Guide to Proper Lossless Rips

Copied from one place and then to another, and now it returns to its origin.

Ripping Guide

How to use Exact Audio Copy

Download EAC from here: exactaudiocopy.org

Before installing EAC you will probably want to install an ASPI layer for your computer.

If your machine has SCSI hardware, and/or built in CD burner, get the one from Adaptec here.

(Note: Although there are newer versions; I do not recommend them. They screwed up my computer; and some other people's; they might do the same to yours. The older version is more reliable.)

If you have Firewire or a USB/USB2 external burner you will want to use the ASPI from Nero (copy this file to the system32 folder on your computer or put it in the directory where you install EAC and reboot). (See attachment: winaspi32.rar)

Required EAC Settings:

Menu Action:

• Select: append gaps to previous track (default)

Required EAC Options:

Extraction tab:

• Check: Fill up missing offset samples with silence
• Check: Synchronize between tracks
• Select: Error recovery quality - High
• Rest: leave unchecked

General tab:

• Check: on unknown CDs, select automatically access freedb database
• Select: use English language

Tools tab:

• UN-Check: retrieve UPC/ISRC codes in CUE sheet generation
• Check: Use CD-Text information in CUE sheet generation
• Check: Create '.m3u' playlist on extraction
• Check: Automatically write status report after extraction
• UN-Check: Activate beginner mode, disable all advanced features

Normalize tab:

• Do NOT use Normalize

Filename tab:

• Naming Scheme = %N - %T
• Various Artists Naming Scheme = %N - %A - %T

Required Drive Options:

Extraction Method tab:

• Secure mode - NOT Paranoid, Synchronized or Burst mode!
• Check: Drive has 'Accurate Stream' feature
• Check: Drive caches audio data
• Do Not Check: Drive is capable of retrieving C2 error information

Offset/Speed tab:

• Select: Use read sample offset correction or test and check your offset with this tutorial
• Do Not Check: Overread into Lead-In and Lead-Out (check only if retail CD is over 80:00 min.)
• Check: Allow speed reduction during extraction New
• Check: CD-Text Read capable drive New

Ripping and creation of CUE sheet:

• put in CD
• ctrl-a (select all tracks)
• F4 (detect gaps)
  Important: If the first track has a gap of more than 2.00 seconds and a start time of more than 0.00 seconds, see the FAQ for how to rip this track!
• EAC, menu action, create CUE, multiple WAVs with gaps (noncompliant)
• F6 (test and copy tracks) or shift-F6 (see later, when EAC is setup to encode to FLAC)
• then click "okay" and explicitly save/create logfile

Don't forget to check the CRCs after each rip, because unlike other errors they do not generate a There were errors message (the log doesn't warn you). The Test CRC and the Read CRC should be the same.

- Encoding with FLAC [/b]
FLAC is a great lossless encoder, arguably the best. Generally what we go for when it comes to encoding wavs into lossless files is how small we can get the files, but then again how quickly they will decode when we want to play them in our favorite audio players. FLAC does both well.


Download FLAC 1.1.2 for Windows (tools only)

EAC options (F9), tab "tools" (example settings):

• Check: on extraction start external compressors queued in the background
• set number of threads to 1 (you can set it higher, but it makes no sense)
• Check: do not open external compressor window (recommended, you can leave it unchecked for debugging purposes)

EAC compression options (F11), tab "External Compression" (example settings):

• Uncheck everything on Waveform tab
• Uncheck "use offset" on Offset tab
• Uncheck everything on ID3 Tag tab
• Click tab External Compression
• Check: User external program for compression
• set Parameter passing scheme to "User Defined Encoder"
• set file extension to .flac
• indicate path to flac.exe
• set Additional commandline options to
  -5 -V -T "ARTIST=%a" -T "TITLE=%t" -T "ALBUM=%g" -T "DATE=%y" -T "TRACKNUMBER=%n" -T "GENRE=%m" -T "COMMENT=EAC Flac 1.1.2 -5" %s
• Bitrate does not matter (FLAC ignores it)
• Decide whether or not you want the WAV deleted after compression
• Check: use CRC check
• Uncheck: Add ID3 tag (FLAC has own tags)
• High or low quality also does not matter for FLAC (is ignored)
• Remember to press shift-F6 (not F6) when ripping, otherwise EAC will not compress/tag!
• Also remember not to close EAC immediately after ripping, look at status bar: the last track might still be compressing

Example of an EAC log from an Accurately Ripped CD

Quote:

EAC extraction logfile from 30. December 2005, 19:43 for CD
Vince Guaraldi Trio / Jazz Impressions of Black Orpheus

Used drive : SAMSUNG CD-R/RW SW-240B Adapter: 1 ID: 1
Read mode : Secure with NO C2, accurate stream, disable cache
Read offset correction : 97
Overread into Lead-In and Lead-Out : No

Used output format : C:\Program Files\FLAC\flac.exe (User Defined Encoder)
128 kBit/s
Additional command line options : -V -8 -T "artist=%a" -T "title=%t" -T "album=%g" -T "date=%y" -T "tracknumber=%n" -T "genre=%m" %s

Other options :
Fill up missing offset samples with silence : Yes
Delete leading and trailing silent blocks : No
Installed external ASPI interface


Track 1
Filename Z:\[Temp]\[Rips]\Vince Guaraldi Trio - Jazz Impressions of Black Oprheus\01 - Samba De Orpheus.wav

Pre-gap length 0:00:02.00

Peak level 81.5 %
Track quality 99.9 %
Test CRC EF678A9C
Copy CRC EF678A9C
Copy OK

Track 2
Filename Z:\[Temp]\[Rips]\Vince Guaraldi Trio - Jazz Impressions of Black Oprheus\02 - Manha De Carnaval.wav

Pre-gap length 0:00:03.66

Peak level 67.1 %
Track quality 100.0 %
Test CRC F4EC9EA3
Copy CRC F4EC9EA3
Copy OK

Track 3
Filename Z:\[Temp]\[Rips]\Vince Guaraldi Trio - Jazz Impressions of Black Oprheus\03 - O Nosso Amor.wav

Pre-gap length 0:00:01.90

Peak level 99.5 %
Track quality 100.0 %
Test CRC 930987F3
Copy CRC 930987F3
Copy OK

Track 4
Filename Z:\[Temp]\[Rips]\Vince Guaraldi Trio - Jazz Impressions of Black Oprheus\04 - Generique.wav

Pre-gap length 0:00:02.89

Peak level 77.2 %
Track quality 99.9 %
Test CRC D23C4017
Copy CRC D23C4017
Copy OK

Track 5
Filename Z:\[Temp]\[Rips]\Vince Guaraldi Trio - Jazz Impressions of Black Oprheus\05 - Cast Your Fate to the Wind.wav

Pre-gap length 0:00:02.56

Peak level 74.3 %
Track quality 100.0 %
Test CRC 61C87A70
Copy CRC 61C87A70
Copy OK

Track 6
Filename Z:\[Temp]\[Rips]\Vince Guaraldi Trio - Jazz Impressions of Black Oprheus\06 - Moon River.wav

Pre-gap length 0:00:03.33

Peak level 48.6 %
Track quality 100.0 %
Test CRC A181EEAD
Copy CRC A181EEAD
Copy OK

Track 7
Filename Z:\[Temp]\[Rips]\Vince Guaraldi Trio - Jazz Impressions of Black Oprheus\07 - Alma-Ville.wav

Pre-gap length 0:00:03.06

Peak level 88.2 %
Track quality 100.0 %
Test CRC D402BBA6
Copy CRC D402BBA6
Copy OK

Track 8
Filename Z:\[Temp]\[Rips]\Vince Guaraldi Trio - Jazz Impressions of Black Oprheus\08 - Since I Fell for You.wav

Pre-gap length 0:00:03.73

Peak level 67.1 %
Track quality 100.0 %
Test CRC 7451388C
Copy CRC 7451388C
Copy OK

No errors occured


End of status report

Things to note about this Exact Audio Copy log:[*]The CD drive is listed and so is the drive offset that was used during the rip. Technically you can look up this CD drive and confirm this drive offset is correct for the drive.[*]You can see most of the settings that were used in the rip.[*]You can see some of the tracks have pre-gaps; which means gaps were properly detected before the rip.[*]You can see that the test CRC and copy CRCs are identical. This could possibly be falsified if someone wanted to be an ass, but the chances of that are unlikely.

Using AccurateRip
AccurateRip may or may not necessary depending on how your'e doing your rips. If you're following this guide entirely and ripping CDs using the Test/Copy method in Secure mode, AccurateRip isn't needed. If you're not doing a Test before you Copy or using a mode besides Secure, AccurateRip is for you since you probably want to be sure that your rip is actually accurate.

Originally Posted by AccurateRip

The philosophy behind AccurateRip is quite simple - each time an audio track is ripped (recorded to computer) it is compared with rips from other people, this allows a confidence report to be generated. The report might say that 4 other people had exactly the same results, this would guarantee your rip was without error, or the report could say that 3 other people disagree with your rip, the likely hood is that your CD has a scratch, or should be cleaned.

I thought of writing my own guide for this but it's pretty much already been fully outlined by the help file that comes with AccurateRip and for me to say anymore would be redundant.

If you're having trouble with AccurateRip (I have in the past) check the FAQ at the bottom of this post.

You can get AccurateRip here.

Burning Guide

Making 1:1 copies
(coming soon)

Backing up as data

FNAQ (Frequently Not Asked Questions)
Q: What if the first track has a pre-gap of more than 2.00 seconds and/or a start of more than 0.00 seconds?
A: If you have the most recent version of EAC you don't have to worry about this. You can continue to read if you like.

95% of all CDs have a first-track pregap of exactly 2.00 seconds. In the rare case that they do not, 95% of these CDs have a silent pregap -- there is no audio data contained. However, it is difficult to be sure that this extra pregap time is entirely silent.

Newer versions of Exact Audio Copy account for pre-gaps of this type with a special setting in its noncompliant cuesheets (which you generate before ripping); HOWEVER, not all pre-gaps are entirely silent. That said, if you download a rip that doesn't have have the pregap ripped, there's a 99% chance that you can still burn a 1:1 copy (entirely identical to the original CD) with such a rip, but there is still a slim chance that the extra pre-gap isn't silent and that audio data was thus NOT ripped and you CAN'T burn a 1:1 copy with the rip.

That said, if you are a true audiophile and perfectionist, you should not automatically discredit the lossless-ness of a download that doesn't have the extra pregap explicitly ripped. However, if YOU are ripping a CD with a pregap of more than 2.00 seconds, you should take a few extra minutes to rip the pregap properly:

This is a specific case where the first track has been given a silent pregap. This is very common on classical CDs. On any normal CD, the first track should have a pregap of exactly 2.00 seconds. If this is off, a specific ripping procedure must take place.
Example:

• First, rip the CD as you normally would, using this guide.
• After doing that, select the first track of the CD
• Action -> Copy selected tracks index-based -> Compressed (or Uncompressed, doesn't matter in the long run)
• After the track is finished being ripped you'll end up with two separate files, one labeled "1.00 - <Trackname>" and one labeled "<1.01 - Trackname>". Delete the labeled "1.01 - <Trackname>" (this would have already been ripped if you ripped the entire CD before the pregap) and keep the other. This is the pregap track.
• Rename the file to something like "00 - Pregap" or just "Pregap"
• In EAC: Action -> Create CUE sheet -> Single WAV file. Save this CUEsheet with a different name than your Noncompliant one! People will most likely want both.

If you later want to burn this rip to make a 1:1 copy, you'll need a utility like foobar to combine all of the tracks into a single WAV file and then use the Image CUEsheet to burn it in Exact Audio Copy.

Q: AccurateRip is being buggy and shit; it's not working as it should when EAC is running.
A: First:

Originally Posted by AccurateRip Forum

Check to see if the option "Show status dialog after extraction" is on. Also be sure *one* of either "Overread into lead in/lead out" or "Fill up missing offset samples with silence" is true. Also "Remove leading and trailing silence" must be false. If any of these options are not set as listed EAC won't check the AR database after extraction and doesn't save your CRC results for submission.

If that ain't it... I run into this problem from time to time. Solution? Baleet AccurateRip and all associated files, then reinstall it. Arbitrarily, it malfunctions based on the timing of your installation in relation to Exact Audio Copy.

Q: But what happens if I don't... ???
A: I'm including this as part of the guide to inform you why the method to this madness is so... Well, methodical.

Why are you going through all this trouble to make a rip? Why are you getting disowned for not following these rules? Well, two issues come into play, here.

One of them is losslessness; you want to give the user everything he or she needs to burn a 1:1 copy of a CD, or as close to that as practically possible.

The second reason is verification of losslessness; sure you say you did it right, but how can anyone confirm this? How can anyone they share your stuff with confirm this?

Why append gaps to the previous track?
Because it's the most failsafe, in a nutshell.

A gap is (usually) a certain amount of digital silence. Digital silence meaning that there's nothing there for your audio player to "interpret" but the concept of "don't play anything here." A gap doesn't "belong" to any track; it exists in a realm between tracks. Of course when we are ripping, we don't want to lose any audio during the process or the concept of losslessness is a fail.

So when ripping in EAC, we append the digital silence of gaps to the track before it. If we don't care about losslessness, we don't need gaps at all. Technically, we could append the digital silence to the beginning of the track after the gap. However, let's say the gap is 10 seconds long. Imagine trying to play a a single song on your PC and you have to wait for 10 seconds to pass before you start hearing music. It's far more annoying than waiting for ten seconds to pass after the song has audibly ended anyway.

Why fill missing offset samples with silence?
There is a feature that is very rare in optical drives that is the ability to "Overread." Basically, unless you drive has this feature, you are losing the last few "frames" of audio at the end of any CD you rip! Now, you'd think this would be bad for losslessness. Thankfully, what you're missing is silent 95% of the time anyway. So you're filling the silence you didn't actually capture with silence anyway.

Why not normalize?
Some CDs are quieter than others. We've all seen this. Some CDs are louder than others. Normalization makes up for this during the ripping process. However, if you modify the actual audio of a file, then it is no longer lossless. That's why we can use ReplayGain to make up for this after we've ripped.

Read sample offset correction? Huh?
Yep. Your drive is going to either start reading a CD a few samples too late or a few samples too early. To account for this we use read sample offset correction; to force your drive to start reading earlier/later than it normally does. If we don't, we'll either end up with duplicate samples or lose some samples.

Although to be honest, a sample is very small! A sample is 1/44100 of a second; you're not really losing anything audible. If this doesn't bother you, don't be upset when you an encounter a rip with no offset correction or incorrect offset correction.

However, for the sake of losslessness, it makes sense to use offset correction when ripping your own stuff. That way, people who do and don't care will be happy with your rips.

Why should I check for gaps?
If you do not check for gaps, the worst that will happen is that it not show up in your logs. Honestly, I'm not sure why the "lossless community" insists on checking for gaps. For one, before you create a CUEsheet, the gaps will be checked (although the logs won't reflect this). Secondly, the audio you rip with gaps checked will be identical if you do not check gaps. Like I said, the worst that will happen is that the gaps will not show up in the logs.

For now, there's no reason not to check gaps. It takes only a few seconds and can be done with the press of a button. Furthermore, more people will be willing to download your rips in the future (if that matters to you).

But if you want to download a rip in which logs were not checked, you can rest assured that it is still lossless (assuming everything else was done correctly.)

Jam it back in, in the dark.

EDIT: Oh, here it is:

1. Action -> [check] Leave Out Gaps
2. Action -> Create CUE Sheet -> Multiple WAV Files With Left Out Gaps

This CUESheet will insert digital silence where it belongs. Gaps are appended for losslessness sake; when the burning takes place, they don't actually stay attached to any one track, they exist in a limbo between tracks. A noncompliant CUESheet generally deals with this.

The reason people don't just rip this way anyway is because some people want to playback the audio - as it should sound on a CD - on their computer.

There's nowhere I can't reach.

I'm actually not sure. I personally use Accurate, Method A, but since I've never seen gaps get incorrectly detected with any given method, I'm unsure.

This thing is sticky, and I don't like it. I don't appreciate it.

I can't say for sure. I know how to deal with ripping audio and data separately, but putting them together for a 1:1 copy? I wouldn't even know how to verify the overall losslessness.

You can try ripping the audio from EAC and just copying the data files manually.

I am a dolphin, do you want me on your body?

I'm not sure what you mean; but I believe this is normal. There is a 2-second pregap before the first track of any audio CD (sometimes longer).

Are you playing this CD on a standalone CD player or on your PC?

I was speaking idiomatically.

Exact Audio Copy is ideal since only it and another program, Burrrn, can read noncompliant cuesheets.

Exact Audio Copy also allows the use of a write offset correction. (Just like your drive reads too early or too late, it writes too early or too late also.) Burrrn does not.

Basically, if you want 1:1 copies, you have to use EAC.

What kind of toxic man-thing is happening now?

I've heard of that tool; I haven't used it in years. I don't know for sure if ripping audio as data yields a 1:1 copy. Would you mind doing a wav comparison to that audio ripped from the CD with EAC and then the audio ripped from EAC from the burned iso generated by CDRWin?

EAC has a wav comparison tool, if you didn't already know.

FELIPE NO

Can your burn BIN/CUEs with EAC? Because I don't know any other software the supports write offset correction.

What, you don't want my bikini-clad body?

Yeah that sucks. On the other hand it is pretty much perfect in every other regard.

Jam it back in, in the dark.

Yes, but in the process of making 1:1 copies of CDs, burning Audio + Data is sometimes necessary.

There's nowhere I can't reach.

Originally Posted by sabbey

Here's a quick question...

I have been using "Installed external ASPI interface" on my rips under the "Use of SCSI interface" option, since I heard it makes for better quality. However, my external DVD burner seems to only show up in EAC with "Native Win32 interface for Win NT/2000/XP" selected. Any idea why, and is there anything I can do to change that? Hell, is there really any difference between the two? If not, I'll make the switch back to the other so all three drives are listed.

Overall, I can't say I care too much, only using it for ripping hybrid SACDs. None of the others in my experience can see such discs when one is in the drive.



BTW, any idea how one can find what the write offset is? That read offset database you linked to is of great help, is there one for the write offset as well?

Sorry for the late response. Actually there's a guide on it here. It takes some time but it's not terribly difficult.

This thing is sticky, and I don't like it. I don't appreciate it.

I don't know, can you decode it to wav, with official flac or third-party tools?

Most amazing jew boots

I used to make par2 files when I backed up music to guard from inevitable CRC errors. But then I realized I was killing my self trying to make them; very time/space consuming. I just gave up. I probably *should* do it for rarer things but for just downloads in general, it's just too much effort.

I was speaking idiomatically.

Still takes a great deal of time to deal with.

Maybe if their was a tool that allowed me to drag and drop several folders and just make par2 files and move them somewhere. Then I could just go ahead and make par2 files for all the stuff I've already backed up and future stuff as well.

But I instead always have to drag/drop chunks of flac files, logs and cues, artwork, etc. Then there's sometimes multiple discs and such; subdirectories. I just got done backing up 200GB worth of lossless and grew tired after making par2 files for the 20th DVD or so. I thought, forget it, if some stuff get's CRC errors on it then so be it. Maybe I'll make par2 files later for rarer stuff (presumably VGM since 99% of the lossless music community doesn't have or share any.)

To help myself I've bought more reliable (expensive) media; as I have most generally gotten CRC errors on cheaper, more generic media. Ideally I'd get some Taiyo Yuden discs, but right now I am using Ritek/Ridata and Verbatim.

EDIT: Also let's not forget that DVDs have a certain amount of error correction native to their design.

What kind of toxic man-thing is happening now?

You might have to detect the read offset in a more manual way.

You can do this by ripping a WAV with your LG drive using its correct read offset. Then, rip the same track with your Philips drive, configured with no offset, to WAV. Then you can use EAC's WAV comparison (Tools > Compare WAVs) to see if there are missing or duplicate samples between them.

I believe if there are duplicate samples then your Philips drive would be reading too early; you want to take the number of sample differences and create a positive offset from that. If there are missing samples then your drive is reading too late, you want to take the difference in samples and create a negative offset from that.

Or it might actually be vice-versa of what I said (positive for missing samples, negative for dupe samples). Keep playing around with it until WAVs ripped from both drives are identical (ie no difference in samples).

FELIPE NO

Oh, I thought it was a no-go with AccurateRip unless your drive was in the database. Finding 3 CDs can be a bitch, though, as there are different pressings of certain CDs, so even though it says Nirvana - In Utero will work, your copy of it might not work and can't be used for detection.

What, you don't want my bikini-clad body?

Okay, that's believable.

Jam it back in, in the dark.

Elixir, you used Burst mode.

Also, Blue_Kirby, AccurateRip has two functions. One is to find your drive offset. It's other function is to compare your rips to the results of other people who have ripped the CD to see if your rips match.

If you're using Test/Copy when you rip your CDs, and the CRCs are identical, AND you have the right offset for your drive, AND there are no errors during the ripping process, the chances that your rip are incorrect are very, very slim. Your drive is essentially attempting to rip the same data twice and getting the exact same results each time, which is a good thing; it means that it's most likely doing it correctly.

If you're not using a secure ripping method or not using Test/Copy, there's no way to really verify that your rip is accurate just by looking at the logs. That's when AccurateRip comes in; it checks your results against other people's results and tells you if your results match up or not.

AccurateRip isn't always perfect for a lot of reasons, BUT it will never tell your that you ripped something correctly when you didn't. But it MIGHT imply your rip is not correct when it is. This is because a lot of CDs have multiple pressings, so the results in the AccurateRip database might be from a different pressing of the exact same album you have. A different pressing of the same CD causes a slight difference in the data.

There's nowhere I can't reach.

Originally Posted by Elixir

Alright, I've just ripped again with Burst off.

The bitrates are identical, however, so I can assume that the songs are also?

I don't think you mean bitrates, I think you mean CRCs from the logs, right?

A CD ripped with Burst mode isn't always secure. I would recommend you rerip whatever you have ripped.

This thing is sticky, and I don't like it. I don't appreciate it.

Well every audio CD has a gap of two seconds (I think; there may be some exceptions, I don't know); it's normal for the CUEsheet to exclude the first two seconds from its pregap flag because it's assumed they exist.

Then you have to take into account that EAC is measuring in seconds when time on a CD is generally measured in frames. It is my understanding that frames are the "correct" unit of measurement for time on a CD.

I believe EAC natively displays time in hundredths of a second (you can change this option in Options -> General tab), which is more precise than necessary, because there are only 75 frames in a second on a CD. So in your log you have 2.44 seconds. In your cue, those 2 seconds are ignored, leaving you with 44 hundredths of a second; which is equal to 33 frames. (100/75 = 44/33). Similarly, 64 hundredths of a second equal to 48 frames, and so on. You can do the conversions for yourself.

Bottom line, just ignore it

Double Post:
Thinking about it, your question would make a good addition to the FAQ.

I am a dolphin, do you want me on your body?