[davi]

I went into an Apple store today and used iTunes to import the same 77 minute CD into 128kb/s AAC on a 1.25GHz Mac mini and a 20" iMac G5 1.8GHz. Both Macs had stock 256MB RAM, no other apps running, processor performance set to Highest, and drives that read "CDs at up to 24x speed" according to the published specs.

The mini took around 6 minutes to rip the 16 tracks on the CD, averaging around 14x encoding from beginning to end. The iMac started at around 7x, then slowly and steadily sped up with each track to end at around 20x, finishing in a little over 5 minutes. I ripped the CD again on the iMac just to double-check and got the same results.

Any ideas why the iMac didn't rip dramatically faster than the mini? Is the low RAM a bottleneck regardless of processor type and speed? And why did the mini rip at a consistent rate all the way through, while the iMac started slow and progressively sped up?

[yikes600]

CD encoding always speeds up towards the end, as optical drives are able to read the audio tracks on the outer part of the CD much faster. The encoding speed is bottlenecked by the drive speed at first. You'll see a bigger difference with a pure CPU test... first import the disc as AIFF (uncompressed format), and time how long it takes to convert them to AAC or MP3.
I can't explain why the Mini's optical drive would read the disc faster towards beginning. Perhaps it has something to do with the iMac's drive being positioned vertically.

[omar96]

My guess is that the mini and hte iMac are using different optical drives, and I'm further guessing that if hte mini never wavered from ~14x that it's being bottlenecked by the CPU, where the iMac is probably being bottlenecked by how fast the CD drive can read the disc. With a superdrive it might be more dramatic of a difference.

[Simon]

Originally posted by yikes600:
The encoding speed is bottlenecked by the drive speed at first. You'll see a bigger difference with a pure CPU test... first import the disc as AIFF (uncompressed format), and time how long it takes to convert them to AAC or MP3.

^^^ Very good advice.

davi, go back to the store and do what he suggested. Then report back. I'm pretty sure the iMac will perform much better.

[davi]

Thanks for your advice folks. Looks like yikes600 was right. I returned to the store and repeated the test as suggested, first ripping the CD to AIFF, then importing the AIFF tracks to 128kb/s AAC, all with iTunes. Again, same hardware: 1.8 iMac with SuperDrive and 1.25 Mac mini with Combo drive.

This time both Macs started the AIFF rip slower and sped up progressively. The iMac took 5.5 minutes to go from 8x to 20x over the course of the CD, and the mini took 4.5 minutes from 14x to 24x.

Converting to the AIFFs to AAC took 4 minutes on the iMac, consistently averaging around 20x, and 5.5 minutes on the mini, averaging 14x.

This suggests that the mini's Combo reads CDs faster than the iMac's SuperDrive, even though they are both specified to read CDs at 24x, but AAC ripping is bottlenecked by the mini's slower processor. On the iMac the SuperDrive reads CDs slower, but the faster processor makes shorter work of the AAC conversion.

[Eriamjh]

Originally posted by davi:
This suggests that the mini's Combo reads CDs faster than the iMac's SuperDrive, even though they are both specified to read CDs at 24x, but AAC ripping is bottlenecked by the mini's slower processor. On the iMac the SuperDrive reads CDs slower, but the faster processor makes shorter work of the AAC conversion.

I think you hit the nail on the head here.

I have seen certain drives have problems reading different media. Superdrives don't seem to read CDs as well as combo drives or even equivalent rated CD-ROM drives.

While the drives might be capable to reading at the same speed, when they have trouble reading, they might throttle down until they get consistent data. This may render a faster drive less effective that its seemingly slower brethren.

[bobx2001]

they are come interesting results, it is a shame that the imac superdrive is slower than the mac mini combo drive. could you try repeating the test with a imac with a combo drive?