EAC workflow

September 25, 2010 10:12 AM

Seeking best workflow for ripping CDs to FLAC using Exact Audio Copy

I have about 500 CDs to rip. And I'm planning to sell or give the CDs away in the near future. By the time I listen to most of my flac files, the CDs will be gone. So I need a process that gives me high confidence that I have accurate flacs, but hopefully without adding too much time to the ripping process.

To the extent that there's a trade-off between quality and time, my priority is on quality. I just don't want to waste time on steps that add little or nothing of value.

My main questions are:

1. If I rip in "secure" mode, should I use EACs "test and copy" (shift-f6) feature? Or is "copy" (shift-f5) sufficient? I've seen some comments to the effect that "test and copy" is overkill in "secure" mode, but can't tell if that's the common wisdom.

2. Am I better off ripping in "burst" mode and using "test and copy," rather than ripping in "secure" mode?

Thanks in advance.

posted by anonymice to Computers & Internet (9 answers total)

Best answer: I have a largish collection (~1500 CDs) that I have ripped to FLAC, and I prefer burst mode with test and copy. As long you know your drive is good (and configured properly), and your checksums for the test and copy rips match, you can be quite confident that you've got a good rip.

Depending on the condition of the disc, sometimes secure mode makes sense, but the vast majority of my collection was purchased new and is in great shape. Secure mode takes longer, and seemed (to me) unnecessary for an undamaged CD.
posted by lefty lucky cat at 11:26 AM on September 25, 2010

This should tell you everything that you need to know.
posted by niles at 11:33 AM on September 25, 2010

This may not be entirely germane, but selling the cd's after you rip them is kind of frowned upon, depending on who you talk to.

Copying them for your own use seems OK, but selling the CD's after copying them and retaining the copies seems to cross a line for many professional musicians. (Let's just leave the RIAA out of it, they're nuts.)
posted by Aquaman at 12:25 PM on September 25, 2010

Response by poster: Aquaman: good point. That actually never occurred to me. My main goal here is just to free up space, so I'll forego the cash and keep my conscience clean.

In a way, I guess even giving them away is ethically problematic -- costing the artist and label a possible sale. Like, if I made a habit of buying CDs I like, ripping them, and then re-gifting them to people with similar tastes, it'd be the same two-for-the-price-of-one problem. But I can't just throw them out.

Modern life is unnecessarily complicated.
posted by anonymice at 12:59 PM on September 25, 2010

Here's another good EAC/FLAC guide.
posted by box at 1:55 PM on September 25, 2010

Best answer: Ripping audio CDs is a tricky process which may seem overwhelming at first but once you understand enough to set up a proper ripping procedure things will go surprisingly smoothly.
Hydrogenaudio is indeed an excellent resource, here's the EAC section.


1. If you're NOT using C2 pointers, secure test & copy is overkill because EAC will already read each sector at least twice.

2. Yes, because it would be a lot faster. In fact, a single burst pass is sufficient as long as all your tracks are verified by AccurateRip.


This should tell you everything that you need to know.

I would avoid that guide as it apparently contains misinformation. Try this one or the Hydrogenaudio wiki instead.
posted by Bangaioh at 2:38 PM on September 25, 2010 [1 favorite]

Best answer: This is the workflow I'm using as I make my way through my multi-thousand-disc collection.

• I use the REACT 2 add-on for EAC. It lets you get multiple output formats in a single pass -- in my case, FLAC images with embedded cuesheets for archival use, and MP3s for portable use. This is a huge timesaver, but it helps to have some Windows scripting experience to get it set up "just right".
• I rip everything in burst mode on the first pass. If that rip comes back with an "All tracks accurately ripped" from AccurateRip, I consider that disc successfully completed and move on to the next disc; otherwise I set the disc and the rip aside for further work later on. 85-90% of my rips are successfully completed in this single pass.
• For rips where AccurateRip returns "Some tracks could not be verified as accurate", or which aren't found in the AccurateRip database, I do a second pass in burst mode on a different model drive and compare the results. If I get identical AccurateRip checksums on both rips, I consider that disc successfully ripped -- having a pressing that's not in the AccurateRip DB is far more likely than two drives with completely different internals returning the exact same checksums from a bad rip. This takes care of 70-75% of the discs that require a second pass.
• If I haven't gotten a good rip by this point, this is when I start using secure mode, slowing the rip down, using other drives, etc. I only have a handful of discs I haven't been able to get a clean rip from, and they're almost always discs you wouldn't suspect were faulty just from looking at them.

I also usually run a couple of extra scripts I've written (these would be overkill for most people):

• A wraparound for cdrdao that re-reads the disc and generates a cuesheet with correct subcode information. This works around the fact that EAC has problems with the ISRC field and can't check for pre-emphasis outside of the TOC anymore. (I own discs with pre-emphasis that were pressed as late as the mid-'90s, so this really does matter.)
• A wraparound for SoX that I run against the FLAC to pull RMS values and more accurate peaks. This is very useful for distinguishing different masterings.

tl;dr: Thanks to AccurateRip, anything more than burst mode is a waste of time on most discs. Save the heavy artillery for the CDs that really require it.
posted by Lazlo at 4:25 PM on September 25, 2010 [3 favorites]

For rips where AccurateRip returns "Some tracks could not be verified as accurate", or which aren't found in the AccurateRip database, I do a second pass in burst mode on a different model drive and compare the results.

One important thing I forgot to add: EAC's current AccurateRip implementation can't check across different pressings of the same CD, so you may get a "no tracks verified as accurate" message every now and then. It's useful to have an app around that can verify these "inaccurate" rips against other pressings possibly present in the AR database: CUETools and/or foobar2000 with the File Integrity Verifier component (set to verbose output) do the job nicely.

Note that when verifying a rip with CUETools you have the option of submitting error correction data to the CUETools Database in case your rip was verified by AR (or of using previously submitted data to correct your rip if it is erroneous). Beware of edge cases, though.
posted by Bangaioh at 5:26 PM on September 25, 2010

I would avoid that guide as it apparently contains misinformation. Try this one or the Hydrogenaudio wiki instead.

Nice to know. Time to upgrade my how-to links. Thanks.
posted by niles at 9:15 PM on September 25, 2010