CAS 5: CD Ripping
Last time we shared our excitement using iPhone as iTunes remote control. This time we will look at a larger topic in audiophile perspective – Import Audio CD, or Rip your CD data perfectly to your hard disk.
Every audiophile knows EAC (Exact Audio Copy) on Windows PC is rated as the best CD ripping program, but not much other similar programs, especially on the Apple Mac side. EAC is indeed a very nice program with its secure ripping features. However, EAC historical background is rooted since very early days, when all computer resources are so limited. Is there nothing else in this market can produce equally good CD rip? We will try to look this topic seriously in this article.
Some people complains iTunes ripped music sonically worse than EAC, they also have an idea that iTunes playbacks quality is inferior. I think audiophiles dislike iTunes is more based on psychological reason, general thought that it is designed for kids to playback MP3.
We need to repeat one more time, the first goal of digital audio playback is “data integrant”. The most important issue is the data transfer from CD to your hard disk, then later output data perfectly. If this chain is transparent, the only issue left is “jitter”. We have discussed jitter on last few articles, there are too many variables to determine how much jitter is created within a computer. There is almost impossible chance just by try and error to identify the cause(s) and fix it(them). Adding audiophile capacitors on motherboard will not produce higher quality music playback. For jitter issue, it can be cleaned and recovered by a good DAC, plus good dejittering unit/method ahead of it.
Audiophiles worry even more that iTunes CD imports is not an exact digital copy of their original CD. I had a search on internet and there are many audiophile forums talking the same thing. It draws my interests to perform some comparisons.
Let’s the test begins:
Mac user should know that when they load in an audio CD, there is a CD icon appears on the desktop screen. Double click on the CD icon, each song is represented by a file. This is a CDDA function built in Mac OS. You can simply drag these files to your computer, each file will becomes a non-compressed 16bit/44.1kHz AIFF file. This is the most basic CD ripping with Mac OS, isn’t that too easy?
But are these files identical to the CD source? We will find out later. But first, I tested 10 times, drag(rip) the same file from the same disc. Compares all these 10 files with a Mac application “File Compare 1.1“.
Results? All 10 files are data identical. It shows my 10 ripping files produce identical result, it’s a pretty good starting. Hold on, will there any chance that actually all ripped files contain the same error? Let’s go on.
Next, I use all different methods and hardwares to rip the same song from same disc. The combination is listed on below:
- “Mac Quad G5″ CDDA->AIFF (Just drag the file from the CD disc on desktop)
- “Mac Quad G5″ internal DVD-Rom – iTunes 8 imports -> AIFF (1st time)
- “Mac Quad G5″ internal DVD-Rom – iTunes 8 imports -> AIFF (2nd time)
- “PC QuadCore” internal Pioneer DVD – iTunes 8 imports -> AIFF (1st time)
- “PC QuadCore” internal Pioneer DVD – iTunes 8 imports -> AIFF (2nd time)
- “PC QuadCore” Plextor Premium Professional -> Extract Audio from CD -> Wav
- “Mac G4 Powerbook 1Ghz” External USB1.0 ASUS CD-Rom CDDA -> AIFF (drag file)
- “PC QuadCore” EAC -> Secure Rip -> Wav
- “PC QuadCore” Wavelab CD import -> Ultra Safe CD import -> Wav
- **Special Method**
Few methods are worth to mention: (1) EAC
EAC Exact Audio Copy – The most famous free CD ripping tools only available on Windows PC platform. The smartest thing it does is to re-read the data and compare each reading until they are exactly the same. And even if a badly scratch disc that cannot recover the data, it will produce a log sheet telling where the problems are. This method is really important in the early days when PC CPU speed is around 166Mhz, 128Mb Ram. The repeat reading and checking data give strong confidence for people that the rip is perfectly identical to the source.
Plextools – Plextor Premium: Plextor CD-Rs/DVD-Rs were the golden standard in computer CD/DVD production industry. The Plextools provides many CD testing features, which only can be found in high cost testing equipment like Eclipse CD Mastering software. It is sad that Plextor left the market and no longer making its own CD/DVD drives. You can still buy Plextor Premium 2 in Japan. The only other CD-R that comparable is StartLab SCSI Sonic Studio CD.1
Wavelab (PC) – Wavelab is a professional CD/DVD-A mastering program, which widely used by CD Mastering Studios such as Bernie Grundman, Doug Sax, Eddy Schreyer and of course in our studio. It has a CD import feature that is called “Ultra-Safe Mode”. With this feature on, Wavelab will read until number of successive reading tests giving the exact same result (before attempting to save the file to disk). By using a very well Plextor Premium drive, we read the file until it provides 8 times exact same result.
Analysis No.1:
After we ripped all non-compress 16bit/44.1kHz AIFF/WAV files, we import (1) EAC / (2) Plextor / (3) WaveLab Ultra-Safe / (4) Mac iTunes 8 CD on WaveLab. Then we align them to exact same timeline position, because each CD-Rom has different space/gap in front/end. This is also a reason you may not able to get identical reason with “File Compare 1.1″
We zoom the heading and further align each stereo ripped file.
As you can see Wavelab allow zoom in to sample accurate detail level. Now all 4 files are perfectly align on the same timeline.
We first focus on first 2 files. We invert L/R phase of one file by ToolsOne plugin. As audiophile, you may already knew some physic. When 2 wave files carries same amplitude, with opposite phase and playback together, they will cancel each other out = No sound. Now we have these 2 ripped files on exact same timeline. We invert one file L/R phase, and playback both files together. If they are data identical, there should be no output. Look at the digital out meter, there is nothing down to -144dBFS. This proves the EAC and WaveLab Ultra Safe Mode files are data identical. Both of them are perfect rip as expected.
Now playback all 4 files together, while Track 2 (EAC) and Track 4 (Mac iTunes 8 CD import) are inverted L/R phases. If any file has even one sample difference, the digital output meter will show.
4 Files play together and the result is again nothing down to -144dBFS. 4 files are data identical. This shows iTunes 8 CD import carries exact same data compare to professional program such as EAC, WaveLab and Plextools.
Well, we have more test files. There are PC iTunes 8, Mac iTunes 8 by Quad G5, Mac iTunes 8 by G4 1Ghz Powerbook, how will they all perform together?
We mentioned on the above. When 2 wave files carries same amplitude, with opposite phase and playback together, they will cancel each other out = No sound. What about they are all in phase? They double up the amplitude. As you can see the digital output meter when we are playing back all 8 tracks together without any phase invert.
Now we invert L/R phases on File 2, 4, 6, 8 to out of phase. File 1, 3, 5, 7 are remain in phase. When we playback them together, if they are all identical, they will cancel each other out completely too. What do you expect the result will be?
They are identical. No matter it is ripped by EAC, Plextools, Wavelab, or by PC iTunes, QuadCore Mac iTunes, Powerbook iTunes, USB 1.0 external Asus DVD drives, they are all bit transparent.
Do you feel surprise with the result? Will this give you a little more trust on iTunes? Have you ever think you spend too much time working on thousand of CDs collection? But the result is same as you click import CD to AIFF under iTunes?
With today PC high performances, a simply drag AIFF file direct from the CD disc can easily achieve bit transparent ripped to non-compressed AIFF audio file. However, there are few things that you should beware of.
- iTunes -> Preference -> General -> CD Import the “Use error correction when reading Audio CDs”
- Ripping CD demands computer resources. Don’t use too many multitasking like watching video clips, MSN, or run softwares at the background. They may draw resources and interrupt the ripping process.
Bonus Section:
Do you remember there is a “**Special Method**” file? This file is created by our PC iTunes real time playback CD via Weiss AFI1 firewire interface -> AES digital output to our Crookwood mastering console -> loop back to Lynx AES16 digital audio soundcard -> capture input by WaveLab software. This test checks a lot of things.
- Is iTunes8 live CD playback bit identical data?
- Is Weiss AFI1 output bit identical data?
- Is Crookwood mastering console completely transparent (when no processing is insert)?
- Is Lynx AES16 soundcard capture bit identical data?
- Is WaveLab record bit identical data?
If just one tiny section (be specific, 1 sample error), the whole signal path is broken (Remember 1 second music carries 88200 stereo samples, each with 16bit word length. A lot of data is passing through this loop), the Wavelab captured file will be different than the original source file.
The above photo shows the “loop captured” file play together with the WaveLab Ultra Safe ripped CD file. Bit transparent result again. Well, we of course very happy that our digital chain is clean.
In this CAS series, we keep on talking an important issue – digital data can be compare, clone and verify. In the file comparison tests, we prove iTunes 8 and up to date computer/CD/DVD drives, are capable to rip bit perfect data same as EAC, Wavelab and Plextools. Moreover, iTunes can output bit transparent digital data as well. I will suggest audiophile make good use of iTunes. It has a lot of good feature, the album management alone is way above all other programs.
Okok, after repeating bit transparent, bit identical, how about the most important element SOUND!!!?
Do you think iTunes rip sound so bad compare to EAC rip? or compare to WaveLab Ultra Safe rip? Please re-read my previous articles for these answers. I understand there are many golden ears in audiophile circle that may argue this case……….okay, I have all these 10 audio files which all are ripped, recorded from different methods and sources. They are proven in my measurement that bit transparent and carry exact same digtail data. Will you like to take this challenge, listen these bit identical files and tell me if anyone of them sounds better than the rest?????
I guess not………but do let me know what you think. If you want my help to check your ripping files are bit transparent before you rip thousand of CDs, email me.
Hi,
Ok I’m totally confident in Itune now
About CD Import I’m unclear, “Use error correction when reading Audio CDs” need to be on ?
What do you think about Apple lossless (vs AIFF) ?
Thanks
Gilles
Hi Gilles,
Thanks for your comment.
I cannot find much details info about that CD “Use error correction when reading Audio CDs” feature under iTunes. But I think it is kind of a more secure way to read the audio data method, it will slow down the speed if there is needed.
I’m using plain AIFF or WAV non compressed file formats. Although it is lossless, I don’t think hard disk space is a matter to me and non-compressed files give me most option and store in the future. We have to beware that building a large library over 1000+ CD will be a pain for any changes in format.
For MAC, You can try MAX from http://sbooth.org/Max which can ripping/convert to many other formats.
Your data does not show a test on scratched, smudged, damaged, or aged CDs. If you just use a new, crisp CD for your tests then any special features for an exact audio copy or wavelab aren’t going to be used.
Hi JM, you are very correct. I have no intention to say that iTunes is better than EAC or Wavelab on CD ripping, it is not. This article is a follow up article to show that bit transparent can be achieved by iTunes, EAC and Wavelab. I guess the Amarra will be the thing for audiophile to playback music.
Hi your site is great!
I appreciate your scientific ways of proving things.
Can you tell me your e-mail? Couldn’t find it on the site =[
I have things that I would like to ask you
Thanks=]
Hello, thanks for your kind words.
Our company website is http://www.designwsound.com / eMail is contact@designwsound.com
In your explanation you have Import Setting in iTunes to: AIFF Encoder, but Custom setting is… Can you tell us what the Custom setting are i.e. bit rate, sample etc?
Great site.
If you select AIFF, you will see no other thing is needed. It will be just ALL Automatic.
But one thing that worth to mention, if you import high resolution files (like WAV / AIFF) to iTunes, and then select “Create AIFF” within the iTunes -> Advance, the new AIFF file will be limited to 16/44.1 only.
Thanks for your kind words.
Yes, but if you select AIFF and all is auto, how do you know what those settings are?
In the Auto box, you can also select Custom and put in the correct settings to assure it is native CD quality.
The custom panel just let you select 8 or 16 bit / Sampling rate from very low to 48kHz / Mono or Stereo.
The default is always 16bit44.1kHz
This creates the problem that I mentioned previously. If I imported a 24/96kHz WAV file to iTunes, that song cannot be embedded with album artwork. If you want to use itunes to export (create AIFF), your 24/96 WAV will become a 16/44.1 AIFF file.
Be aware if you have high resolution files.
This is also the reason we use AIFF on our records.
Kent,
How can I print articles such as “CAS 6″? If I do Print Preview, only two pages show.
Thanks.
Hello Frank, we have add a print option. You can see this on the bottom of all posts and page. Hope this help. Thank you for your support.
XLD seems to be the current “best ripping app” for Macs. It’s based on the CD Paranoia engine, and it supports AccurateRip.
iTunes *may* yield a bit perfect rip, and it *may not*. That is the general problem with it (as far as ripping goes): you never know, since it doesn’t give you a log after ripping.
I’ve found that if it can’t rip a scratched disc/track, it gives up and moves on, leaving you with an incomplete file. That in itself isn’t so bad, but iTunes doesn’t even tell you that it couldn’t rip the track!
Of course they’re identical — IF the record is free from scratches and you manually align the start offset! Those are two pretty big if’s. The whole point of EAC is that it does there two things for you: it re-reads tracks until the rip is bit perfect, and it compares rips to a database to find out the gap offset and correct for it.
A bit perfect ripper for unix is cdparanoia. It has a Mac frontend called Pillage (from bratproductions.com). The least you could do if you have to rip under OSX is to use that. Remember, friends don’t let friends use iTunes! (It turns out the encoder is the bigger problem with iTunes, optimized for speed and not quality. LAME always scores better in blind testing.)
I do not worry about the gap offset. The “starts” and “ends” of any track are zero (no content) anyway. It does not affect the bit transparency at all. There are other OSx rippers which do similar like EAC.
I do not understand what “encoder” do you mean about iTunes when comparing to LAME.
May I send you 3 files ripped from EAC, Itune and Adobe Audition 1.5?
for comparison becos i found that the error is happened!
hello freeman,
yes, please send me those files to take a look, thank you.
server something wrong, will upload later for your download
http://www.badongo.com/file/16253629
http://www.badongo.com/file/16253838
http://www.badongo.com/file/16254004
http://www.badongo.com/file/16254210
I uploaded, pls downaod and compare by EAC
Just wondering if you would get the same result using Apple Lossless instead of AIFF under iTune. Cheers,
Hello Freeman,
Sorry for the late reply. All your 4 ripped files are bit transparent. The EAC is the same as the iTunes ripping.
Hi,
Thank you for this site and the great information.
I have had quite the opposite experience with ripping with iTunes and no longer use it for ripping, just playback. This could have been due to the superdrive hardware in the Mac, but iTunes will not let you know if there was a hardware problem, EAC does. Even iTunes rips from Mac and PC can differ for the same file format. iTunes ALAC and WAV files differed on my Mac but not on my PC. Using a new cd, fresh out of the box and Foobar bit comparison program to compare the ripped files this was an unexpected outcome.
HI:
I use Itunes and like it very much. Currently I use it on a PC with XP and would like to move to Vista or Windows 7. In your examples what OS were you using on the PC with Itunes?
I use a USB to Digital Coax converter into my DAC and want to make sure that VISTA and WIN7 will be as good as XP. Any suggestions or comments, Thanks.
Joe
Hello Joe, Thanks for your message. I am using XP with iTunes on the PC. I think the Vista and Windows 7 should be as good as XP if your USB device is update to date and working correctly.
It appears that Itunes 9 witht the latest Quicktime now supports Windows Audio Session (WASAPI). In the Quicktime Preferences/Audio tab you can select Windows Audio Session. Doing that and manually setting the sampling in Quicktime I beleive should give bit-perfect results, although I have not tried it. It is encouraging. Not sure if this is possible in Windows 7. Someone said the Windows Audio Session button was grayed-out in Windows 7. In any case iTunes 9 and th elatest Quicktime along with XP and a Hagerman USB/Coax converter give me wonderful results much better than my Quad CD player using just the DAC in the Quad.
The new Quicktime option for Windows Audio Session is with Vista, not XP. Forgot to mention that.
More info:
Looks like iTunes works with WASAPI only on Vista 32. Vista 64 and Windows 7 not supported. Maybe Quicktime still has another update to release.
I’ve used iTunes to rip apple lossless files (osx) on a lot of ocassions and hav noticed that it adds blips/pops to the sound. I’ve done the same rip in XLD on osx and it didn’t give me any pops in the same section of the song. I have tried this numerous times to test the integrity of the rip and hav to conclude that iTunes is horrible at cd ripping. Maybe this doesn’t apply to aiff format but if ur ripping to apple lossless, I would recommend using XLD for mac. MAX is not bad as well but it is slower at ripping CDs.
Thanks for your comment JoeR. I am glad to hear you are having great result and enjoying the computer playback.
Thanks for your comment sound_wave. You are right that if there is ripping problem, iTunes will not notice the user like XLD, EAC, MAX, neither it does secure ripping (ripping more than 1 time to verify data transparent). However it is easier to use and manage from ripping to album arrangement. And it provides bit transparent ripping. Ripping is not a big deal for today CD reader anymore, however I agree there are more secure methods out there.
I have compelling evidence that the offset matters. For testing pursposes a colleague of mine and software engineer generated three tracks using the same data block, but with different offsets. Each of these three tracks sounds different in my system and the system that I exhibited at RMAF in Denver this month. I also have feedback from others that can hear the differences with both NOS and Sigma-Delta type D/A converters.
I have distributed these tracks to a variety of industry experts for examination. BTW, the engineer that developed these tracks could not hear the difference on his own system, but is interested in getting to the bottom of this as I am. If you want the tracks to examine, email me with your mailing address.
Steve N.
Empirical Audio
“We need to repeat one more time, the first goal of digital audio playback is “data integrant”. The most important issue is the data transfer from CD to your hard disk, then later output data perfectly. If this chain is transparent, the only issue left is “jitter”. ”
Jitter is certainly important, as I have also written white-papers on it. However, I dont believe that a data compare is sufficient to guarantee accurate playback of a ripped CD with iTunes. The reason that I dont believe this is Amarra. Amarra player combined with iTunes plays real music. The fog of iTunes is lifted. Again, at RMAF I demonstrated for hundreds of attendees that iTunes and iTunes with Amarra sound vastly different, with the nod going to Amarra.
My conclusion is that rippers matter, ripping drives matter and particularly the playback software matters, and there is more to this than just data compares.
Steve N.
Empirical Audio
Dear Kent,
I understand that high quality rips can be made using iTunes and that up to date computer/CD/DVD drives, are capable to rip bit perfect data. Being capable to rip bit perfect data does, however, not mean that bit perfect data is in fact always and under any condition being riped, i.e. if the CD is not in perfect condition. With this regards the following questions (which somehow unexpectedly remained unanswered at Computer Audiophile so far):
(i) Will it make any difference what drive I use (DVD/CD – CD only etc.) – as long as the drive is in good condition? It seems that you have been using the Plextor Premium 2 for your tests. You also mention the StartLab CD.1. Both are difficult to get here in Germany.
(ii) Furthermore, does the speed of the drive influence the quality of the rip? Generally, I read that slow speed is better than faster speed. Sounds reasonable to me; I presume that less error correction might be required.
As to (i) – and only if you want to comment on third party products – I would also be interested to know your opinion with regards to the AQVOX CD drive (www.aqvox.de). As to (ii) – and of course only if speed matters – I would like to know if you have any recommendation how to slow down CD drives under Mac OS 10.5.8. (DiscRotate does not work properly with my (internal) DVD/CD drive. Scharping works only if iTunes is used as the CD player (I have not tried this, though, and, despite the above, “feel” more comfortable using Rip). No clue how to handle hdparm and cdcontrol…)
Thanks for this very informative site.
Philipp
Hi Kent,
If itunes can do the job properly, why need to spend a few thousand dollars for Amarra?
@philipp
Hello Philipp, I agree that if iTunes can more securely rip CD will be great.
In this ripping test, I also used one very cheap external Asus DVD-Rom hook up to a Powerbook via USB1.0 connection. In normal term, ripping in lower speed may be better if there is problem on the disc, however this is not necessary in today computer standard.
The AQVOX seems like quality piece of gear.
@V Chan
Hello V Chan, this article has very objective scientific tests on digital audio bit transparent on CD ripping. I think this concept is important to audiophile.
Regarding Amarra, it is a pure MAC software and it is capable to automatically switch sampling rates regarding the native file sampling rate. This avoids any processing or SRC from iTunes, which is very bad sounding. The volume fader and EQ of Amarra is also EXCELLENT. Being subjectively look at its cost vs. other audiophile products (cables, racks, tip toe etc) I think it is a bargain. However if you look from the perspective that iTunes is free, and Amarra is HK$8000, that is not easy to justify.
I told Sonic people that the only way to sell Amarra is to perform demo…….people will judge by themselves if it is worth.
water.logic, your statements are over the line. I have deleted both your posts.
The test could be better with the same work with error correction off just to compare ?
@audioengr
How can offsets change the sound of audio? If you cut off the first second of a CD (ie. first second of the first song), will the rest of the CD sound different? Of course not! You just miss out on the first (few) note(s).
While it is good to be offset corrected, far too much importance have been placed upon offset correction. The author of EAC (Andre) did his best to ESTIMATE what “point zero” should be, but by no means should it be considered to be absolutely correct. At best, “point zero” (ie. when the offset is corrected) is good so that all CD rips, from different rippers, have a common point of reference.
(All other rippers that support offset correction, and AccurateRip, are based on EAC’s point zero.)
PS. XLD (X Lossless Decoder) seems to be the preferred ripper for Mac OS now.
And anyone who thinks iTunes is as good as secure rippers … Well, your loss. The point of a secure ripper is that if it reports no errors after ripping, you can be fairly secure (no ripper can guarantee 100%) that your rip is error free. iTunes offers no such security, which is why it and other non-secure rippers are not favoured.