When DVD-ROM drives started hitting the market around 1997, many hailed them as the indubitable successor to the CD-ROM drive. A similar transition occurred when CD-ROM drives first appeared. Remember when programs used to come in CD-ROM and floppy versions? After a certain point you had to mail in a certificate to get the floppy version, then eventually, floppy versions of most popular software titles were no longer made.
Three years later, the DVD-ROM is far from replacing the CD-ROM. While we see the occasional new title in DVD-ROM format, for the most part the technology has been used to re-issue older titles in an "enhanced format." Titles such as Riven and Baldur's Gate, originally filling up four or more CD-ROMs, fit on fewer (or even single) DVD-ROM discs. Full-motion video (FMV) sequences are often lengthened, or visual quality is improved. But the CD-ROM obsolescence movement has yet to arrive.
Nevertheless DVD-ROM speeds continue to improve even though DVD-Video movies play at only 1X speed. Faster DVD transfer rates can benefit during movie playback when scanning and searching, but the thrust of faster DVD-ROM speeds is for pushing computer data faster. With such an underwhelming selection of computer DVD titles thus far, most folks will buy a DVD-ROM drive to watch movies on their PC, not to play computer games or transfer computer data.
Presently 6X DVD-ROM drives remain relevant, even though 12X-16X drives are already arriving on the scene. Enter Pioneer's DVD-303S and Toshiba's SD-M1212.
Pioneer DVD-303S - Firmware revision 1.10:
From Pioneer New Media Technologies, Inc. comes one of the fastest-rated SCSI DVD-ROM drives currently available, the DVD-303S. Rated at 6X DVD, 32X CD-ROM, Pioneer specifies a speedy 80ms access time, with a rather large 512kb buffer and a standard 1-year warranty.
As you'd expect from a drive that reads only up to 32X, noise levels were quite low for our test unit. Seeks were very soft and while full-rpm sounds were audible, they were definitely on the low side. During stress testing, our unit was just slightly warm to the touch after extended use; again, to be expected for the comparatively low RPM. Pioneer installs only a standard 4-pin analog audio out.
Toshiba SD-M1212 - Firmware revision 1R14:
Toshiba brings us the SD-M1212, a previous-generation ATAPI DVD-ROM drive from their lineup, with 6X DVD speeds and 32X CD-ROM speeds (their newest DVD offering does 12X DVD and 40X CD-ROM). Toshiba specs this unit slightly slower than the Pioneer, with 85ms access times and a 256kb buffer. The SD-M1212 is protected by a standard 1-year warranty.
Running the Toshiba through our access time stress tests, we noticed that it was slightly louder than the Pioneer. Seeks were muted, but the noise and vibration accompanying high-rpm operations were a bit more audible. Like the Pioneer, the SD-M1212 was only slightly warm after extended testing. Toshiba has, however snuck in a CD-digital line-out. Although they only list analog audio-out in their spec sheet, there is an unlabeled 2-pin connector next to the standard 4-pin analog connector, which, when tested with our SB Live! proved functional. While we cannot guarantee that other units will also grant this "secret toy surprise," we were nonetheless pleasantly surprised with this undocumented feature of our test unit.
Here we see a clean sweep by the Toshiba. In the access time test, Toshiba's ATAPI unit ducks in .1ms under its rating, while Pioneer's SCSI unit finishes 3.3ms later than the Toshiba (and over 8ms slower than rated). We see the trend continue in transfer rate tests, with the Toshiba virtually tying the Pioneer at the inner tracks, and then pulling ahead by 3% at the outer edge of our test CD. Note the Pioneer's transfer rate graph, where we see quite a bit of spiking as we near the outer portion of our test CD. While this didn't affect our performance measures, it does cause a bit of concern. One must question if the Pioneer truly had problems reading our CD, or perhaps it is some sort of benchmarking anomaly. Although it appears to be a significant issue, the lack of any consequences in the upcoming CD Winmark 99 and Copy tests should do a lot to alleviate concerns.
All of this bodes well for the Toshiba in the Winmark tests. As we have discovered, Ziff-Davis' CD-ROM Winmark is highly dependent on low access times. Will the Toshiba continue to pull ahead? Let's find out.
CD-ROM Winmark 99
Surprise, surprise. For the first time in our optical comparisons, access times did not accurately predict the outcome of a Winmark comparison. A closer look is warranted. While the Pioneer's scores deviated by a rather small 3.5%, the Toshiba proved to be very finicky about our four test CDs. It scored an amazing 1660 with our first test CD (this compares very favorably with some of the fastest CD-ROM drives we've tested to date), but averaged in the low 900s with the other 3 discs. It appears the Toshiba unit has a more difficult time than the Pioneer when faced with a variety of CDs to read. The same three discs that the Toshiba struggled with, the Pioneer averaged 1367. It's only when we compare our disc 1 results that the Toshiba comes out on top, 1660 to 1300. As always, when we encounter unexpected results, we retest. Our retests showed a very similar trend in scores, validating our final outcome.
Although the Winmark test posits that access times are of utmost importance, there are many situations where access times play a smaller role in performance (and sometimes no role, such as in purely sequential reads, which occur much more often in optical drives than in hard drives). Let us now move on to our file and disc copy tests, where such situations arise.
File and Disc Copy
Our file copy test generally correlates well with measured sequential transfer rates. Although our low-level results suggest the Toshiba should come out on top (just barely), it's the Pioneer that edges the SD-M1212 by a miniscule 1%- certainly well within normal deviation. For all intents and purposes, they tie not only in sequential transfers, but also in the file copy.
However, when we compare the disc copy results, we see a definite victory by the Toshiba. Through our previously tested units, we've shown that of drives that are of the same speed rating, the one with the lower access time generally does better in the disc copy test. This test requires the drive to read through multiple files nested within multiple folders and sub-folders. Speedy random accesses will of course benefit in such situations; the Toshiba is the swifter of the two. In the end, it copies our test CD nearly 10% faster than the Pioneer.
As we conclude our look at the traditional read performance of these drives, we have a very tight race. On the one hand, the Toshiba shows a definite advantage in access times and disc copying; on the other hand, the Pioneer appears to do a better job of reading a variety of discs with consistency. We tend to place more emphasis on the latter. As we move forward into less traditional areas of optical drive performance, we will see if the Pioneer can maintain its (slight) lead.
Digital Audio Extraction
The Pioneer is essentially a 12X CLV audio extractor. There are times when its extraction speeds dip slightly under 12X, so you almost certainly would have difficulties burning on the fly with a 12X writer. However, the Pioneer should be able to feed an 8X burner easily. The same cannot be said for the Toshiba. Even its maximum extraction speeds are a woeful 6X. Unless you burn at 4X or less, forget about recording on the fly.
Both drives extracted with no errors in our DAE Quality test. Subjective listening of some .wav files ripped from our test CD-DA confirmed CDSpeed99's results-to our ears extracted audio sounded no different to the source.
However, one should be cautious when using the Toshiba for audio extracting. CDSpeed99 reports that the SD-M1212 does not support accurate streaming. What is accurate streaming? A bit of background is in order.
Data on audio CDs is not grouped in a strictly sequential nature. There is some interleaving of data, which helps to protect the CD-DA from physical damage. If there is a scratch on the audio CD, a tiny portion of a number of audio sections may be damaged, rather than a significant portion of a single audio section. The end result is that you are far less likely to hear any audio defects when a scratch does occur. A by-product, however, is that interleaving makes it more difficult for the drive's head to quickly and accurately position itself. Optical drives which support accurate streaming are able to more quickly and accurately position the drive's head because the commands necessary to do this are hard-coded into the drive's controller chip, rather than depending on the DAE software to perform the complex algorithms. This greatly enhances the speed and efficiency of audio extraction and is probably one reason why the Toshiba extracts so slowly. It does not support accurate streaming, so more time must be taken to provide an error-free extraction.
Now that we have examined pressed CD and CD-DA performance, let's see how these drives read recordable and rewritable media.
CD-R Media Compatibility and Performance
CD-RW Media Compatibility and Performance
Returning to the Toshiba, we see that it drops read speeds considerably when faced with a CD-RW: It turns into a 10X-20X CAV reader. However, it does so with grace- the transfer rate graph is placid.
Now that we've thoroughly examined each drive's CD-ROM performance, let's see how they perform when reading the type of media they were primarily designed for.
DVD-ROM Performance Results
Currently, the latest version of DVDSpeed99 has an incompatibility with certain drive models (the SD-M1212 among them) when running the seek tests. Because of this, we have elected to skip over the seek test scores for all drives until this issue is resolved in a later revision. Our main focus will be on transfer rates, to determine how closely each drive performs to its specs. As we can see, both drives yield very similar transfer rates. The Toshiba has a slight advantage, reaching 5.75X, while Pioneer's unit peaks at 5.59X with our test DVD. Neither drive reaches its 6X rating, though both come very close.
Looking at the CPU utilization scores (measured at 1X DVD speeds), we see the Pioneer with an astonishing 43%, while Toshiba's ATAPI drive averages 7%. In researching this huge discrepancy, I was informed by DVDSpeed99's author that this issue is specific to the 1.10 firmware revision of Pioneer's DVD-303S. Unfortunately, our Pioneer evaluation unit was returned before we could attempt any testing with other firmware revisions. Our testing with other DVD drives thus far has shown results more in line with the Toshiba, so perhaps there is an issue specific to Pioneer's 1.10 firmware.
Does the slight transfer rate advantage of the Toshiba, along with its apparent CPU utilization dominance affect the end result...watching a movie?
Subjective Playback Observations
Our subjective impressions are based on watching portions of two DVDs-Twister and The Matrix. We focus on two particular scenes. In Twister, we closely examine what we've coined the "wheat field fly-by" (scene 2). Specifically, we watch for any jerkiness in the camera panning as the scene rapidly approaches the moving vehicle driving swiftly through the wheat fields. In The Matrix, we focus on scene 15, the Morpheus/Neo Matchup. Again, we watch for the level of smoothness in rendering this fast-paced fighting scene.
Both drives performed outstandingly. Playback through WinDVD on our P3-450 testbed did not show any noticeable jerkiness. Color saturation and clarity were excellent on our 21" Sony 500PS monitor. Overall, both drives provide excellent movie playback capabilities. The Pioneer showed no signs of taxing CPU resources.
Overall, both drives perform quite similarly except in DAE, where the Pioneer was clearly superior. The Toshiba has speedier access times and disc copy results, but performs inconsistently when faced with a variety of discs. The Pioneer is a tad slower in access times and the disc copy, and its spiked transfer rate graphs raises some questions, but it is a much more consistent performer than the Toshiba in application-level tests.
In the end, recommending one over the other is difficult due to the different interfaces. Generally we compare within interfaces rather than between, but those without SCSI host adapters should not feel compelled to move to SCSI strictly for the purpose of getting the DVD-303S. However, those who already have a SCSI-based system would almost certainly be happier with the Pioneer. The consistent performance and DAE superiority are sufficient enough to rank it ahead of the Toshiba, when interface is no object.