Well, I use mencoder in Linux and in OSX so I can compare; on the same dvd on an iMac G5 1.8gHz I get a mere 10 fps encoding from dvd.
Exactly the same command line on a Toshiba Satellite P4 2.4 gHz gives me 35 fps.... uhm, and I believed the bs about the ghz of the G5 being worth more! In fact I can confirm (also checking with md5sum's, maple etc..) that a single G5 at 1.8 ghz is about half or a third the speed of a P4 2.4ghz. Sorry for the bad news!
A good news is that for anything else (OS related) my iMac is far superior to a PC-based OS. Ah, if only OSX would be released for PC's!!!
Also note (a bug?) encoding from dvd is MUCH slower (about half speed) than encoding from hd (I mean I first copy on hd the content of the ts folder and then encode). I can't say if it the dvd's fault or mencoder, or OSX.
Mencoder OSX
Direct DVD to DiVX, XVID ripper
Version: 0.7.7
Pathetically slow on dvd
Feedback Type: Commentary
Contributed by: Linuxophile Monday, December 13 2004 @ 06:51 PM PST
Product Platform: MacOSX
Used Product For: Less than a month
Recommend Product: YES
Comments
Doesn't say ANYTHING about Mac speed - logicnazi
I realize you were probably just *censored*ing about slow encoding but I wanted to make sure no one was confused by your comment and in any way thinks this is a fair comparison of PowerPC and x86 processors. The differences in the processors are quite subtle and complicated and couldn't be fully revealed with one piece of software even if it was a good choice for comparison. For instance while x86 chips tend to be quite strong on integer operations the PPC chips tend to be very strong on floating point calculations and SIMD. Thus making them particularly suited for multimedia and scientific (though not in a price/performance way) applications.Moreover, this is a horrible program to compare processor speeds. Ideally you want a program which is entierly processor bound, i.e., the processor is never waiting for incoming data, so you can compare the processors instead of other hardware or drivers. As your comment about encoding being twice as fast from disk illustrates this program is clearly spending a great deal of time waiting for data. Thus your observed difference in speed is as likely to be caused by a difference in DVD readers or even software drivers as it is to a difference in processors. Also if this program isn't well optimized it may be making inefficent disk reads and thus depend more on the disk cache settings than anything else. Finally one shouldn't forget that all sorts of compiler differences, optimization choices, and the use of native vs. non-native libraries can all affect performance as well.
There are tons of benchmarks out there that will give detailed accurate comparisons of PPC vs. x86 performance broken down by type of task. I'd look at these rather than a DVD encoding program working from a DVD drive.
Tuesday, September 13 2005 @ 08:41 AM PDT
Doesn't say ANYTHING about Mac speed - Linuxophile
Maybe G5 is such a powerful processor but anyway my wishes apparently will be soon granted since Apple is migrating all the platforms to Intel!And out there already people are alpha testing osx on X86!
Anyway I realize that I must have touched a sensitive issue:
just keep it cool: what makes an Apple an Apple is not the G5, is the OS!
(and the design too)
Sunday, November 20 2005 @ 08:42 PM PST
Doesn't say ANYTHING about Mac speed - Linuxophile
Maybe G5 is such a powerful processor but anyway my wishes apparently will be soon granted since Apple is migrating all the platforms to Intel!And out there already people are alpha testing osx on X86!
Anyway I realize that I must have touched a sensitive issue:
just keep it cool: what makes an Apple an Apple is not the G5, is the OS!
(and the design too)
Sunday, November 20 2005 @ 08:43 PM PST
mac's vs Pc's - jenorth
I'm still running my Mac g4 with dual 533mhz processors.My son has an HP, and a Gateway that are 2 gig and up.
I hate waiting for the PC's to do anything.
I take my Mac anyday.
Friday, June 08 2007 @ 07:51 PM PDT
Pathetically slow on dvd - Pac Mac
Out of curiosity, what is your performance setting in the the Energy Saver preference panel? If it is set to "Automatic", try setting it to "Highest" and retry your test. I found that in benchmarking things the iMac/1.8 was about 30-40% slower when set to Automatic than when set to Highest. It might help.Reply to This
Sunday, May 08 2005 @ 09:39 AM PDT