradation” reference, we merely have a “does this disc look okay” reference.

The rule of thumb for LDC/BIS scans is LDC average <=13, and BIS peaks <=8. A LDC scan and a TRT scan takes about 32 minutes for a single layer disc – so this post took a really long time to make … All scans were made December 2012.

TDKBLD-RBB-000

These are the “real” TDK BD-R’s (before they started re-shipping CMC stuff). Almost 4 years old …

Sample Disc 1 [Burnt *around* early 2009]

Signatures: GGW-H20L YL05 (released Nov 2008), Imgburn v2.4.2.0 (released Jul 2008)

That’s not the worst scan I’ve seen but it’s pretty bad. Don’t tell me I have to migrate my data off the 40-or-so TDKBLD-RBB-000 discs of mine …

Transfer-Rate test (TRT) looks perfect though, but the warning signs that the disc isn’t that healthy are on the wall. So why don’t we grab another sample of this MID code.

Sample Disc 2 [Burnt *around* early 2009]

This disc has the same fingerprint as the other disc – in fact, it was the next sequential disc in the set.

So the LDC test gives quite a few spikes, but isn’t initially too bad considering the disc could be slightly dusty (these can contribute thin spikes on the test). Looks somewhat different to the first disc – maybe the first disc was a particularly bad burn (the bottom has regions of different darkness, suggesting the burner stepped laser power during the burn, whereas this one was much smoother). Regardless, there are still regions of BIS out of spec, and LDC is high … especially towards the outer edge. Not what I’d want to see.

TRT is perfect as expected. I’m not sure we’re gleaning too much info from the TRT itself.

Sample Disc 3 [Burnt *around* early 2009]

This disc is a completely non-related disc, and wasn’t burnt in the same set, but with the same fingerprint. This can be explained by the fact that I buy my media in batches and tend to burn through a spindle of one MID before moving to the next.

Don’t ask me why, but this LDC scan looks better than the above two. Same drive, same firmware, same test drive, same media code, same batch, roughly the same age, same storage condition. But still, not good towards the outer edge.

TRT perfect as usual. So it looks like maybe the TDK discs and the LG burner weren’t a good combination to begin with, but the graphs suggests that these discs might not have too much time left up their sleeves.

TDKBLD-RBA-000

These discs are the 2x versions of the ones above. I only ever owned two of them, one of them succumbed to oxidation of the reflective layer, and the other one seems to be okay. Burnt around the same time – early 2009.

Sample Disc 1 [early 2009]

So we are getting a bit of “error growth” in a band towards the outside, with some spikes on the inside. Maybe the LGs just aren’t getting along with the TDKs.

TRT fine as usual.

Sample Disc 2 [early 2009]

This is why there isn’t a scan of this one, by showing you a *different* sort of scan of the disc:

Here’s a closer look if you can’t tell what’s happening:

That’s the reflective layer being eaten away by moisture. Looks like the adhesive sealing at the edge of this disc is damaged or improperly formed, so moisture got in. Data that is lost this way is lost for good, but at least you can physically see this happening (unlike the data loss with the Riteks). LG discs are showing some signs of this too – but the data that’s on the discs looks good.

VERBAT-IMc-000

This particular disc was bought as a no-brand reprint top, so it’s not likely to be the best quality. Almost 3 years …

Sample Disc 1 [Burnt January 2010]

Signature: GGW-H20L YL05, Imgburn v2.5.0.0 (July 2009)

The scan itself doesn’t perfectly meet our stringent requirements with spikes above the required levels but is pretty close and looks quite good. Looks like Verbatim media proves itself, again, to be superior.

No trouble from the TRT as one would expect.

Sample Disc 2 [Burnt January 2010]

Same signature as above, for the same reasons described earlier.

A few spikes but generally not too bad. In fact, with results like these, I see no need to test a third sample disc.

TRT is perfect, as expected.

CMCMAG-BA3-000

These disks were from a batch of TDKs which (dissapointingly) ended up being CMC’s. Which might not be such a bad thing when the TDK performance (above) is considered. Almost 2.5 years old.

Sample Disc 1 [Burnt Early-Mid 2010]

Signature: GGW-H20L YL05 (Nov, 2008), Imgburn v2.5.0.0 (July, 2009)

Interestingly, the error rates seems to hold a consistent error floor, with a few spikes here or there, until the outer edge where it raises somewhat. It’s not perfect but at least the error rate seems to be quite stable. This might mean the disc (when it fails) will fail as a whole … The error rates for LDC are a bit high, for BIS seems to be reaching its limits.

TRT is all good. I guess testing with TRT has limited value as I’m yet to see anything but a perfect TRT on my samples.

Sample Disc 2 [Burnt Early-Mid 2010]

This scan starts off better, the same error floor seems to show. Occasional spikes can be seen as expected, this disc initially keeps within the error guidelines, but towards the end, things seem to start falling apart again …

And TRT is perfect, as expected.

Sample Disc 3 [Burnt Early-Mid 2010]

So lets take a third disc. And look at that, it’s a consistent story!

And TRT is perfect too.

PHILIP-R04-000

These discs came from a spindle labelled Optical Quantum. As far as I know, they may be made by Moser Baer India, and have a mixed reputation. One and a half years old.

Sample Disc 1 [Burnt July 2011]

Signature: GGW-H20L YL05 (Nov, 2008), Imgburn v2.5.0.0 (July, 2009)

The scan looks okay, but then again, the discs themselves are quite young. Error rates increase towards the outer edge, and the errors start hitting the limits at the outer edge.

TRT looks just fine.

Sample Disc 2 [Burnt July 2011]

The scans look very similar, so I think it’s a fair representation of the disc state. I don’t have many of these discs, so two discs will do.

TRT looks just fine as well.

LGBRA-S04-000

These were individual jewel cased discs, I only bought four of them, so testing two is more than enough. Almost 2.5 years old – the discs are showing some oxidation from the edge of the reflective layer. Some of my TDKs have been lost that way. These don’t have a good reputation around myCE as they can come from several sources. This was a non-printable disc, 4x rated.

Sample Disc 1 [Burnt Early-Mid 2010]

Signature: GGW-H20L YL05 (Nov, 2008), Imgburn v2.5.0.0 (July, 2009)

The scan itself looks exceptionally good for BIS compared to the others, but the LDC has lots of spikes. The average is still not too bad.

The TRT is perfect, as per usual.

Sample Disc 2 [Burnt Early-Mid 2010]

Degrades slightly at the end, but is otherwise quite consistent with the other LG disc. Strangely, the LG discs seems to be the best of the pack so far!

TRT needs no comment.

INFOME-R30-000

These were unbranded from dexwebweb on eBay – the lowest cost blanks which were available a while back, decided to get 50 of them to give them a go, knowing that some people had problems with these discs on the outer edge. This is the youngest set of media I have and I only have samples burnt with the iHBS212 5L09. These discs are only 9 months old.

Sample Disc 1 [Burnt April 2012 with the LiteOn iHBS212]

Scan of the first disc shows a fairly good burn actually. But then again, it is the youngest batch of media I’ve come to own.

TRT is just fine as usual.

Sample Disc 2 [Burnt April 2012 with the LiteOn iHBS212]

Scan of the second disc shows a consistent behaviour – I didn’t expect it to be like this given the “bad” reputation this media had. Then again, my Riteks did seem okay when they were still alive …

TRT is perfectly fine as usual.

MEI-T01-000 (DUAL LAYER)

These would have been my earliest dual-layer discs, burnt mid-2009. Unfortunately, due to (what seems to be) a bug with the iHBS212 – error scans freeze the drive (hard freeze) requiring a reboot right at the layer change, so only TRTs for these discs. The discs were printable top 1-2x Sony branded discs.

Sample Disc 1 [mid 2009]

TRT shows a big dip right at the layer change, depending on the drive, this can be expected as it represents the time needed to refocus from one layer to the next.

Sample Disc 2 [mid 2009]

Uh oh. TRT reported an error at the layer change, probably lost some data here. I’ve always been a bit wary of using DL for data archival because increasing errors at the layer change seem to be fairly common.

[UPDATE: The following results using the LG GGW-H20L appear to be affected by an unknown glitch, which makes their results invalid. The disc appears to still likely be functional. More details in my addendum.]

Why don’t we shove it back into the LG GGW-H20L and look at what CD-DVD Speed Scandisk looks like:

Okay, so that’s interesting. The LG thinks this disc is fine with a few yellow blocks (quite common on my LG scans, seems to be nowhere near the layer change) … so maybe lets look at the TRT from the LG and see what that looks like:

And no kidding, LG has a (repeatable) issue in reading this disc, but somewhere where the LiteOn had no trouble at all. This disc is peculiar. Why could it pass a scandisk with a few damaged but not unreadable sectors but not a TRT?!

Sample Disc 3 [burnt mid 2009]

So this time, we get a tiny dip at layer change, this suggests the big dips above are unusual. Surprise surprise, MEI discs not doing so well?

TDKBLD-RFB-000 (DUAL LAYER)

These were bought off eBay from j-stop, and came in sleeves, three discs only. Only about one year old.

Sample Disc 1 [Burnt August, 2011 by LiteOn iHBS212 5L09, Imgburn v2.5.5.0]

So the TRT isn’t perfect with a small dip in the second layer. Lets see what happens with a second disc.

Sample Disc 2 [Burnt August, 2011 by LG GGW-H20L YL05]

This second disc was burnt by my LG instead, and seems to be okay as well. Not perfect though, but close.

Conclusion

None of the presented burn graphs are quite perfect in terms of meeting the LDC and BIS guidelines. Some are worse than others, but (almost) all discs showed perfect readback and (mostly) perfect TRT. Would I trust BluRays for archival? Uh, well I’ve already committed, but I’m not sure it was a wise choice. The longevity of these discs could be under question, so it seems – the margins for errors with such small pits and lands is really not much. Dust on the disc during a burn, or a test can throw these error charts towards a spiky mess. Okay, sure, two or three discs a brand is hardly representative, nor is having most of the discs burnt by an “early” model of BluRay burner.

Feel free to discuss and comment …

[Update: Some discoveries were made in regards to dual-layer disc scanning, with some results now available – please see addendum which will be updated as results arrive.]

About lui_gough

I'm a bit of a nut for electronics, computing, photography, radio, satellite and other technical hobbies. Click for more about me!
This entry was posted in Computing and tagged , , . Bookmark the permalink.

9 Responses to Recordable Blu-Ray Disc Longevity – Periodic Survey

  1. M.Elea says:

    Why did you do the quality scans with 8x speed and not with 4x?

    • Gough Lui says:

      The science of Bluray scanning is still not as developed and standardized in the hobbyist community as the DVD era. 4x was a standard back for DVDs because it minimised the scanning time, while minimising spurious error readings towards the outer edges due to drive focus and tracking shortcomings.

      As for the standards, they may have been defined for 1x reading only as they probably do not govern how higher speed drives operate – merely that discs within certain error boundaries should be readable. It is already known that “amateur” scanning using consumer drives yield results which deviate significantly from that of standards testing equipment and only is a gauge of how good a disc reads in that particular drive.

      As for Bluray scanning – I have seen 4x suggested as a “good speed”, but others have chosen to use 6x and 8x. It can be expected that discs will see slightly elevated error rates at higher scan speeds (again, due to drive influences), however, one reason I chose 8x as the scan speed is due to the sheer number of scans required. As it was, it was already a very long undertaking. The second reason is that I have scanned discs at 8x for as long as I have cared to scan them – pressed discs show reasonably good error rates scanned at 8x, so why not recordable media?

      There was also a myCE thread earlier that alluded to 8x scanning being similar to 4x for results but saving a mountain of time – but this was when I began owning my LiteOn, and since then, developments may have changed the philosophy of things – I will rescan as much as I can justify spending time on at 4x for a better view.

      Thanks for your suggestion!

      • M.Elea says:

        In my opinion 8x scans could be fine as long the results are within the expected limits. But as you can see – most problems are in the outer areas. A reason could be that the disc was burnt too fast. But scanning them with a lower speed could could give better results.
        And as long you do not have an original scan (after burning) you can hardly say whether they aged or not.
        But I like most of your results.
        Btw: when I scan with 1x or 2x speed I get sometimes worse results as with 4x.

        • Gough Lui says:

          All discs were burnt at rated speed as most were burnt on the LG GGW-H20L which isn’t really capable of overspeeding most media.

          Stay tuned, I’m about 50% of the way through scanning all media at 4x. The results show reduced errors, but trends seem “broadly” similar – especially peak BIS. LDC values are about halved.

          I’ve also managed to scan the only 2 readable Riteks from my 16 discs to show what “still readable” Riteks look like – the surprises will be unveiled sometime in the next 2 days.

  2. Pingback: Recordable Blu-Ray Disc Longevity – Periodic Survey Part 2 | Gough's Tech Zone

  3. Pingback: Addendum: LG GGW-H20L DL Performance Oddity | Gough's Tech Zone

  4. Pingback: A More Detailed Study of Blu-Ray Scanning | Gough's Tech Zone

  5. cd_pirate says:

    8x scanning is fine. The best burns still look good at 8x and marginal burns will looks worse at 8x but better when lowering speeds. The worst burns looks bad no matter the scanning speed.

    8x is realistic as when you rip the data back to hard drive, you aren’t going to do it at 1x or 4x. The drive will try max speed, which is 8x most of the time.

Error: Comment is Missing!