hello,

is there any danger in ripping two CDs at once?

i want to convert my CD collection to MP3s/APE (about 100-150 CDs) and being able to rip 2 simultaneously would be a big help. i tested it last night (using CDex - LAME 3.92, VBR alt-preset standard - link) and it seemed to work fine. however i don't want to have to test every MP3 after its burned. I'm trying to establish a cost benefit analyses. if i save time ripping 2 at once, am i going to just waste it at the other end by needing to test every MP3 (and reburning when there is an error) or is it pretty risk free?

finally, will it matter if i'm using CDex to make decent quality MP3s or ripping lossless files with EAC?

if system specs matter, my current system is:
a new Dell Dimension XPS 3.2GHz P4, 2GBs RAM, a 74GB 10K rpm SATA hard drive, an 8X DVD+R/RW CD-RW Combination Drive and a 16X DVD-Rom Drive.

thanks

System specs do matter for this sort of question, but yours are more than adequate to the task, and assuming you are logical about the number of other processes simultaneously running, that all your RAM is good, and that you don't have any dying sectors on your hard-drive, the process should be quite safe. (Unless, of course, you share my own paranoia about the quality of your archive! In that case, you'd find yourself carefully listening to every encode anyway, and wouldn't save any time whatsoever. ) But remember, that every program running in the background is using some amount of resources, whether it's Notepad, Mozilla or Quake.

- M.

Hmm... no. They do not. Apps that are not being used, does not use any CPU time. And assuming your page file is big enough, they will be paged out, so it will not affect your available physical RAM either.

I stand corrected; my post should have read "any program actively doing something in the background..." (and I should have left out Notepad under those conditions, since it's difficult for Notepad to do anything "in the background"). Silly me!

- M.

Searching a word in a giant text file

I just tested LAME on a very, very similar PC. I have not tested the MP3 quality -- but I trust LAME did what I said, so I'm going to assume all worked out great -- but I did measure the speed.

I ran two separate copies of LAME on a test suite of 169 already-ripped WAVs.

The results?

With one copy of RazorLAME, --alt-preset standard
--> 80 minutes

With two copies of RazorLAME, --alt-preset standard
--> 54 minutes

I'd call that more than a little significant, wouldn't you?

Heck, I've used 3 DVD drives at once to rip CDs. (To .ape, though)

My biggest problem was keeping up with the dang drives, having to correct/fill out track names and genres and whatnot for 3 discs at once. (Especially with singles)

And that was on a P3 866. So, no worries, yo.

However, I did use EAC for that.

Is it better to use DVD drives to rip audio cds?!?! Because I have a Samsung SD-612 DVD-Rom and an ASUS CRW-5224/A here and I use the asus as it seems to be the best. I had an ASUS CD-S520 but I killed it (plugging drives when computer is powered on *is* a bad idea).

I just used the DVD drives since I wanted the future-proofness. In my experience, DVD drives tend to have smaller and more consistent offsets (EAC thing, track offsets vary by drive) than CD drives.

Ripping multiple cd's at once is fine. Encoding is different. It's not worth it. I'd recommend doing all the ripping and then just running lame overnight or something.

DVD-Rom's can be better, depends on the drive. Usually DVD-rom's don't cache audio.

I don't personally have a P4, but the P4 3.2ghz should have HyperThreading (I believe that all p4's over 2.8ghz do) and that should make an encoding difference. I know that is a whole lot of "should's", but take for example my dual athlon 2000+ machine:

1 Lame (--alt-preset standard -q 0 -V 0 -Y --replaygain-accurate) ~=5-7mins
2 Lame Concurrently (Same settings) ~=6-9mins

Now it's true HT doesn't have the same resources as full blow dual processors, but as there is some definite I/O dependency coupled with the p4's overpipelined architecture, gains should not only be possible but probable. Furthermore I would expect a top of the line computer like this to have dual channel ram and an 800mhz bus to better feed data to the processor. The 8k L1 D-cache may inhibit your encoding gains somewhat but it is hard to say without some more in-depth study.

So, imho as long as you are able to keep up with it on your end (or make a really cool batch file), you computer should encode faster (total wall time start to finish) running 2 Lame executions concurrently. As for the ripping....errr not my area.

Running two shots of LAME via EAC, or RazorLame may work as intended although I haven't personally tried it. Just a short while ago when it was talked about on here about making a Hyper-Threading compile of LAME and the end result being "just run two copies of LAME" I tried it with ALL2LAME. My result had me wondering what the heck was happening as one LAME window was encoding the same stuff the other one had just encoded a few seconds earlier.

I realize now though that it must have been go.bat that ALL2LAME creates for LAME to use for encoding which caused the problem, but nonetheless it was strange.

I tried ripping two CD's on my 3.20Ghz P4 with HT the other day, I came to the conclusion that I'll stick to ripping one CD at a time since I'm paranoid about the ripping quality versus the speed. I'm too paranoid to even browse the web while ripping is taking place because when EAC gets into some serious error correction set at 'Normal extraction and compression priority level' I've seen even my 3.20Ghz P4 with HT stumble a little until error correction is done.

Hmm...

I do all kinds of stuff while ripping... I even play games some times, and I've yet to see (or hear) a problem. I don't think it's an issue.

thanks for the help everyone!

i was already going to create a two tiered archive. on the low end, decent quality MP3s, on the high end, quality APEs. based on your advice and my personal quirks, music i really care about and that seem like they'll stand the test of time i'll rip one at a time using EAC and MPC. music thats a bit marginal, or only worth keeping for campy nostalgia i'll rip concurrently using CDex, and LAME (alt-preset standard).

if anyone thinks this is a bad approach for some reason, please let me know. otherwise,
thanks again.