In the last post, I described my glee at receiving a very competent Sony Optiarc AD-7280S. As it would turn out, in this post, I give it some torture as I relive a bit of a hobby from my past, while trying to set a new record – the most oversped DVD burn which is still readable. We’re not talking going from 4x to 6x or 8x to 12x. This drive is going to let me take things to an extreme level.

Why Do Discs Have a Write Speed Rating At All?

We’re probably used to the fact that blank discs are sold with a rated speed or range of speeds supported by the media. For blank DVDs, 1-2x, 1-2.4x, 1-4x, 1-8x, 1-16x are all common speed ranges for recordable discs of varying vintages. But have you ever once stopped to ponder the question of why do recordable optical discs have a write speed stated at all? After reading a good number of ECMA standards and through a few practical experiments, the answer seems to be for a few reasons and not just one.

The first reason would most likely be the dye itself. The dye is the chemical layer which is altered (burned) by the laser – in DVDs, it is often a purple colour and is likely to be some formulation of a metal-AZO, Oxonol or Metallic Dipyrromethene (the latter seemingly having relatively little information). The specifics of how the dye is formulated and how it responds to laser light is the key to whether a certain dye is suitable for high speed recording. A good dye for high speed recording needs to be highly sensitive as the laser power is limited from the drive (as semiconductor lasers are limited in power and lifetime/cost concerns prevail), ideally have a wide power margin (so as to be less sensitive to laser power drift) and still be highly stable in the presence of lower-powered read laser and environmentally normal levels of light. If the dye isn’t sensitive enough, the drive simply cannot produce a decent “mark” in the short window it’s firing its laser at the dye.

The second reason is due to the required metadata. Discs have embedded within their stamping, metadata that describes the burn strategy to be used. This information basically indicates the target power at a given speed, laser pulse timings and offsets necessary to produce a good burn. Usually, older lower-speed discs don’t have this data embedded in the disc. While there is the embedded data, many burners do not use this data, instead preferring the route of identifying a disc type through the Media ID and matching this up with a table of write strategy data located in the firmware. In many cases, the firmware tables have entries for the rated speed, occasionally some drives can over-speed selected media by default, based on what the drive manufacturer has optimised the drive for.

The third is the mechanical construction of the disc. Different discs manufactured by different manufacturers simply “handle differently”. Higher quality discs are often more rigid, returning excellent tracking error (TE) and focus error (FE) scores in media tests – this means that they are often created with better stampers and are better balanced and spin flatter when at high speeds. This is important as disc warpage and wobble on the outer diameter, where the disc is furthest from the spindle which supports the disc, is at the highest data rate section and thus can compromise the write process if it’s not stable enough that the servo mechanisms can compensate for it. This is slightly less critical for readback, but can also cause speed fall-backs during reading if severely lacking.

It seems that these factors are the predominant reasons behind discs having a speed rating at all – but drives can and do routinely disobey them. In the earlier days, drives were factory-permitted to overspeed on selected high-quality media. Later-on, with the advent of BenQ SolidBurn adaptive write strategy and Overspeed, later even with Lite-On’s Hypertuning and Overspeed, users could try overspeeding media themselves, even unsupported media, but to a more limited extent. Finally, overspeeding could also be enabled through firmware modification – most commonly a strategy rename or strategy swap which assigns a given media code the strategy for another code that may provide higher speed burns – but this means (implicitly) using the strategy for that higher-speed media on a lower speed disc. In many cases, quality burns could not be assured and mild overspeeds (e.g. 4x -> 6x, 8x -> 12x) tended to be most successful. Higher overspeeds were rarely achieved consistently, at least in my prior experience, with this hobby dying out as later drives rarely had firmware support in modification tools such as Omnipatcher or MCSE.

I understand this is a bit technical for those who may not be optical media enthusiasts, so here’s something that might help …

A Paper Analogy

To best explain the concept, I would like to create an analogy. For this analogy, you will be an optical drive. The laser power is the sun, you have a cover you can use to stop the sunlight from going through and the optics are the magnifying glass in your hand. The dye layer is represented by a piece of crinkly paper, with different types of paper representing different speeds. The paper itself has lines ruled across it – these are analagous to the spiral groove on a disc. 

To “burn” data to the paper, you need to create a series of well-defined burn-marks of a particular length, in the right place relative to the pre-ruled lines. If the lines on the paper are too feint, or if they deviate from straightness too much, that makes your job of following them just a bit harder – this is the importance of tracking error and stamper quality. If the paper itself is curled at the edges or crinkled a bit too much, you’ll have difficulty following the hills and valleys of the paper to make sure the sunlight is in focus to burn your spots – that’s representative of focus error. Of course, the paper itself is clipped to a clipboard at one edge and moving while all this is happening (like a disc is rotating), so the paper may flap around a bit – that’s representative of disc warpage that occurs most on the outside diameters of the disc where the disc is not supported and imbalance can cause vibration. So ideally, you’d like a flat sheet of paper with perfectly sharp ruled lines that stays flat while it’s moved about. That is the “physical” aspect of the media construction.

The next would be the paper itself. What is it made out of? How does it burn? If it’s tissue paper, as soon as you ignite the paper, the whole thing “runs away” like a chain reaction – it’s just too sensitive. On the other hand, if you have a bright white thick wax paper, maybe it won’t burn at all. Ideally, you need a type of paper that starts to burn almost immediately when exposed to focused sunlight and stops burning immediately once the sunlight is removed by you covering up the magnifying glass. In reality, this is almost never going to be the case – so instead, we can relax this to say that you need to know how early you need to pre-heat the paper to achieve a burn spot by a particular point and how early you need to stop heating to make sure the burning stops by a particular point. But to make this even more tricky, your special sheet of paper also needs to be resilient to half-way sunlight, so that someone can read those lovely burn spots you put on the paper, without itself going brown. This is the dye response characteristic and the dye stability characteristic – this is what the “write strategy” of a drive or disc optimises to make sure the marks are well formed and stay that way to keep your data safe. Thankfully, your paper may have a brand on it (the media ID) and some instructions on the outside on how to best use it (the embedded metadata).

But it’s a little more complex than that. In reality, you might have the secret formulas to burn spots onto your paper at just the right time, but in reality, those spots start and end just slightly off the intended location. This is jitter and it can happen because your sunlight isn’t steady, your timing isn’t quite right, maybe the paper isn’t moving at quite the right speed or the paper itself isn’t made the same way. Depending on the power of the sun, the paper could turn a bit too dark and cause nearby light spots to go brown, or maybe not dark enough and the burn spots could fade away. This is the “beta” or symmetry of the burn, in a rough way of describing it. Thankfully, even if the marks are a little off, you can usually still work it out thanks to error correction. But luckily for you, there are a few practice pages so you can get your technique right before having to burn a whole volume of data. That is analogous to the power calibration area of a disc.

Let’s consider the problem of burning the paper quickly. The faster you move your spot over the paper, the less sun energy is given to the paper, but you still need to make a good burn mark on your paper. That means, ideally, you need a paper that is as sensitive as possible without being so sensitive that it is disturbed by ordinary light. The timings for a given mark length on the paper get shorter because it’s moving faster, so you have to react more quickly and make sure you’re still starting and stopping at the right times (or find paper which is a bit more tolerant to imperfect timing). The paper, moving quickly, makes it more difficult to stay in the lines – so those line markings need to be extra straight and clear. Finally, the paper needs to stay flat, so that your sunlight “spot” burns properly, in focus. Only when all these stars align do you get yourself a nicely burned piece of paper (or disc). It’s a miracle it even works at all.

Why the Sony Optiarc AD-7280S?

The game of overspeeding was often played using older drives as the firmware was well supported by modification tools. Unfortunately, such older drives are starting to become rarer but also, as with my recent investigations into high speed burns, it would appear that newer drives have an advantage of better stability at high speeds and a more adaptive and flexible approach to write strategies. Drives tend to use the power calibration area in a more intelligent way to tune the write strategies, rather than just blindly following the data in their Flash ROM, which also explains why coaster rates have gone down even with counterfeit media achieving usable results at speed. That was definitely not the case in the early days of DVD recording where such media with stolen media IDs would almost certainly produce problematic results.

So why did I choose the AD-7280S to run this experiment with? My previous 24x burn experiments with the Asus DRW-24D5MT showed that MediaTek chipset was definitely burning “adaptively” having constantly changing beta and jitter throughout the burn. But the problem with that drive was that I didn’t have a way to extract the firmware and even if I did, the modification tools would not support it. My experiments with the iHAS324 @ 624 B were more promising, as a Lite-On, the firmware is supported by MCSE just fine. The problem with that drive was that it was struggling to give a good result at 24x on Japanese-made TYG03. This suggested the drive may not have enough laser power or servo compliance to make it a good contender for overspeeding slower media (which would have characteristics worse than TYG03).

The AD-7280S impressed me during the commissioning tests. But even more impressive is that the firmware was extractable and supported by MCSE. This was everything that I needed. As a bonus, I had some more strategies than just the Taiyo Yudens to choose from, if I ever needed it – 24x is available for RITEK R05-012, YUDEN000 T03-000, MBI 01RG50 and TYG03. But in principle, I stuck with T-Y codes for the strategy swaps.

Results

Candidate 1: Maxell DVD-R (MXL RG04) 16x @ 24x

This is a new media favourite and one I felt was deserving of 24x burns. The TE/FE shows it to be borderline on the old BenQ DW1640 but that’s good enough for the majority of later 16x drives to do 16x on. Can we do 24x?

The write completed in 4 minutes and 2 seconds – the standard burn time for a 24x burn. A small hiccup near the middle but a solid result at the end. Welcome to the 24x club!

The AD-7280S reads it back just fine and the quality scan indicates an excellent burn, even not considering the write speed.

The BenQ DW1640 similarly likes the burn and gives the jitter an amazing score for the speed.

No problems on the DH16A6L either with a smooth readback. The jitter shows more ordinary levels but the PIE/PIF results are absolutely amazing.

My old SHM-165P6S is usually quite picky and it’s not so happy on the jitter. Virtually no discs can meet the guideline on that drive, but this is still a bit higher than I’d like to see. The beta level is a bit low too, but it’s in the acceptable range most of the time. For reading back on modern drives, it seems unlikely to cause issues – hence the PIE/PIF rates are quite low.

In all, I think I’d call this a success – on the AD-7280S the MXL RG04 can overspeed to 24x with a strategy swap to TYG03.

Candidate 2: Mitsubishi DVD-R (MCC00RG200) 2x @ 20x

In my previous review of the disc, the discs were remarkably well-made and stable. The one I happened to grab wasn’t so good on the inner diameter on the tracking error, but the outer diameter is no worse than the Maxell. It seems possible for 24x in theory, so I requested that but the drive wasn’t able to make it.

A 20x burn on a 2x disc is still far faster than any overspeed I’ve ever experienced. Writing took 4 minutes 40 seconds – the drive did spin up to 24x speeds, but at the outer diameter, during calibration, the drive spun down one notch to 20x before the burn started. This suggests to me that the drive was unable to meet quality or tracking targets on the outer and thus limited the speed. But we already know that the tracking seemed okay – so I suspect the dye is the cause.

The written disc is readable and the quality scan from the AD-7280S suggests it’s a fair quality burn – not great on the inner and outer, but still within limits and with excellent PIF scores.

The BenQ DW1640 read it back just fine too – but the quality scan shows quite problematic PIE that exceeds limits on the inner diameter. I suspect this is a representation of radial tracking error induced burn quality deviations.

The DH16A6L agrees that the inner diameter is not as good in terms of burn quality but this time, the PIEs are within limits. The more concerning issue is the increasing jitter with increasing diameter – this suggests to me that the dye might not be sensitive enough for such high speeds, but the recording is still readable at present.

The SHM-165P6S shows a record level of jitter and a similar trend to the DH16A6L, but the drive also shows a declining beta as well, despite starting off at a reasonable value. I suspect this suggests to me that the dye isn’t responsive enough for such high speeds.

In a bit of greed, I tried again a few times to see if I can improve the situation, but it turns out, those other discs resulted in speed fallbacks at the outer edge. I suspect the hypothesis is correct – this one falling back to 18x managed to hold the beta stable.

This second attempt falling back to 4x managed to haul the beta all the way back to zero.

Finally, this one with multiple fallbacks to 4x in the end also showed an improving beta situation with decreased write speed at the outer.

This is a record overspeed, going from 2x to 20x, on a readable burn basis. However, it’s not done reliably as it likely runs into either a laser power limit or the dye itself simply doesn’t respond to such short bursts of laser power, so the recording becomes worse towards the outer diameter. The incomplete burning of the dye could result in stability issues as well – so I don’t recommend doing this for actual burns you care about, but it’s a demonstration of how a disc which may be mechanically stable enough for high-speed burning may not have a responsive dye that enables quality results.

Candidate 3: Unifino DVD-R for Data (UTJR001001) 4x @ 24x

I previously reviewed these discs too and found the “for data” discs to have excellent mechanical construction characteristics and a decent dye layer. While others had found the media hard to achieve good results with on burners of the era, late model burners actually achieved excellent results at rated speed and even at overspeed – achieving 12x on my DW1640. I thought an overspeed from 4x to 12x was pretty good already, but can we make a defunct media manufacturer join the 24x club?

The answer is, yes. The solid construction of the disc and the response characteristic of the dye allowed the drive to proceed with a burn. It took a few seconds longer to dial in the characteristics, requiring 4 minutes and 9 seconds (so about 5-7 seconds longer than usual) but it managed it in the end.

Best of all, the recorded disc is fully readable. The quality scan in the AD-7280S is well within limits, having excellent PIFs, but the PIEs do show some impending issues just prior to each WOPC adjustment which seemed to bring it back into check.

My DW1640 was happy with a smooth read and an excellent quality scan result including good jitter.

The DH16A6L also agrees.

The effects of adjustments can be seen in the SHM-165P6S’ jitter and beta curve. Once again, the beta is a touch low, but it’s still within the range you might find from other burns.

This is undoubtedly a major success in my books – Unitech Japan’s disc can now be said to have joined the 24x burn club, well after they’ve gone defunct, simply by strategy-swapping with TYG03 on the AD-7280S. The drive simply adaptively takes care of the rest. But beware – different batches do have different construction quality as shown in my review – it’s not a “cop-out” as suggested by The Digital FAQ.

Candidate 4: Taiyo Yuden (TYG02) 8x @ 20x

Taiyo-Yuden was well recognised as the originator of the best media, so what about their previous 8x generation media? Will it be overspeedable in a significant way? Mechanically, the results suggest the disc is good enough for at least 16x on a modern drive.

Strategy-swapped, the AD-7280S decides it’s only good for 20x, completing a burn in 4 minutes and 45 seconds. This suggests to me that the dye response isn’t sensitive enough to go faster.

The resulting burn is readable, but the quality scan suggests somewhat average PIEs and good PIFs. It’s acceptable but not picture-perfect T-Y quality.

The DW1640 seems to read back fine but with a tiny wiggle near a W-OPC point. The quality scan suggests that the drive was changing something significant with the burn strategy, resulting in a choppy looking PIE curve. The jitter seems to decline as the burn proceeds, suggesting the drive was actually “tuning in” to a better burn as the speed increased. Perhaps choosing a different strategy would help, or perhaps if the drive has an adaptive strategy memory, subsequent burns might show improvement.

The DH16A6L seems to show a smooth read curve and a similar choppiness to the PIE curve, but it’s much lower and in the acceptable range. But contrary to the DW1640, the jitter seems to peak in the middle but is mostly stable at an elevated level.

The SHM-165P6S says the jitter is high but mostly stable, increasing towards the outside as the beta is fine most of the way, only becoming slightly low towards the outer.

It’s a success as far as burn completion and readback is concerned, but the result is suggestive of a drive that’s not quite tuned into the best burn strategy and was making some adjustments that caused some severe steps in error rates. Perhaps the dye’s laser margin is less forgiving, making such adjustments more visible. While the 8x overspeed to 20x is possible, I’d not recommend it.

Candidate 5: Taiyo-Yuden (TYG01) 4x @ 20x

While the 8x version didn’t do stellar, I decided to try the 4x-era disc as well because it’s also fairly well constructed.

The drive managed a 20x burn as well, in 4 minutes and 39 seconds.

Once again, the result is a readable burn with a quality scan showing somewhat average PIE values and decent PIFs. On this basis, it might not seem so different to the TYG02 result, but the other drives may not agree so much …

The DW1640 reads it fine with some short dips which might just be the ageing drive. But the quality scan does show fail-level errors on PIE and PIFs while jitter seemed acceptable. Once again, the DW1640 seems to suggest the writer was making some adjustments mid-burn which reflected poorly in the burn quality.

The DH16A6L managed to read it back completely, but the quality scan showed a high PIF spike and relatively high PIE levels in very specific regions. The jitter level was elevated and stepped.

The SHM-165P6S seemed to have difficulty scanning this disc – high jitter, horrible histogram and oscillating beta although not severely low.

In my judgement, while this 4x oversped to 20x reads back okay, it’s metrics suggest something severely wrong with the burn quality which is only saved by the fact that drive chipsets are likely more tolerant than the “reference drive” scanning condition when performing quality scans. Perhaps strategy swapping older generations of T-Y dye to the newer T-Y strategy is a bad idea – maybe the response is close enough that the drive is confused, whereas swapping to an obviously-wrong strategy might have the drive more “comfortably” force itself to go adaptive.

Candidate 6: Ricoh DVD+R (RICOHJPNR01) 4x @ 20x

It’s a trip down memory lane with this RICOHJPNR01. They were famous, of their era, of being capable of overspeeding from 4x to 8x – the discs themselves seem well built but not quite 16x-grade. While the RICOHJPNR02 managed 8x and overspeeds to 12x, I never had much success getting RICOHJPNR01 past 8x at all (and this was with the then-new 16x drive, a Lite-On SOHW-1633S@1693S).

The AD-7280S decided 20x was as fast as it would go – I would have expected slightly less on account of the construction of the disc.

The burn was readable in the AD-7280S but the quality scan shows an average quality PIE-wise and great quality PIF-wise.

The DW1640 read it back just fine too, but both PIE and PIFs are a fail and jitter is high.

The DH16A6L manages to read it back smoothly and seems to suggest acceptable but not great levels of PIE and excellent PIFs, but at a high level of jitter.

The SHM-165P6S doesn’t love it – some adjustment steps are visible in the jitter and the histogram is very muddy, but surprisingly, this is the first overspeed where the beta is within-limits and slightly on the high side rather than on the low side.

In all, the burn is readable on this 4x to 20x overspeed, but it’s not a recommended result, as it does show some quality issues. But as a vintage media famous for its overspeeding from 4x to 8x and no further, the fact that a readable burn results at 20x just shows that something has changed with the drives and their write strategy capabilities. Perhaps it’s stronger lasers and faster switching, but also better algorithms and the ability to adaptively tune the write strategy rather than just tuning power alone.

Candidate 7: Pro-Feel DVD-R (VANGUARD) 4x @ 12x

Finally, I decided to try some VANGUARD. I’ve previously reviewed this media and they appear more of an ordinary 4x media to me. The TE/FE curves are mostly acceptable at 4x when tested in the previous review, but grabbing another few discs for a higher speed test, we can see that they are mechanically deficient for high-speed burning. This will be a good demonstration of what happens when your discs aren’t mechanically stable enough. At 16x, it definitely shows major issues but at 8x, perhaps a tolerant drive might be able to handle it.

Is the AD-7280S smart enough to tell? I chose 24x and let it rip and we managed two coasters. The drive gave up at around 14x-15x, still having not burned half of the disc. As a result, this shows that write failures can still happen with these smarter drives (perhaps they don’t assign enough weight to tracking stability or they get upset by a spot-defect in the stamper).

Reducing the speed to a more leisurely 12x manages a complete burn, but with a fall-back to 4x on the outer edge. Whether this is tracking or dye related is unknown, but it could be a mix of both. Still, a 4x to 12x overspeed would have been significant for the time had it not been for my results before.

The lack of disc stability is telling – readback has dips despite a good quality scan at low speeds.

The DW1640 shows outer dips attributable to mechanical issues but also an earlier dip mid-burn which seems to be the drive mal-adjusting its burn strategy. The tell-tale “stepped” PIE curve which exceeds limits suggests that the drive was adjusting its burn strategy and “fell off” the band of goodness before returning and cycling in and out. The jitter is also higher in that segment.

Even the usually bulletproof DH16A6L did not read this one smoothly – the outer diameter instability was evident, but this drive seemed less perturbed by the mid-burn adjustments.

The SHM-165P6S seems to show some of this adjustment behaviour in the jitter curve which dances up and down. When the drive falls back to 4x, the jitter curve falls into the acceptable zone and the beta trends back to zero – the dye quite likely is “best at 4x” but the drive can push it, at least most of the way, to higher speeds.

But this is a good demonstration that the drive is not a miracle worker that can make every disc burn at 20x or above. The disc has to be good enough, the dye needs to be responsive enough and ultimately, the drive’s adaptation has to be consistent enough such that the burned result is stable for it to be a reliable overspeed. In most cases, I don’t think all the criterion are met, but the result for the MXL RG04 and UTJR001001 at 24x are both very commendable.

What About Rewritable Discs?

Rewritable discs are much more sensitive to power levels and timings as media formulations are made for very specific heating/cooling times. I’ve been long told that overspeeding rewritable media is simply not possible. Having done strategy swaps on MediaTek drives, I noticed that the swap to a higher-speed strategy is simply ignored and the speed selections remain the same as prior to a swap.

But for this Renesas-based drive, that was not true. Swapping to another rewritable strategy with a higher speed resulted in the higher speeds being selectable, so let’s find out what happens in practice!

Candidate 1: Maxell DVD-RW (RITEK W01) 2x @ 4x/6x

Attempting to write this 2x disc at 4x with a swap to a 6x-capable strategy of the same vendor (RITEK W06) resulted in error 053005 CANNOT WRITE MEDIUM – INCOMPATIBLE FORMAT.

Trying again, it decided to throw an 037304 PROGRAM MEMORY AREA UPDATE FAILURE. This tells me that it cannot write in a way that the drive can read it back – the PMA is an area where the drive keeps track of the discs’ status in terms of formatting and currently written sessions, so a failure may leave the medium unusable.

Trying 6x, it threw 037303 POWER CALIBRATION AREA ERROR, thus confirming that the changed strategy simply doesn’t result in the drive being able to find a write power that can produce a readable result.

Candidate 2: TDK DVD+RW (RICOHJPNW11) 4x @ 8x

Another type of disc, another power calibration error. No can do it would seem, at least based on using the RICOHJPN W21 strategy.

Candidate 3: TDK DVD+RW (PHILIPS041) 4x @ 8x

In this case, this 4x disc managed to burn smoothly at 8x – the smooth burn indicates the drive is burning on a strategy basis, most likely non-adaptive, after calibrating once on the PCA using the PHILIPS RW8 strategy.

Unfortunately, as you can see, while the disc has been written, the drive itself reports the disc is empty. The underside has been darkened, but after ejection, no drive could recognise or reformat/erase the disc. It’s been permanently damaged.

Candidate 4: TDK DVD+RW (CMC MAG W02) 4x @ 8x

This disc was burned at 8x using the CMC MAG W03 strategy at 8x, successfully and “blindly” as is common for rewritable media. The slightly-better result here is that the disc is recognised as containing data now.

But the write quality is so bad that the AD-7280S cannot read it back, throwing 030200 NO SEEK COMPLETE. Ejecting this and returning it to the drive or any other drive simply would not read, nor reformat. Yet another disc permanently damaged.

Based on these results, it seems the claims that rewritable discs cannot be oversped is perhaps true, based on direct strategy swapping without any understanding of the material dynamics or fine-tuning of actual strategy parameters (power, timings). The lack of success might be, in part, attributable to the fact that rewritable burns appear to be non-adaptive, based strictly on strategy data.

Conclusion

The Sony Optiarc AD-7280S was definitely impressive and its firmware definitely exhibits adaptive strategy behaviour which can be exploited for overspeeding of recordable media. While the burn quality isn’t ever going to be perfect when overspeeding discs, the result was still super-impressive as the majority of oversped discs were readable. Because of this drive, I’ve managed to elevate the still presently-produced MXL RG04 (Maxell TW likely Ritek Pro) and the now defunct UTJR001001 (Unifino JP, Unitech Japan) to the “exclusive” club of achieving a 24x burn. To think that a 4x-rated disc managed to complete a 24x burn is amazing but does jive with my previous testing – that disc was just seriously well-made. These two burns actually fared quite well in quality testing too, so I’d say this is a genuine achievement.

Another major result was the MCC00RG200 (Mitsubishi JP) discs, which were rated at 2x. One of them managed to burn at 20x as well – unfortunately, it seemed that the drive deemed the tracking error on the outer tracks to be too high to permit a 24x attempt, as multiple attempts often tended to see the drive fall-back in speed towards the outer with a notable “bare” spot in the burn. I suspect the issue here is a combination of the disc not being stable enough, but also the dye not being sensitive enough to maintain a good recorded signal.

Despite the possibility of overspeeding, I wouldn’t necessarily recommend this as it is hard on the drive’s laser. The discs that were oversped tend to show low beta values which only recover after the drive does a fall-back in speed, suggesting that the drive was struggling to maintain enough laser power to make a solid “mark” on the dye and this only got worse as the speed increased towards the outer edge of the disc and as the laser itself heated up.

Still, the results are undeniable. I don’t think I’ve ever seen overspeeds of this magnitude documented anywhere else, nor do I think that there have been other successful 24x overspeeds as well. It might as well be a world record … but alas one nobody cares about. It’s unfortunate that while the drives will write at 24x, they will only read them back at 16x. But that’s been a whole lot of fun and a lot more successful than overspeeding in the older non-adaptive strategy days where strategy swapping already known media usually resulted in quality problems rather than acceptable results.

This was, also, the only drive I’ve ever tried swapping codes for rewritable discs that actually had an effect. Doing it on the MediaTek-based drives, in my past experience, just resulted in the drive ignoring the swap and enforcing the media’s pressed speed limits. While it was an interesting chance to try such an overspeed, as I had been warned, the results were either a failure to calibrate or a destroyed rewritable disc. Perhaps the discs could be revived, but most drives will get stuck trying to work out the state of the disc and simply reject formatting.