One thing that made owning a late-model BenQ DVD recorder worthwhile was the decent writing quality it had out of the box, along with the amount of tweakability the unit had. Knowing what I know now, an earlier Plextor drive (that was designed in-house, and not rebadged) was the one to die for – with its fancy PlexTools with GigaRec, PowerRec and other features, it had the widest flexibility amongst all optical drives. That being said, the BenQ DW1640 did become rebadged as a Plextor PX-740A, implying the BenQ drive was especially worthy, even if it wasn’t as great as the “real” Plextors that came before it.
Most of the tweakability for BenQ drives is controlled through QSuite, a small tool supplied with the drive and downloadable online which allows you to configure the options. In this post, we will look at all of the Qsuite functions and do some Qscans of present-day media. As far as I know, version 2.1 is the final version prior to BenQ’s exit from the optical drive business, selling their assets to LiteOn, and helping to form Philips LiteOn Digital Solutions (PLDS). In all, the BenQ and Philips joint venture lasted just four years.
On starting the program, you are warned that it only works with certain BenQ drives, and that changing settings may have side effects. It also warns you that the test results can vary, as a disclaimer as the tool mightn’t be consistent across drives and different media – it’s consumer grade hardware, not a high-end optical media analyzer.
The body of the program consists of a tabbed interface which defaults to the Information tab. When a supported BenQ drive is selected through the drop-down, the information tab allows you to view information about the drive and the disc inserted. It gives you the media code (with some bugs for DVD-R discs) but more importantly, whether the media is supported by the Media Table within the drive. Only discs supported by the Media Table within the drive (i.e. containing a write strategy) are likely to give consistently good quality burns.
The next tab is used for changing the Book Type of discs written by the drive. This is necessary to make DVD+R/RW discs more compatible than DVD-R discs by setting their booktype bits to appear as a pressed DVD-ROM. Earlier writing software were not aware of the commands necessary to control the booktype, and thus this provides an ability to set the preferences on the drive itself. DVD+RW discs can be booktyped to DVD-ROM as well, but doing so can prevent some software from writing to or erasing such discs, so the bottom part allows for the booktype of a DVD+RW disc to be rewritten.
The QScan tab is by far the most important of the suite of tools. This is a tool that allows you to gauge the physical construction quality of the disc in terms of tracking and focus error. Through a process of simulating writing at a given speed, the TE and FE are plotted versus a set criteria to judge whether it is advisable to record a disc at a given speed or not. The process does not consume the disc and does not record anything to the disc. It can even be carried out on a recorded disc, but is only effective for DVD±R (and DL) discs. The test has a setting for the test speed, as well as the sampling rate. Even with it set to full, the tests are much quicker than an actual burn as it takes scattered point-samples instead of continuously evaluating the TE/FE.
After running the test, you will get a recommendation as to whether the disc is suitable for writing at a particular speed or not.
Tests done at 2.4x seem to run into a problem and never finish, causing the drive to drop off the IDE bus as it takes a long time to complete some action “on its own”. As a result, the message above may appear. Because the QScan tool is so useful, I will look at it in a separate section below.
WOPC, short for “walking optimum power control”, is a strategy that allows the drive to pause writing from time to time to check the write quality and adjust the laser parameters to attempt to maintain writing quality. This process increases burn time slightly, and can cause some step variations in the writing quality of the disc, but generally helps keep burn quality consistent. In the cases where it doesn’t help or a faster burning speed is desired, it can be disabled.
OverSpeed is an option that lets you burn 4x or 8x media at 8x or higher speeds. By default it is turned off, but the BenQ DW1640 generally has excellent overspeeding properties and can write a wider variety of older media at higher speeds.
Part of the reason for its excellent overspeeding capabilities is the Solid Burn algorithm. This is a technology that uses the inner and outer test areas of the disc to optimize the write strategy prior to the burn starting. Optimized parameters are stored in the EEPROM and improved burn-to-burn. As a result, even unknown discs with no write strategy support can see a decent burn on these recorders. Solid Burn is generally unnecessary for discs which have media table support, but I find burn quality to be improved with it left on. Its use is not without drawbacks, which I will elaborate in a future posting. In case the drive’s self-learning features have been led astray (e.g. by mixing discs with “faked” media codes with genuine blanks) and the burn quality is problematic, the clean tool can reset the learning data to allow the drive to start fresh. This technology now exists in LiteOn’s HyperTuning/Smartburn, although traditionally with most drives, only low speed recording is possible for unsupported media with absolutely horrible quality.
For changes to these settings, they can be made temporarily (i.e. in the burner’s RAM) or persistently (i.e. in the EEPROM). The software allows you to make this choice to avoid unnecessary writes to the EEPROM which could shorten its lifetime.
The Test Write tab turns on a “simulation mode” for the drive, ensuring that no write actually occurs. In the CD-recorder days, a specific command existed to allow for simulated writes – this allowed for testing to ensure the source was fast enough to prevent buffer underruns, to ensure the laser could follow the disc’s track during recording, and to ensure the recording software didn’t encounter any errors to prevent wasting a disc. This specific command did not carry over to the DVD era, and thus in many cases when a simulation is attempted, a drive will actually burn a disc. This allows you to run simulations anyway, by disabling the drive’s ability to write, so you choose to write a disc as normal in your software to run a simulation. This does have a danger that you could inadvertently leave it switched on, and thus not burn anything you thought you were burning.
Aside from the ability to save screenshots, that’s basically all the QSuite has within it. Further drive capabilities are available with third party software – something we will look at in a later post.
I decided to run a bucketload of QScans on my DW1640 with the blank media I still have in my possession to demonstrate the advances in disc TE/FE with improved manufacturing for higher speed discs. High speed burning is enabled not only by more sensitive dyes, but also with better physical disc construction in terms of precision centering, balance, track definition, warpage, etc. With Over Speed enabled, the drive is able to run TE/FE tests at 16x, 12x, 8x, 4x and 2.4x (or 2x in the case of DVD-R). I suspect the display of 2.4x for DVD-R is merely a bug with the software that was never corrected. Of course, good QScan results do not guarantee good burn results as the laser power and waveform need to be correct to get a good response from the dye, but a poor QScan result generally guarantees a bad burn as the drive will have problems even getting the laser to the dye properly. The results will vary subtly from disc to disc (depending on manufacturing variance) and from drive to drive (in case of wear, or different models).
The oldest disc in the collection is an MCC00RG200, which is a Verbatim branded 2x DVD-R which is media code supported by the drive. The focus error is high for 16x and 12x, but tracking errors are not too bad at 12x. By 8x, with the exception of a spike of FE, it seems possible to track the disc accurately. At 4x, it seems absolutely fine, which matches my previous experience of overspeeding these discs to 4x with absolutely no problems at all. Interestingly, the FE gets worse at 2x, which may have to do with mechanism resonance or some other effects.
Another old sample is an LGE04, an LG branded 4x DVD-R that is not natively media-code supported by the drive. This disc shows universally high FE and TE values at 12x and 16x, only settling to mostly within thresholds at 8x. It’s only completely within thresholds at 4x, and looks excellent at 2x. So it seems that over-speeding on this disc might not be that advisable, but 8x may well be achievable. It also shows the variation in blank disc quality from vendor to vendor – even if it is a “higher speed” rated blank.
I managed to dig out an old CMCMAGAF1 4x DVD-R, sold under the Imation branding and supported by the media table of the drive. Like the LG disc above, the TE/FE readings are wild and elevated at 16x and 12x. By 8x, it settles down with the exception of the inner radius which seems to be high regardless. At lower speeds, a spike in the FE appears which may be a disc defect, and the BenQ does not “pass” the disc at any speed, although it seems 2x may be the safest bet.
The next-generation version of the disc is the CMCMAGAE1, an 8x DVD-R under the Shintaro brand and also supported by the drive. It has a slightly less purple coloured dye, but more importantly, this one has much improved TE/FE figures. It’s probable that burning at 12x may be safe.
The next sample is an MXLRG03, a Maxell branded 8x DVD-R also natively supported by the firmware. Maxell have had quite a positive reputation in the community for good quality media, and this shows as the media only just exceeds on TE at 16x, and seems possible to burn at 12x without issue.
The newest is an MCC03RG20, sold in a bulk spindle of printable discs rated for 16x. This disc is also natively supported by the firmware. While it is claimed to be 16x capable, it seems the FE/TE exceed limits at the outer edge when burned at 16x, so restricting to 12x seems advisable. This is not uncommon, as with most burners, using the maximum speed is often a bad idea.
The oldest +R disc I have to hand is a RICOHJPNR02, an 8x DVD+R under the TDK brand, natively supported by the firmware. While the previous generation R01 4x discs did well for overspeeding, my experience was the R02 wasn’t especially great towards the outer edge, and the graphs seem to show this. I suppose if you don’t fill the disc, you can get away with 12x or 16x, but best to stick to 8x.
Another 8x rated disc is the CMCMAGE01 which I also got under the Shintaro brand, also natively supported in the media table. Unfortunately, this one has rising TE/FE towards the outer edge, making overspeeding not recommended.
The latest discs I have are the MCC004 16x rated DVD+R discs under the Verbatim label, also present in the firmware media table. These seem to have a bit of a spike in TE about 5% in, but it seems that 12x burning may be possible, with 16x a bit of a stretch at the outer edge.
Due to a bug in the software, all of the DVD+R DL discs are claimed not to be supported by the media table, but through other means, it was determined that all tested discs were indeed part of the media table. The oldest is a 2.4x rated RICOHJPND00 branded under the TDK label, where no overspeed was offered. The test showed some spikes in TE/FE but it generally was within limits.
The next-generation was a RICOHJPND01 8x rated disc also under the TDK label. This disc showed higher TE on L1 versus L0, although FE remained relatively constant across both layers. Burning at 8x looks borderline based on the TE results, but probably not impossible.
Next-up is a RITEKS04 under the Laser Co brand of Bulkpak. This is another 8x disc. This disc shows a pronounced increase in TE towards the layer-change point at the outer edge of the disc. While overall TE is lower, the spike at the edge suggests problematic burn quality at 8x at the edge, so best to run 4x or less.
Finally is a MKM003 of 8x rating from Verbatim, which is considered a quality media. Contrary to the other discs, it seems that FE is an issue at 8x, with some pronounced TE spikes (as noted in other MCC discs as well). Burning at 8x is probably possible as the exceedance is limited, but not recommended.
Does it work on a BenQ DW1650?
While the DW1640 was my favourite drive, I have been gifted a “new-old-stock” spare OEM DW1650 with a newer Philips chipset. While I’ve dabbled with the DW1650 and found it slightly inferior, I never checked to see if it could do a QScan. I’m happy to report that it works practically the same as the DW1640 although it seems 2.4x selection is unavailable for most media.
Here is a scan of another sample of MCC 004 which shows a similar peak in tracking errors about 10% and 25% of the way in, which seems to be a peculiarity in the batch of Verbatim 16x DVD+R discs that I have. According to the tool, it’s not suitable at any speed due to the spike in TE, but generally speaking, burning at the full 16x is overly optimistic, and 12x is probably going to end poorly.
QSuite was an important part of making the most of your BenQ DVD burner. While the burner generally had good “default” choices put in, the software had the ability to change these options if you wished. The main attraction was the QScan feature, which allowed you to assess the physical construction quality of a disc through measurement of focus and tracking errors at simulated write speeds without writing to a disc. It even worked on already recorded discs (although results can vary slightly after recording). This could save you coasters from being too optimistic with your write speed choices. The majority of the software was demonstrated with a DW1640, although a DW1650 with a slightly later Philips chipset was also used and found to work just fine.
As for the media, it seems that higher speed discs universally showed better TE/FE values, as necessary to ensure quality burns. This implies the manufacturing of the discs were improved along with the dye formulation to make high speed recording a reality. That being said, some of the older media were surprising in its TE/FE stability – a look at the Maxell 8x DVD-R or the Verbatim 2x DVD-R shows that not all older discs were as “wobbly” as the others. That being said, the newest discs still didn’t give perfect TE/FE values towards the outer edge, which probably represents a limitation with the mechanism of the burner combined with the difficult physics problem of keeping the outer edge of the disc in perfect tracking and focus.
Of course, the drive is capable of even more interesting things which I will look at in a follow-up post, but interestingly, FE/TE scanning ability was present in some LiteOn drives which were supported by DVDScan, although it seems that firmware hacking has hit an all-time low with the irrelevance of optical media, lack of willing hacking expertise and the encryption/restriction of firmware distribution all playing a part in making this a thing of the past.