EAC & dBpoweramp Users – Which do you use/prefer & why?

Of course. Just have EAC generate a cue sheet as multiple wav files to simulate the track based method of ripping. You may have to manually modify the file names if you ever use the cue sheet. Also, CUETools can be used to quickly correct file names and generate a single image file for burning.

Current versions of EAC and dBpa can both detect pre-emphasis if it is in the TOC. If it is indicated only in the sub-code, neither will detect it. You have to go back to EAC v.095pb3 for this to work. This can be found on the web with a search and it can also be installed on a computer along with the current version of EAC - just remember to install to a different path.

 

Why didn't you just put your answer in the body of your post, instead of making me have to scroll back up to read the title?

 

Thanks Scott! Just so I am clear, when you say "archiving your CDs" & mention EAC's CUE sheet capability, you mean making CD-R copies, etc.?

What I want to do is basically digitize my collection. I don't have a music server set up at this time. What we do have are several iPods in the family - and we have no problems using MP3s on those. My thought was to rip to FLAC, convert them to MP3s and store both. That way I'll have a lossless library when I need it & an MP3 library for the iPods.

So there are no conflicts with using EAC simply to generate a CUE sheet but rip in dBpoweramp (if I were to make that switch).

It sounds like you have multiple versions of EAC just to use to check for pre-emphasis.

I appreciate this help!
Dexter

 

My next question has to do with the differences between EAC's approach to ripping compared to dBpoweramp's approach. Both make use of Accurate Rip, but from what I understand EAC does things a little differently. If someone could explain the differences in a simple way, I would appreciate it!

The only version of dBpoweramp that I have right now is the "Power Pak" version which I don't think is offered anymore. If I'm correct, my version only rips in "burst" mode.

I did a series of tests using EAC 1.0 Beta 3 - changing the FLAC compression levels. What I found surprised me - that is, unless I did something wrong.

I chose the 1987 version of Sgt. Pepper for the test - thinking that it should have plenty of records in Accurate Rip. I timed each rip with a stop-watch.

At -8, the rip took 20:46 & the folder size was 264 MB.
At -5, the rip took 20:44 & the folder size was 265 MB.
At -0, the rip took 20:46 & the folder size was 289 MB.

The difference in the amount of time it took was negligible. So what's the big deal here (especially between -8 and -5) ... or did I do something wrong?

Also, in all 3 cases, the Accurate Rip confidence in EAC was "6". Does this mean that EAC is only finding 6 matches in Accurate Rip to my rip?

In identical tests I did in dbPoweramp (in "burst" mode because of my s/w version), naturally the time was MUCH faster - taking an average of 1:54. The folders were identical in size to EAC. But the number in the Rip Status column was in the upper 80s to low 90s for each track!

Why would EAC's result be so low - especially for this album? Might it be that EAC's Accurate Rip results are "reset" with every new version of the software? Or am I mis-interpreting the result?

If I were to upgrade dBpoweramp, *approximately* how long to you think it would take to rip the same album in a secure mode?

Lots of questions I know, but I do appreciate the help!

Dexter

 

For those titles where pre-emphasis is detected, does dBpa document it for you? Whether it be in the tags or the log for confirmation and reference later on? I don't recall ever encountering any info like that, but it could be I haven't ripped a pre-emphasis disc since starting to use dBpa.

 

There is no big deal here and you did nothing wrong. Here we see small resulting size differences, with -8 being the smallest and the files getting larger as you get down to -0 setting. I think most recommend going with the highest compression setting,-8, because it is the most space-saving. That's what I use. But as you can see from your example results, the differences are small enough that you can't go wrong with any of those settings. They are all lossless no matter which setting you choose. There may be some small differences in how long it takes to rip a disc from one setting to the next, but it's too minor for me to care about. We're probably talking a few second here or there, at most.

Part of the inconsistency of your above -8 vs -5 vs -0 timeline could merely be due to some other processes happening in your computer which have influenced (slowed down) one rip to the next. Something other activity tying up your memory or cpu a little bit, pulling resources away from your rip. No big deal.

Yes, only 6 matches. But one is even enough.

The Accurate Rip database gets updated periodically. Not sure how often that is, however. I'm sure someone else would know. I'm also forgetting if you are allowed to "opt out" of having your ripping results added in to the database. Other than a disc simply not being that popular, this ability to 'opt-out' of providing your own ripping results might explain a low number of matches.

I do find it curious myself sometimes. Discs which I would suspect would have had a ton of prior matching examples only have a few and then others which aren't that popular sometimes seem to have a ton of Accurate Rip matches. very odd, at times.

 

You have to manually add a "Track Technical" column to the dBpoweramp Ripper to see if the disc has pre-emphasis or HDCD.

Put a CD in. You'll see columns of info with "Title" "Length" "Rip Status" etc. Right-click up in the column heading area and it will let you add a column for "Track Technical".

dBpoweramp doesn't have a DSP to do de-emphasis. All it can do is tell you that pre-emphasis was detected. If you go ahead and rip anyways you'll have a rip that will sound too bright. I don't believe dBpoweramp adds any info to the log that de-emphasis was detected. So the only indication that there is pre-emphasis is in the "Track Technica" column that you'll see during the rip. Bummer. Means that if you're doing a bulk ripping session you have to watch the "Track Technical" column if you're ripping stuff you suspect may have pre-emphasis (like older classical CDs). Bummer. I really wish dBpoweramp handled that better and could do de-emphasis as one of its DSP actions.

 

Ok, thanks Ham. I've added that column. Even just knowing whether the pre-emphasis is there or not is of some help.

 

It's not clear from your post what exactly you think is wrong with your results. Taking the possibilities one by one:

• The FLAC compression levels won't have any effect on how long the rip takes, because FLAC compression doesn't even start until after the rip is completed -- it's a separate process, and the time spent on it isn't counted in the time reported by EAC. So you would expect the rip times to be similar regardless of the compression level, as long as all the rips were done in the same mode.
• As with any lossless compression scheme, you hit diminishing returns pretty quickly on higher FLAC compression levels. So it's not a surprise that a -8 file isn't that much smaller than a -5 file.
• And your rips are probably taking 20+ minutes because you're not using burst mode. There is absolutely no benefit in using anything but burst mode by default in an AccurateRip-enabled ripper -- all you're doing is wasting time and putting unnecessary wear and tear on your drive. If you can't get an AccurateRip match with the burst-mode rip, than yeah, go ahead and try a secure-mode rip. But a solid AccurateRip match is a better sign of a good rip than a successful rip in secure mode.

EAC's AccurateRip results require you to have set up the offset on your drive correctly (the Offset/Speed tab in Drive Options). This compensates for how different drives will return audio data starting at slightly different sample positions on the disc. In most cases this offset read position doesn't cause you to lose any actual audio content, because on most titles all you lose is zero-value silent samples, but it does result in different checksums when you validate the data file. So if your drive has a read offset but you haven't set up EAC to compensate for it, you'll only get matches back from the AccurateRip database for rips that have the same uncompensated offset as yours. (There are also cases where an otherwise-bit-identical mastering will have a slightly different amount of zero-padding at the start of the disc, which will cause the same thing.) Not having set your offset values is the most likely reason for you to see a low confidence value on a popular title like Sgt. Pepper.

dbPowerAmp's numbers, on the other hand, are probably showing all submitted rips with bit-identical content regardless of offset values. If you run a lossless version of your rip through CUETools it'll probably show you all the different offset values reported for that mastering.

Here's an illustration of what that means in real-world terms. I ripped Jon and Vangelis' Short Stories and got this result in the EAC log:

Code:

AccurateRip summary
 
Track  1  accurately ripped (confidence 11)  [65D2D8D5]
Track  2  accurately ripped (confidence 13)  [04DFFD5C]
Track  3  accurately ripped (confidence 13)  [7EF8D00C]
Track  4  accurately ripped (confidence 12)  [13B79658]
Track  5  accurately ripped (confidence 13)  [603024B0]
Track  6  accurately ripped (confidence 12)  [B75386B8]
Track  7  accurately ripped (confidence 12)  [EE103D18]
Track  8  accurately ripped (confidence 13)  [459C347E]
 
All tracks accurately ripped

So a clean match, but not a very high number for a relatively popular title. The reason for this becomes clearer on a CUETools report:

Code:

[AccurateRip ID: 000e2b24-005deda5-5d0a9a08] found.
Track   [ CRC    ] Status
 01     [65d2d8d5] (11/84) Accurately ripped
 02     [04dffd5c] (13/87) Accurately ripped
 03     [7ef8d00c] (13/87) Accurately ripped
 04     [13b79658] (12/86) Accurately ripped
 05     [603024b0] (13/87) Accurately ripped
 06     [b75386b8] (12/86) Accurately ripped
 07     [ee103d18] (12/86) Accurately ripped
 08     [459c347e] (13/87) Accurately ripped
Offsetted by -762:
 01     [208cad94] (06/84) Accurately ripped
 02     [e83cd391] (06/87) Accurately ripped
 03     [304ebb06] (06/87) Accurately ripped
 04     [8477d9ab] (06/86) Accurately ripped
 05     [7c02dcaa] (06/87) Accurately ripped
 06     [7d805213] (06/86) Accurately ripped
 07     [0b1468ec] (06/86) Accurately ripped
 08     [8a0b9e5a] (06/87) Accurately ripped
Offsetted by -744:
 01     [3655d10f] (03/84) Accurately ripped
 02     [97ee5dc9] (03/87) Accurately ripped
 03     [aabc1d7c] (03/87) Accurately ripped
 04     [3670202f] (03/86) Accurately ripped
 05     [160c22ce] (03/87) Accurately ripped
 06     [2455d3e7] (03/86) Accurately ripped
 07     [db2eed48] (03/86) Accurately ripped
 08     [c0464cfb] (03/87) Accurately ripped
Offsetted by -80:
 01     [ccd7e650] (16/84) Accurately ripped
 02     [8ca9268b] (17/87) Accurately ripped
 03     [5bee0504] (17/87) Accurately ripped
 04     [7fa54328] (17/86) Accurately ripped
 05     [10d97858] (17/87) Accurately ripped
 06     [0e8b7563] (17/86) Accurately ripped
 07     [34e8f9d8] (17/86) Accurately ripped
 08     [25923d1d] (17/87) Accurately ripped
Offsetted by 7:
 01     [c6649e6d] (06/84) Accurately ripped
 02     [7aa1d407] (06/87) Accurately ripped
 03     [35383c55] (06/87) Accurately ripped
 04     [5a4abb05] (06/86) Accurately ripped
 05     [9a87161d] (06/87) Accurately ripped
 06     [35bc81e3] (06/86) Accurately ripped
 07     [af1d5a93] (06/86) Accurately ripped
 08     [9c8c3e42] (06/87) Accurately ripped
Offsetted by 584:
 01     [4424bc61] (34/84) Accurately ripped
 02     [4d64aeed] (34/87) Accurately ripped
 03     [0fc5bd23] (34/87) Accurately ripped
 04     [83ed518a] (34/86) Accurately ripped
 05     [51d20ff6] (34/87) Accurately ripped
 06     [37e5f504] (34/86) Accurately ripped
 07     [9f017b9b] (34/86) Accurately ripped
 08     [e69748b7] (34/87) Accurately ripped
Offsetted by 910:
 01     [1dedfea6] (00/84) No match but offset
 02     [b04415d6] (02/87) Accurately ripped
 03     [40aecaf5] (02/87) Accurately ripped
 04     [4a283f91] (02/86) Accurately ripped
 05     [583816a1] (02/87) Accurately ripped
 06     [f2ad2484] (02/86) Accurately ripped
 07     [cf69dd8c] (02/86) Accurately ripped
 08     [44402fd6] (02/87) Accurately ripped

What this means is that there are seven entries in the AccurateRip database with bit-identical audio content for this title (not counting leadin/leadout zero samples), but with varying offsets in either the mastering or in the drives that people used to perform their submitted rips. My rip matched the first of these entries, which is all that my version of EAC knows how to do. (I'm still on 0.99b5, as I rely on REACT2 for postprocessing; the 1.0 release may behave differently.)

 

I have no idea what the earlier versions and/or the free versions of dBpa are capable of and I strongly recommend that you upgrade to the latest version. It has a free and fully functional 21 trial period and that’s what you should be using for comparisons.

Pre-emphasis
In the current full version of dBpa (v 14.2), you can enable it to (not set by default):
1. See pre-emphasis detected in the CD Ripper display by enabling (right click bar) "track technical."
2. Have it write this info to a log file saved with the rip by enabling a log file under:
CD Ripper\Options\Ripping Methods\Secure Settings\Secure Extraction Log\Write To File (check box)
3. Write a tag to the audio file indicating pre-emphasis:
CD Ripper\Options\Meta Data\Options\Write ID Tags\Pre-Emphasis (and HDCD) (check box)
　
AccurateRip\Confidence levels\Secure mode
Confidence levels are the actual number of discs reported to the AccurateRip database that match yours.
There are now 2 versions of the AccurateRip database, with version 2 using a higher level of algorithm to verify the results. Version 2 is newer and consequently has lower number of CD’s (i.e., confidence level) in the database. The current versions of EAC, dBpa and CUETools use both. It would seem as though you’re using an outdated version of dBpa, which is only reporting V1 of AccurateRip.

Yes, burst mode is much faster. The length of time to rip a disc in secure mode between the 2 programs varies depending on the disc, the drive and the errors encountered. dBpa will typically rip slightly faster in secure mode because it’s relying on AccuaretRip in a different way and it won’t re-read an error as many times as EAC will, but the difference is not significant and not predicable. i.e. - don’t worry about it.

 

Ok, looks like I've already had those two boxes checked off for some time already. I take it then that I've been getting this information saved all along, within the tags, but not in the log file, as I've only added the "track technical" column today. No biggie though. As long as that info has been saved somewhere (i.e., the tags, thus far).

Great info, thanks as per usual, GreenDrazi!

 

Maybe I should start a new thread instead of piggybacking upon this one, but...

Which DSP "extras" do you guys have turned on for dBpoweramp?

Here is what I have been using. Would you strongly suggest adding any more than these?:

HDCD
Trim Silence
ID Tag Processing (not sure if having this on makes it easier to edit tags after the initial rip or not - just figured it couldn't hurt to turn it on)
Multi-CPU Force
ReplayGain (both album & song)

Still confused about these three: "trim silence", "hidden track silence removal" and "silence track deletion" - I have only "trim silence" active. Not sure if I read up and that one seemed to be the most useful or if I just randomly chose it.

 

I use dbpoweramp 99% of the time, which is why chose that. On rare occasion, I use EAC for hard-core cases.

I like dbpoweramp because it's easier to configure and use, and it does the same job as EAC. It will also do many different formats. I can add or change tags and graphics. It does HDCD, ReplayGain, and other things.

I do not bother with cue sheets and stats. I rip the CD or convert formats, and listen to the music.

You did not want to hear about anything else, but I also use Foobar 2000 for HDCD and many other things too. It's also totally free.

 

Neat. I didn't know the current version is able to write pre-emphasis detection and HDCD detection into the log and/or tags. The older version I used when doing my big ripping project didn't do that.

So with the current version you can do a bulk style ripping project ripping stacks of CDs as fast as you can. Then grep the logs or search the tags to find out which discs or tracks have pre-emphasis or HDCD. Then process the pre-emphasis or HDCD separately.

It would be nice if dBpoweramp did that as part of its default setup. Instead you have to manually add the "Track Technical" column, manually enable logging, manually add pre-emphasis tagging and HDCD tagging. Most people are not going to know to do that.

 

The "trim silence" can cause problems with gapless tracks. For example a classical album that has intentional silence and an intentional pause between movements. You don't want to trim that silence out.

Trim silence can be helpful if you listen to rock or pop and want to have consistent crossfades from song to song. If a song happens to end with a few seconds of silence then crossfades with that song won't work as you want them to (you'll be crossfading with silence rather than the end of the music). The proper way to handle that is to use a media player that can trim silence during playback. Good media player software is able to do that. If you trim the silence during the rip then the damage has been done and you can't get that silence back if you want to play it as it was on the CD.

Silent track deletion would be for albums like Dave Matthews Band "Under the Table and Dreaming" where some editions have song "#34" at track 34 with twenty-some empty tracks before it. I prefer to keep things like that as they are on the CD rather than delete the empty tracks.

Similar for CDs that have a hidden song at the end of the last track with several minutes of silence between the end of the last track and the beginning of the hidden song. You can choose to have dBpoweramp delete the silence if it detects that.

 

Thanks!! Whew!

The timing I got was done with a stopwatch from the moment it began to the moment the window popped up indicating that it was completed. From what I can tell, the FLAC compression on each track starts after each wav has been ripped - so compression tasks are taking place while the next wav is being ripped. (This computer can handle that - i7 processor 6 GM RAM - and I don't usually do other stuff on the machine while ripping.)

The offset stuff confuses me. The EAC setup guide I was using mentions that "Accurate Rip has configured my drive settings". And indeed it shows a value of +667 in the box when I get to that tab. Yet when i uncheck the Accurate Rip checkbox & click "detect read sample offset correction", it comes back with a different value (-580). To make matters more confusing is that the AccurateRip database listing shows +667 for my drive!

All of this makes me wonder if dBpoweramp might be easier to setup. <sigh>

 

Use Dbpoweramp always. The interface is more user friendly and the program offers more codecs and options. EAC is an excellent program and does what it is does well , but is some in many aspects for my usage.

 

Thanks for the explanations.

For this one, I assume you're referring to the first on the list of DSP options, "Hidden track silence removal". I think I had that activated for a while, but encountered a couple CDs where this feature didn't do anything at all with a long gap between a last song and an untitled hidden song that came after. So, I think I turned that particular DSP feature off and turned on the "trim silence" DSP in hopes that this would eliminate those kind of gaps. Details are hazy now, but with one cd, there was a gap of at least a minute until the hidden track kicked in. The "hidden track silence removal" did nothing at all for it.

 

I need to ressurect my thread a little here because I have a couple more questions - based on your excellent comments:

1) Even though I do not really make CD-Rs of my CDs, I do see the value in creating a CUE Sheet in case I would decide to do so at a later time. So, with that in mind, is there anything wrong with ripping a CD with dBpoweramp & then using EAC to create a CUE Sheet?

2) Can you rip multiple tracks as a single file in dBpoweramp - similar to the Copy Range feature in EAC? For example, "Sgt. Pepper's Lonely Hearts Club Band / Lucy In The Sky With Diamonds" or the entire Abbey Road Medley?

3) If the indexing on a CD is slightly "off" between 2 songs, is there a way to correct this when you rip in dbPoweramp?

4) Is dbPoweramp significantly easier to set up?

5) A question about offsets: In the newest version of EAC, they give you the option to install the AccurateRip feature, which naturally I did. It appears that EAC automatically gave my HL-DT-ST DVD+-RW GH50N drives a Read Sample Offset Correction of +667. This is also the value listed in the AccurateRip database. Yet when I uncheck the "Use AccurateRip with this drive" checkbox & click the "Detect read sample offset correction" button, EAC comes back with Sample Offset of -580. Now honestly, this is all "Greek" to me - I don't really understand what this all means - I was simply following a setup guide. But does this make any sense? Could someone please explain it simply? Which is "right"?

6) In secure mode, approximately how much of a difference will I notice in the amount of time it takes to rip a CD with dBpoweramp. Let's say it takes 20 minutes to do it in EAC as a benchmark.

7) It's been mentioned that EAC & dbPoweramp differ in the way they approach secure mode. How are they different?

Thank you!

 

1. Wrong -> No. Just realize that dBpa typically rips in track mode (not single image) and you’ll want to do the same for the EAC cuesheet. File names will likely not match exactly. CUETools can help with clean-up if you ever need to use the cuesheet to burn, etc.

2. You will have to try it with the Cue Sheet Image option. I don’t use this in dBpa. And I only append tracks after ripping.

3. You can try Apply Track Number Offset, but again, it’s better to correct this after ripping, IMHO.

4. Yes. But instead of asking here, why not just download the full-featured 21-day free trial.

5. Did you try to determine the value manually? I recommend sticking with the AccuaretRip number.

6. It depends on too many variables based on your hdwr and disc. Again, just try dBpa for the 21-day trial period.

7. dBbpa does not setup secure mode by default. Also, if an error is not encountered by the player, dBpa essentially relies on AccurateRip after the first pass and will move on if verified accurate. Re-read/compare and re-verify if not. EAC re-reads errors (higher number of re-reads allowed by default) to correct encountered errors and only verifies with AccurateRip after ripping.

 

I switched from EAC to dBbpa last week and I love it!

 

Well, I downloaded the reference version of dBpoweramp 14.2 yesterday evening. I'm not sure what I was expecting, but I am totally confused! There must be a better setup guide than the one on their site. There seems to be no end to the guides for EAC out there (most...all?...are woefully out of date) - but I can't seem to find a dbpa setup guide that is easy to follow & written for someone who is not as technically savvy about these things as the author or the Hydrogen Audio folks.

 

OK...first 3 questions....

On the Encoder tab, I have FLAC selected (actually, it was the default - I didn't have to change it). I am leaving the compression level at "5". (My tests with EAC showed me no reason to go with anything other than the default.) What does the checkbox labeled "After Encoding Verify Written Audio" refer to & what does it do when checked?

On the DSP tab, should I add any DSP Effects? He suggests adding ReplayGain. I have only a passing knowledge of what ReplayGain is - it sounds like an audio compression program, which I would think would be counter to what this forum is all about. Do you all use ReplayGain?

On the old files that I ripped to MP3 a few years back, I've found that most of them have a ReplayGain value on them now - I am sure that I didn't rip them this way & have no idea how it got on them. To some degree, I guess I can understand its value on an iPod or for listening in a car - but I can't understand why someone would want to put it on their FLAC files.

He also suggests adding the HDCD DSP Effect?

What kind of file naming convention do you use? Up until now, my rips have gone to Artist\Album Title\TrackNumber TrackTitle. Only when I rip a Various Artists CD do I put the artist's name in the file name...Various Artists\Album Title\TrackNumber TrackArtist TrackTitle.

More to come.....thank you for your help!

Dexter

 

I'm using it! ReplayGain isn't modifing your audio data - it is just a meta information, how the perceived loudness of the audio file is.
The use of it isn't to apply compression, but to adjust the volume of tracks or albums from different sources while played back (if your mediaplayer is able to use the information).

Martin

 

It doesn't matter. Lossless is lossless. The numbers only affect the amount of time needed to rip and compress the file. The end result is identical in terms of sound quality. The difference in file size is absolutely trivial, especially in a time when you can buy very affordable 4TB drives.

Not a clue. I don't bother. If it's an Accurate or Secure Rip, I'm 99.9% confident that it's fine. I strongly dislike EAC because I think its user interface is kludgy, and the support on the dBPowerAmp User's Forum is extremely good.

Not necessary, and you can always do it later. I'd rather leave the album as it was on the original CD, bit for bit. If I need to crank everything down later, I'll do it as a separate process. The key to me is just to get a good rip, good file names, good song titles, good artist name information, organized in the right folders. I do use ReplayGain (or similar tools) on a case-by-case basis.

No funny business for me. I do no effects at all, but I will sometimes rip mono CDs as mono -- but make sure you take the level down -3dB, because if it combines left and right to create mono, you may clip the end file. When in doubt, I'll just use one channel and leave it alone. If I decide to do any processing later (which is very rare), I do it by hand and don't rely on automated tools, except maybe for de-ticking and de-crackling from vinyl sources.

My preference:

1) one set of drives for individual artists; one set of drives for Various Artist compilations [set up just like music stores used to be]

2) individual artist file folders as: Artist Name / Album Title (that is, one overall Artist folder, then individual Albums inside that folder).

3) song files as Artist Name - Song Title. I make absolutely sure each file has metadata tags for Disc Number (1 of 2, etc.), and Track Numbers, as well as Category (Rock Vocals, Pop Instrumental, etc.), to aid in organizing playlists later on.

4) V/A compilations are all organized by Album Title, and song files done as above. Every file also gets a 600x600 JPEG of the original album art.

5) I also include a report of the final rip (with or without errors) in the album folder, and in some cases, a medium-rez (readable) PDF of the CD booklet.

On separate drives, I have duplicates of certain hit songs arranged by year and by Billboard peak chart numbers, so I can listen to (say) all the Top 100 for a given year, or my favorites from a given year, or every charted single from a given year. The file names are all changed to reflect years and peak numbers for each individual year. Different file folders are used to store hits from each set of charts: Pop, R&B, Modern Rock, and so on.

You'll find a lot of interesting discussions on Chris Connaker's Computer Audiophile website, where he has an entire document devoted to CD Ripping Strategy & Methodology. I don't agree with everything he says, but the highlights are good, particularly the need to get bit-accurate rips, to make sure all the metadata is right, and to do lots of backups. I disagree with doing rips as uncompressed AIFF or WAVs, and firmly believe that Apple Lossless is the best way to go. ALAC is the only high-res format that can be easily played on Windows, Macs, and all iOS devices (iPods, iPads, etc.), and can also hold all available metadata and image files.

Finally, I've said this before, but: be sure to back up, especially during a massive CD-ripping project. Drive costs are minimal; your time is not. If it costs $150 to back up 2000 CDs on two sets of drives, this is negligible compared to the weeks of work it would take to do them all over again.