Tested: Quality Variation in Blank DVDs & “Bad Batches”

While doing some prior research on optical media, I came across the Digital FAQ’s post which discussed the concept of “bad batches” as a cop-out for when someone’s DVD quality scans don’t match the expected high-quality exhibited by others using the same combination, although peppered with a grain of truth where production issues and sub-par media arrived to the market. While I do agree with some of the premise, I think that the truth is a bit more nuanced than this, so I decided to put my equipment to good use and help shine some light on the issue. After all, we’ve already seen just how a “batch” of drives can differ quite a bit on recording quality using discs from the same spindle, in this post, I’ll be looking at the media side of the equation.

Production Truths

If you have a spindle of discs, how do you imagine they came about? Perhaps you might imagine a single machine, making discs one by one, each of which go into the spindle sequentially. So, let’s say you buy two spindles, presuming the discs are serially numbered, you might imagine that one spindle has discs 101-150 and the other has 251-300.

In reality, from what my understanding of how discs are made, things are a lot more complicated. Large companies have many lines operating in parallel and they are not all independent. It’s likely that a number of lines might feed into a given “spindle” due to the way the factory is set-up, so you might already be receiving a mix of discs from different machines. Each of these machines will be at a different “cycle” in their production lifetime as they likely do not receive maintenance all at the same time and consumables, like stampers, may not wear evenly or may only be replaced when absolutely necessary. Quality control is based on sampling – not every disc can be tested thoroughly, so perhaps only quick automated visual checks can happen, but every so often, someone will pluck out a disc and shove it into an analyser to see if the plant is still operating within its quality control limits. Based on this model of how a manufacturer might operate, a spindle could contain a mixture of individuals from different lines, only some of which have been quality checked. After all, those who work in electronics well know that no two components are the same. Perhaps the only exception to this may be small-scale manufacturers who only have one production line.

Beyond this realisation is also the understanding that certain parts of the system are consumable and will degrade. The most notable one is the stamper which forms the land/grooves and lead-in pre-pits (DVD-R) or ATIP wobbles (DVD+R) that are used by a drive to identify its position and place its marks accurately. Stamper “sons” are built through a multi-stage process from a glass master, to a “father” negative, to a “mother” positive, then to a series of “son” negatives that are used in the production machines. This process ensures that a single glass master can be faithfully replicated to a large enough number of “sons” that do the work. Each “son” reportedly has a lifetime of around 30,000 to 50,000 discs, meaning that the discs produced from a range of “sons” may look the same (in terms of stamping code) depending on whether the manufacturer does anything to differentiate them (e.g. laser marking) but are actually from different machines. I suspect that’s where some of the variations in ink-dot printing or centre hub code pressings are indicating. As each pressing is made, the stamper will likely slowly deform and the sharpness of its edges will slowly round-off, resulting in a less-defined positioning signal. The stamper itself may warp slightly and positioning information can be corrupted slightly by accumulated debris. While drives can compensate for this to some extent – the accuracy of the servoing information provided by the stamper needs to be highly accurate to avoid drives throwing write errors or falling back in write speed. Add to this, the fact that the production machines are mechanical and they wear over time, so there may be some differences in positioning tolerances as well – this could result in the finally produced discs having dimensional variations leading to tracking run-out, disc vibrations or layer unevenness.

As a result, it’s probably better to think of each line as “somewhere” in-between brand-new to “needing maintenance” and your spindle could have a mix of discs from several lines operating at different points in their lifecycle. An added complication is the concept of binning or grading – manufacturers may choose to have “better” discs separated for their premium brands or offerings (e.g. medical, first-party brands) and offer the “acceptable” discs to other OEMs who will rebadge them, often offering such discs at lower prices. This can result in instances where the same media code, but across different brands, can appear to have quite different characteristics.

Because of this, I’m more inclined to think of discs as individuals, but if a factory is operated well and under tight quality control, the differences between individuals should be small, resulting in consistency (within limits). My personal hypothesis regarding the appearance of the “bad batch” concept relates to the rapid growth of DVD writing popularity – the high demand may have incentivised manufacturers to be more lax with their quality control and it was claimed that some low-grade discs had made their way to the market and been rebadged as a higher-grade product. That may have been the start, but as margins further decreased due to competition in the optical media market, the incentive to maintain quality (which comes at a cost) may have declined along with the exit of prominent Japanese manufacturers. This was potentially somewhat offset by later writers becoming more advanced and capable of handling more marginal products. As a result, I suspect that my observation of some later batches of media seemingly being “difficult” may relate to an overall decline in the industry. But this is just my hypothesis.

How to Tell if a Disc is Good or Not?

This logically leads us to the question of how a blank disc might be judged. For an end user, burning a disc and running a quality test may be considered the ultimate litmus test, but the result is the product of many factors. A good burn may be because a given drive has better compatibility with the disc in the form of a tailored write strategy, might simply be better at handling whatever defects the disc may have (but other drives may not), or you managed to pluck a “golden disc” that was somehow better than the others, whereas a consistently good media should usually burn well across many drives (assuming it was supported). In many cases, for the best result, to satisfy my curiosity and to depict every media in its best light, I take the time to test across a number of drives to determine the optimum combination(s) to use. While I can’t say this is entirely invalid, as a test, it’s not very specific as to the cause of issues when things do go wrong.

In industry, it would be more common to use disc and stamper analyser systems to determine the quality of unrecorded discs and stampers to monitor the production line on a sampling basis. Such devices are relatively rare, expensive and require calibration. Unfortunately for me, I don’t have access to such calibrated quality analysers, but the next best thing (in my opinion) is to perform a tracking error and focus error (TE/FE) test.

Because this is done at a read-level of laser power and it is done by sampling sections across the disc, it is relatively fast and non-destructive. It can even be run on discs which have been recorded (and the recording itself should not interfere much with the result). The data is a dimensionless quantity, which usually indicates the amount of deviation that the servo systems have to deal with for tracking (inside-outside) and focus (up-down) directions. While these parameters are not the only parameters which quantify the quality of an unrecorded disc, they are the only relatively-easily accessible ones – that said, only a certain range of drives that are capable of it – BenQ’s DW1620, DW1640 and DW1650 via Qsuite and selected Lite-On MediaTek-based drives via DVDScan 004 beta and certain Plextor drives with PlexTools are the only ones I am aware of, however, the results will vary across drives. In my experience, I found the DW1640 to be the most sensitive and consistent, so I tend to use it most often.

But perhaps more importantly, while the TE/FE test will provide us with information if a disc is good for recording at a particular speed based on whether the servo can follow the pre-formed servoing information well, the more interesting thing to note will be the variation between discs of the same type – as with all physical systems, we do expect some variation from disc-to-disc in the same spindle and also between the same disc media ID across different spindles. Whether a spindle has a particular characteristic pattern that differentiates them from another spindle, or whether they are indistinguishable from one another, will be a key indication of the concept of “batches” and quality control.

Methodology

For this test, a single BenQ DW1640 DVD writer was used with its Qsuite software. When I started, I performed tests by hand but quickly, I wrote a python pyautogui screen-based automation to start the TE/FE scan every time a valid disc is inserted, take a screenshot of the completed result and eject the tray to minimise test time. Candidate discs included all blank DVD media that I have on-hand in spindles, preferably with 20 samples or more (except DVD+R DL where 10 samples was enough). Many spindles have been partially used, resulting in odd numbers of discs (e.g. not 20, 25, 30, 50, 100). No retail jewel-cased discs were tested due to the labour involved in opening each of the cases, but also as this operation risks introducing dust onto the media surface impacting on recording quality.

Repeatability

While the TE/FE test is something I’ve anecdotally experienced as being valid and correlating well with speed fall-backs during writing, its validity deserves further checking. While I cannot determine its absolute accuracy, I can at least determine its repeatability by scanning a single Verbatim-branded CMC MAG.AM3 16x DVD-R disc ten times sequentially and comparing the results. This will provide a baseline that illustrates what to expect from a “zero-variance” case and should indicate the “noise” of the measurement system.

Speed

One potential issue lies with the scanning speed. By default, Qsuite allows you to scan up to the supported write speed for the media. However, scanning above the rated speed is possible by enabling the Overspeed feature (but not for DVD+R DL). Scans produced at slower speeds take longer and typically are flatter (less sensitive), so a higher scanning speed is generally preferred (assuming the results are not off-scale most of the time), however, it is noted that lower-speed rated discs may appear poorer in TE/FE quality as the TE/FE margins for recording are generally tighter than those needed for successful 16x readback.

To investigate the impact of speed, a single 8x media type (Imation Forcefield CMC MAG.AE1, 30 discs) was scanned at both 8x and 16x for comparison.

Variability

A total of 1200 discs were scanned, belonging to 30 different spindles (some of which have not been featured on this blog yet), to determine variability across different spindles. Of the 30 spindles, a total of 15 media codes are represented, which include DVD-R (10 types), DVD+R (2 types) and DVD+R DL (3 types). Unfortunately, some media codes were only represented by one spindle, so only disc-to-disc variability can be examined, while some other media codes are represented by multiple spindles, allowing comparison of TE/FE trends across spindles which may differ in plant and production date.

To best present these results, the screenshots have been annotated using a separate python script using pillow and assembled into an animated GIF with Adobe Photoshop. The results are visually interpreted, owing to the lack of raw TE/FE data when collected this way. Note that some TE/FE curves seem to end early – this is because the drive times out when scanning very poor discs and this results in an early abort of the test.

Results

Following are a series of animated GIFs that present the results of scanning 1200 individual discs with some commentary. The process itself took around three days of work in total (which I was frantically doing before my flight out to Singapore).

Repeatability

When it comes to repeatability, it appears that this test is very repeatable. The lines have a very similar shape, bouncing up and down by a unit or two either way. Some variation is expected due to the mechanical nature of spinning a disc and movement of the optical pick-up, but the fact the values are so close suggests to me that the test is highly repeatable – therefore, significant multi-unit moves or changes in shape of the graph are likely to represent actual deviations in the disc itself rather than random noise.

Speed

The Imation Forcefield 8x DVD-R (CMC MAG.AE1) scanned at its rated speed appears above. While this section is not entirely concerned about variability, it is clear there is some variability especially with regards to TE with numerous failures even at 8x. There is minor variation in FE, mostly at the inner diameter where a few discs seem to have a bump and others do not. It seems there might be a few types of disc – one with an inner, mid-outer and outer TE bump, one with more of a mid-outer and outer TE bump. Regardless, this spindle itself is not internally consistent – a couple of discs also have repeatable spikes.

The same discs scanned at 16x show higher values due to the higher demands that the faster rotational speed places on the disc. The same bumps in FE are seen, and the trends with TE are exaggerated, especially towards the outer diameter. The fact some values are off-scale may make comparison a little more difficult, but given the substantial time savings (~20s/disc), it does suggest that higher-speed scanning can be appropriate.

Variability

To best present the variability section, I’ve grouped the results by media code, with the previously presented result excluded.

Variability – Type RITEKF1 (16x DVD-R)

A total of four spindles of this media code were scanned, the first three belonging to Hi-Disc, a Japanese brand belonging to Mag Lab which I suspects rebadges media of mixed grading.

The first spindle is mostly consistent, with a slight inner FE bump and a noticeable TE hill at the outer. There appear to be an occasional outlier – one disc with TE spikes in the middle, another that has a sharp FE excursion on the outer and a third with a more minor FE excursion on the outer. Perhaps these discs might have suffered some damage in transit. Otherwise, the discs in this spindle seem to follow a single characteristic pattern.

The second Hi-Disc spindle seems to have a mixed characteristic. Some samples have a pleasantly sloped TE/FE curve with few bumps, but there are discs with an inner TE bump and a bit more elevated TE towards the outer. Some discs in this spindle are a much better candidate for 16x burning than others, but it doesn’t look anything like the first spindle.

The third Hi-Disc spindle is yet another distinct entity. This one seems to have a majority of the spindle showing a double-TE bump in the inner diameter portion and a TE hill towards the outer diameter. Some discs have notably worse outer-diameter performance in both TE/FE while some also have elevated FE baselines. This suggests this spindle has at least two different characteristic patterns. Three spindles so far and three different behaviours, each showing outliers within the spindles but also distinct characteristics across spindles despite having the same branding and media code.

The fourth spindle comes from Maxell, which is an “in-house” brand of Ritek after they acquired the name. While one might expect the best discs to be reserved for their own names, this spindle seems to show higher FE with some discs exceeding limits and a noticeable double-bump on the FE in the mid-inner diameter and towards the mid-outer diameters. The TE is noticeably lumpy throughout but flatter and not as wild towards the outer except for one particular disc. Again, individual outliers are visible, but it seems the spindle does share a similar characteristic pattern.

Variability – Type TYG03 (16x DVD-R)

For this venerable type, one of the few consistently supported at 24x speeds, I have three spindles, one of which is Made in Japan and the other two, Made in Taiwan post-CMC acquisition.

The JVC-branded Japan-made spindle is surprisingly not as consistent as I might have imagined. The TE values are excellent for a disc scanned at 16x, remaining under BenQ’s extremely stringent limit lines. The FE however, does not always do so, dancing around especially towards the outer diameter, occasionally touching, with individual outliers exhibiting TE spikes or FE “hills” towards the outer. There are a few characteristics – some of them seem to have a double-FE bump, others look more like a linear slope.

However, compare this to the Verbatim “black” Japan-technology (but Made in Taiwan) spindle of discs that would be representative of CMC Pro era production and it seems everything is different. Not just marginally worse – but some individuals were so bad that the drive errored out while scanning and off-scale results were seen, both in TE on inner and outer diameters. The FE result is better across most of the disc but exceedances towards the outer diameter are visible. There seems to be a mid-inner diameter small bump in TE across most of the discs – I wonder if that’s an upstream stamper defect that got replicated across to many sons. The inner diameter jumps are rather shocking.

As I didn’t know it yet, I snapped up a second spindle thinking it was good quality stuff. This second spindle seems a little different – it has a slightly slopier TE with a mid-inner-diameter bump as well, but there are individuals with FE bumps in the mid-outer and mid-inner. Perhaps the quality control is not as good overall, but there is at least a correspondence to the fellow discs in the same spindle for the most part (excluding the occasional outlier).

Variability – Type TYG02 (8x DVD-R)

Another venerable type of media, the predecessor to the TYG03 above. This time, I had three spindles, all OEM duplicator grade stuff, one with a white top and two with a plain silver (unprinted) top.

Considering it is scanned at 16x, the result for TYG02 is surprisingly good overall, with some discs achieving levels even the Japanese-made TYG03 didn’t quite achieve. That being said, there seems to be some discs which have an FE tendency to go high from the mid-outer diameter, one outlier seems to have a mid-inner diameter FE bump, while the TE is mostly well controlled except for a single outlier. Overall, the spindle is mostly consistent but there are a few outliers that defy the trend.

The first silver-topped spindle also has excellent characteristics overall, but the FE behaviour seems to show individuals with noticeably worse outer-diameter performance. The TE values are mostly consistent, with only minor bumps in the mid-diameters on some discs, but nothing particularly noteworthy.

The second silver-topped spindle seems to be a little more consistent on those FE variances, although there are a few outliers too. But it would seem Taiyo Yuden, at least in this era, had quite good quality control, at least in their OEM duplicator line media. This matches well with my own experience, as it was the media most consistently recommended to professionals.

Variability – Type CMC MAG.AM3 (16x DVD-R)

This is a much more ordinary type of media, one that was formerly ubiquitous. Of this type, I have four spindles under different brands.

The first comes from the Hi-Disc brand of Japan. While most discs of this brand seem to follow a pattern that is stable except with elevation in both TE and FE towards the outer edges, there seems to be more than a few discs with a prominent mid-diameter FE bump. There is also some discs which have a prominent inner-diameter TE bump. As a result, it appears there might be three different characteristics in this one spindle, with the latter two categories being significant risks of failure for high speed recording.

The second comes from Verbatim, but sourced via Japan as well. The characteristics of this spindle are mostly much better – actually meeting the DW1640’s stringent standards for TE/FE at 16x. There are a couple of outlier discs with significant outer diameter exceedances, but note that these are not the top and bottom discs of the spindle but are somewhere mid-spindle (discs are numbered in order from top to bottom).

Another spindle also sourced from Japan, but under the Victor brand. This time, there are notable middle-inner diameter FE bumps on most discs, but the exact peak varies slightly across discs. One outlier has a significant mid-diameter peak in FE. A very different characteristic with the bump being a bit earlier than in the Hi-Disc examples. I suspect this is all indicative of stamper wear.

Finally, a locally (in Australia) obtained lot of discs under the Imation branding left-over from the heydays of DVD burning. This batch seems to have higher FE overall with some discs showing an inner diameter FE bump while most show an inner-mid diameter FE bump. Iin all cases, TE exceedances on the outer edge seem consistent. Not as much in the way of obvious outliers, this one seems rather consistent in quality – but consistently marginal.

Variability – Type CMC MAG-M01-000 (16x DVD+R)

Only one spindle of this type, also locally acquired in Australia, the +R counterpart to the above. Interestingly, it has a much better characteristic suggesting it is mostly meeting the DW1640’s stringent TE/FE requirements for 16x burning. There are no obvious large bumps, although outer edge exceedances do occasionally occur – maybe this is pressing-to-pressing variation. The TE/FE values are much lower overall. Part of this might be down to the drive and the superior tracking of ATIP wobbles in the +R format, but another could simply be the lower demand for +R media compared to -R resulting in less wear on the production stampers at the time of production, but that is just a guess.

Variability – Type MCC-004-000 (16x DVD+R)

Perhaps a more common quality DVD+R from the end of the era, I have two spindles of under the Verbatim brand.

The first spindle is mostly a case of a sloping FE and stable TE, getting close to and occasionally crossing the limit line. But there are a couple of outliers – mostly a case of an outer diameter big hump which in one case turned out to be a defective disc that had delaminated on the outer diameter (could be physical impact during shipping). The others had more minor excursions with a bit of undulation on FE in the inner diameter and elevated TE/FE on the outer, one seemingly also in the mid-outer diameters. Despite being Verbatim and locally sourced in Australia, there are still the occasional outlier discs.

This second locally-sourced batch seems to have a bit of a tendency to show mid-diameter FE bumps in quite a few of the discs, alongside occasional inner-diameter TE elevation on some of the discs. The elevation in TE/FE towards the outer edge is also there too. I’d say this is probably a slightly worse result than the one above, again, showing the point that the same media code can see differences in characteristics across spindles, despite both being locally sourced logo-top type discs.

Variability – Type MCC 03RG20 (16x DVD-R)

Now, we change over to the -R format version of this disc. While formerly ubiquitous, I’ve had many issues with the media type as of late (new-old-stock potentially poorly made or poorly stored). As a result, I only have two spindles to compare, both of which are Made in India under license by the now defunct Moser Baer.

The few logo-top discs I had seemed to have slight TE elevation inside with some elevated TE on the outside too but FE is stable. The number of samples is small, so it’s hard to pull out any patterns in terms of groupings.

A brand-new spindle of Lightscribe-capable discs show a very noisy and peaky TE on the first half of the disc, while the FE is mostly stable with a few showing slight increases near the inner diameter. The whole spindle seems internally consistent within itself (after all, LightScribe media is somewhat specialty, so perhaps it has a smaller production line to itself).

Variability – Type MBI 01RG30 (16x DVD-R)

Speaking of Moser Baer, I do have some of their own media under the Kodak brand in two spindles – one logo-top and the other printable-top.

The logo-top one seems to show a few discs with mid-diameter FE elevations, while others show outer-diameter TE elevations. Most discs have a stable FE. Some of these discs are quite stellar, meeting the DW1640’s stringent requirements but others fail, so this is a bit of a mixed batch although not catastrophically off-scale in terms of TE/FE values.

The printable-top version spindle seems significantly worse. The FE values exceed limits from about half-way in the disc, while the TE is higher too, crossing the threshold earlier. None of the discs meet the DW1640’s criteria, but some of them have noticeable inner-diameter TE/FE humps, suggesting there are a few characteristic groups in here. The media is not consistent.

Variability – Type FAKE MCC 02RG20 (8x DVD-R)

Moving onto smaller and lesser-known or lesser-reputable manufacturers, it’s the Imation i-Pro 8x DVD-R discs which carry a fake MID.

These discs were not rigid enough to get a decent scan at 16x, so the speed was reduced to 8x, honouring its original rating. Most discs look okay under this arrangement, but there is one disc with TE/FE issues in the mid-diameter and a couple with elevated outer diameter readings. The result is fairly consistent, but with outliers.

Variability – Type VANGUARD (4x DVD-R)

I looked at these discs before and have one spindle to hand, so I decided to try it. It’s not quite good enough to do higher speed scanning, so this one scanned at 4x (rated speed). Despite being quite lax in that sense, it is clear that there is a group of discs with a TE hump towards the outer. Some discs also have elevated TE relative to all the others. FE is generally good, but there is also the occasionally spiky disc. I suspect this is the result of multiple plants mixing together based on the ink dot printing I mentioned in earlier posts about this media.

Variability – Type GSC003 (8x DVD-R)

A relatively uncommon find, this was under the Royal Shadow (Japanese) brand. Gigastorage media was most commonly found under the Melody brand. Nevertheless, these discs were scanned at 8x because they could not consistently tolerate 16x. The results are good for 8x, but there are clear differences in FE with some laying mostly flat, and others having a mid-diameter bump. The TE characteristic seems similar across most discs, but in the inner diameter, some discs have just one bump and others have two.

Variability – Type CMC MAG-D03-064 (8x DVD+R DL)

I did not intend to cover dual-layer discs, but seeing as I had the script ready and had purchased some for an earlier post (that still hasn’t been assembled), I decided to give it a try. I have two spindles of this type – one under the HP brand (usually reputed to be the “best stuff”) and the other under Intact Media branding (a local media reseller).

Scans of the HP-branded discs seem to show a consistent TE bump exceeding the threshold limits on the outer diameter of L1, just after the layer change and elevated TE towards the end of L1 in the inner diameter. Some discs show a pronounced “shoulder” in terms of FE towards the inner diameter, appearing as a step on the beginning and end of the scan. This seems to be characteristic of this media type. Overall, the results within the spindle seem mostly consistent except for a disc with more spiky TE, but some discs seem to have higher FE than others, with some discs having a bit of waviness in FE, despite all remaining within limits.

Surprisingly, the Intact branded discs were free of this major TE bump, instead showing a consistent thin TE spike on both L0 and L1 in a mid-diameter position. Some discs of the Intact-brand seemed to fail scanning prematurely – I’m not sure entirely why. The FE characteristic “step” is still there, but I’d argue these discs are likely better than the HP-branded ones.

Variability – Type RITEK-S04-066 (8x DVD+R DL)

When it comes to DVD+R DL, Ritek discs are often a cause of a lot of pain. I have two spindles of these discs, both under the Ritek Excellent branding.

The first spindle seems to have mostly decent discs, with the peak in TE aligning with the layer change, but often in L0 rather than in L1 as per the CMC discs above. Many of the discs are decent, but there is one significant outlier which show both TE and FE deviances (perhaps the disc is somehow damaged?).

The second spindle seems to have a relatively consistent characteristic. The layer-change TE increase is there, but the elevated TE spreads a little into L1 with an elevation in FE on some discs. But more importantly, in L0 near the inner diameter, there is a consistent hill of TE which distinguishes this spindle from the other. A few of the discs also have isolated TE spikes and the FE may elevate or show “waves”. Therefore, it seems clear that there are spindle-to-spindle differences, even if the discs within the spindle are somewhat similar.

Variability – Type MKM-003-000 (8x DVD+R DL)

For DVD+R DL, Verbatim is usually the go-to choice. I only had 10 discs to test, so here they are:

The TE is impressively low and consistent for L0, less so for L1 where exceedances occur. But consistency-wise, some discs show spikes and waviness in the FE, while L1 having a mid-inner-diameter hill seems to be a characteristic of this spindle, save for possibly just one of the ten discs which doesn’t show it prominently.

Conclusion

While you might think shoving 1200 blank DVDs into a drive, one-by-one, while trying your best not to get any dust or fingerprints on them is not a fun task, it was something that needed to be done (at least, in my opinion). This task actually proved quite illuminating, as it confirmed what I had long suspected – sampling a single disc and running a TE/FE scan on it may not be representative of a whole spindle depending on the manufacturer and your luck as there are outliers. Each disc has its own individual distinct character, although some spindles might have discs that share similar defects or variations more than others. But even in such spindles, sometimes you can see two, or more different patterns, or in some other cases, everything seems fine but you get one significant outlier.

That is not to say that certain media codes aren’t better than others, as there are other factors that go into making usable quality discs (e.g. dye type) and even more factors that go into making a quality burn. But even within the same media code, there is the potential for significant variation. As a result, it’s not unexpected that high speed recording might see speed fall-backs due to hitting one of the less well-made discs. Based on the absolute TE/FE figures, some discs appear quite marginal or even well beyond failure for recording at their rated speeds, suggesting that the manufacturing plants might have had occasional process variation that wasn’t caught by sampling quality control checks or that such outlier discs may have degraded suddenly in a way that the remainder of the spindle did not.

It also reminds us that verifying burns after making them is a good idea – you never know when you might hit one of these outlier discs and get a “bad burn” despite doing well on the remainder of the spindle. But if your drive supports running a TE/FE scan, then running it prior to recording might be a good way to reassure yourself about the physical quality of the disc prior to selecting a burn speed. But having a good TE/FE doesn’t guarantee a good burn – it’s just a prerequisite factor necessary to deliver one.

But based on these results, I can definitely see bad individuals spoiling a spindle and bad spindles spoiling the reputation of a media code. Maybe they’re not just a single spindle, but may affect boxes of shipments under various brands, but they’re still out there … so blaming the media for a bad outcome is not quite as much of a “cop-out” as some might believe. After all, everything works together in a system.

Bonus: Site Update

Hello again and sorry for the silence. I was expecting to get a few more posts out before I left to Singapore, but alas, things got very busy very quickly so posts had to wait until my return to Sydney. At the moment, I am still recovering from my trip, but I felt well enough to get started once again on sharing things that (perhaps only) I care about.

Nevertheless, I managed to enjoy two non-stop weeks of touring Singapore with my Mum which she also enjoyed. The weather held-up for us most days and it was amazing to see just how much it had changed in the nine years since my last visit. It’s still hot and humid, but with a bit more haze and a lot more construction happening. The infrastructure is amazing, food is very decent while being affordable and their wildlife reserves remain world-class. Thanks to all that made our trip a memorable one and for the friends who have hosted my visit as well.

There’s a lot of stuff to process, but I will blog about the trip in due-time. But until then, there’s still a backlog of posts to get through … hopefully before I start work again (which will be soon, as I’ve accepted an offer). I hope you’ll stick around for more of the same, random mix of technologically-related postings and as always, thanks for all your support.