DVD caching experiment

[[{"type":"media","view_mode":"media_large","fid":null,"attributes":{"class":"media-image size-medium wp-image-1263","typeof":"foaf:Image","style":"","width":"286","height":"300","alt":"Acidrip Screen 1 from DVD cache test"}}]] Acidrip Screen 1 from DVD cache test

Back in May 2012 I wrote an article about DVD ripping and encoding for Full Circle Magazine Issue #61. One of the more surprising results was when a couple of our dual core computers beat out our two four core (and an 8 core computer I tested later) in the ripping portion of the test. This led me to believe that certain DVD drives performed much better than others.

Every month I stay a few hours later at work to host the Kitchener/Waterloo regional Drupal content management user group, so I decided use the time between when Computer Recycling was closed and the Drupal meeting started to pit a number of refurbished DVD drives we had against one another in a DVD caching "death match."

To keep the experiment fair I used the same computer and software throughout the entire process. For the caching/ripping I used Acidrip on Xubuntu 12.04. I also used the exact same DVD for each experiment. The one change I made each time was the location of the cache file. To ensure Acidrip wasn't reading from some pre-existing cache I created a different folder on the desktop to each DVD drive's cache. When the experiment finished I deleted both the folder holding the cache files as well as the avi Acidrip started to create. I then emptied the trash to make sure the cache and the avi partial file were gone. Lastly I kept almost all the same settings, with only the cache file location difference, for each DVD drive tested.

I've taken screenshots of the Acidrip General, Video, and Settings screens so you can repeat the experiment if you wish. The DVD I used is the 1987 Louis L'Amour's The Quick And The Dead (the Sam Elliot version, not the Sharon Stone version).

[[{"type":"media","view_mode":"media_large","fid":null,"attributes":{"class":"media-image size-medium wp-image-1264","typeof":"foaf:Image","style":"","width":"287","height":"300","alt":"Acidrip Video settings screen for the DVD caching experiment"}}]] Acidrip Video settings screen for the DVD caching experiment

On the video screen I unchecked the Crop and the Scale options. I also set the bit rate and clicked the Lock Bits/Px then raise the bit rate until the Px was 0.150 (out of the red zone/poor quality zone). This meant the bit rate was locked at 1549 for all DVDs tested. I kept the codec at the default lavc setting and the Passes 1.

For all of the tests I used the stopwatch program to keep track of the time from when I started to when I finished. It's not exactly a perfect experiment because there's a slight delay between when I click start on Acidrip and when I click start on the stopwatch, but it's really about 1 second and most of the results vary much more. What I was looking for was significant differences in time  between the various DVD-ROM drives and DVD Writers.

Now you might think why go through all this trouble for a few minutes? I explained my main reason back in the article for Full Circle Magazine; if you're ripping one DVD you don't really care about a few minutes, but if you have to rip a collection of a few hundred DVDs, those few minutes can save you a lot of time (200 DVDs times a 3 minute difference is 600 minutes overall, or 10 hours). And here's the best part, the difference between the fastest and slowest so far is more than 10 minutes (200 DVDs times a 10 minute difference is 2000 minutes or 33.33333 hours). Over a day more of constant ripping which doesn't even taking into account the amount of time it takes to encode the video [which is not DVD drive dependent, but takes much longer].

[[{"type":"media","view_mode":"media_large","fid":null,"attributes":{"class":"media-image size-medium wp-image-1265","typeof":"foaf:Image","style":"","width":"287","height":"300","alt":"Acidrip Settings Screen from the DVD caching experiment"}}]] Acidrip Settings Screen from the DVD caching experiment

On the Settings screen I changed the location of the Cache path each time (in the screenshot it's /home/linuxuser/Desktop/TOP) I put a new DVD drive into the computer. I unchecked Delete cache and checked Eject DVD. I should note that I clicked stop on the stopwatch when the video stopped caching, not at the point the DVD drive opened (a second later in each case).

Another reason why I wanted to undertake this experiment was to see if the rumor that certain manufacturers limit the speed of DVD ripping on their drives was true or not. I didn't have a lot of newer drives, but I did include a couple of Serial-ATA (SATA) drives among the Parallel-ATA (PATA) drives. Whether a drive was SATA or PATA didn't seem to matter as much as the drive itself.

Age also didn't seem to be a factor, in fact the overall winner was one of the oldest drives (then again the second slowest was the oldest).

I tested 8 different drives. One of the newer drives in the test was bad and refused to do much of anything (we see a fair bit of faulty equipment in Computer Recycling, but you would think it would be older equipment).

You won't be able to go out to your local Futureshop or Bestbuy to pick up any of these drives, but you can probably find them in any computer refurbishing place or surplus stores selling computer equipment. I hope this also encourages you to drop by a computer refurbisher (old is not always bad technology). Here are the results:

DVD Ripping Times (3.9GB movie)

DVD Mfg & Model
Mfg Date
Minutes to Cache (less is better)

AOpen DVD1648
November 2001

Hitachi-LG DC-4522B
April 2005

Hitachi-LG GDR-H30N
July 2007

February 2007

Lite-ON SHW-1635S
September 2005

Sony DDU220E
February 1999

September 2001

Toshiba Samsung TS-H493
January 2008

As you can see from the results the fastest drive, the AOpen, was 1 second faster than the Lite-On drive. These results can't be taken as a ringing endorsement AOpen or Lite-On, but compared to the slowest 3 drives they're more than twice as fast (more than 3 x faster than the slowest Hitachi-LG drive). When I wrote my article for Full Circle Magazine I also noticed that LG drives seemed to be slower than other drives. These results appear to confirm that companies like LG and Sony degrade the ripping/caching on their DVD drives. I only had one Sony drive to test, but we already know Sony is a big fan of crippling technologies like DRM, so it isn't surprising that the only Sony drive I tested was the second slowest. Sony can't really blame the 1999 age of the drive because both 2001 drives in our test were more than double the speed (LG and Hitachi have even less of an excuse).

One could argue that because the drives are used and used by different people the results could be significantly skewed. However I've tested DVD ripping on several more LG DVD drives over the past few months and they're all within the 10-14 minute mark.