aIn FCM#61, I wrote about DVD ripping and encoding. For the article, I tested the ripping and encoding speeds on single core, dual core, and four core systems. As expected, the four core systems had the fastest speed from start to finish. But there was a surprising result as well, the DVD burners in the dual core systems ripped (cached the DVD to the hard drive) almost 3 times faster than the quad core systems (and just less than double an eight core system armed with a SATA Blu-ray I tested later). From this, I surmised that the performance of the DVD players/writers had more to do with the caching speed of the movie than the speed or number of CPU cores.

Why does all this matter? DVD caching (commonly called ripping) takes only a fraction of the time that’s needed for compressing and encoding the video, but it still adds several minutes to the process. Ideally, you want the best outcome on the best possible hardware. If you have 100 DVDs to back up (caching is used in backing up DVDs as well), and one DVD drive takes 10 minutes longer, that process is suddenly 1000 minutes (16.67 hours) longer.

Logically, you would think the latest technology would give you the best results. But what about across brands?

To compare DVD ROMs and writers, I used a selection of DVD drives from different years, using different interfaces, and from different manufacturers. I used Acidrip to cache the DVD. Each time I cached a DVD, I made sure to use a new directory for the cache (deleting any old directories to ensure Acidrip wasn’t borrowing from any other existing cache).

The results are shown below.

In this case, the newest technology didn’t necessarily mean the fastest. Although I didn’t test a Blu-ray player for this experiment, I did test one prior to this experiment using the same DVD I used in the experiment in issue 61, and (on the 8 core) it cached better than most, but still a couple of minutes slower than the older DVD writers in the dual core systems, just over 6 minutes.

Interface also doesn’t seem to make a big difference. Though I tested only 1 SATA DVD drive (and the SATA Blu-ray I mentioned), a couple of PATA DVD drives beat the SATAs.

Both the slow Hitachi-LG and Sony DVD drives had small buffer sizes. You could surmise that the buffer size has something to do with the speed of the caching, except for the fact that the fastest AOpen DVD drive had a smaller buffer than most of the other drives.

Made in 2001, the AOpen DVD drive is also one of the oldest drives, but it posted the best caching time, so age isn’t necessarily a factor (though wear and tear could be).

What about the Maximum DVD read speed? The Sony DVD drive was the slowest in the group and the second slowest performing drive (not to mention the oldest), but it still beat a drive eight years newer and with the fastest Maximum DVD read speed.

One rumour I heard was that certain manufacturers purposefully cripple their drives to make it more difficult to copy DVDs. Looking at the statistics, the worst performers are Hitachi-LG drives, and the Sony drive. While Sony is a well known supporter of Digital Rights Management (DRM), and LG drives were among the slowest in the test conducted in FCM#61, I’m skeptical this is the reason why these particular drives performed the worst.

Doing more digging, I discovered that the maximum DVD read speeds published by manufacturers are not necessarily the maximum DVD video read speeds. This is where it gets tricky; it’s difficult to find accurate DVD video read speeds for drives.

I found that the LG DC-4522B, rated for 16x, actually transfers video data at a maximum of 4.8x. The maximum DVD read speed is also part of a group of statistics known as Data Transfer Rate (DTR) which includes burning as well. Drive manufacturers either no longer have – or have never – published DTR video statistics for the other drives in this experiment.

At this point, I went looking to other sources and found www.cdrinfo.com – a site (with banner ads) that publishes a variety of information about optical media technology. The site is Windows-oriented, relying on proprietary programs to do a lot of the testing, but their numbers seem to jive with the data I gathered. The AOpen DVD1648 was among the fastest of the DVD drives they tested copying DVDs using CSS. While cdrinfo.com didn’t have the same LG drives tested here, the LG 16x drives I looked at were among the slowest drives reading CSS encrypted DVDs.

So what does all this mean? If you’re interested in archiving your DVDs or Blu-rays, try to get as much information about the DVD video read speeds beyond the simple maximum read and burn speeds published on the websites of most manufacturers. A speedy DVD or Blu-ray drive could save you hours of work.

Important Links

http://www.cdrinfo.com/ http://www.btc.com.tw/ http://www.lg.com/ http://global.m.aopen.com/