5. Configuration - Drive Options
- Hit F10 or click "Drive Options" in the EAC menu.
Note: If you want to configure more than one drive for ripping, you will need to save the settings for the first drive, then re-do the drive settings for the next drive, and save those settings under a different name. See Saving Profiles. Choose the drive you want to configure by selecting it in the drop-down menu in the main window.
Also, if you upgrade the firmware on your drive (flash the firmware) you will need to set up the drive options for EAC again.
If you have elected to perform CacheX and Feurio tests, you also need to do them again if EAC still reports that your drive does not cache audio data.
You need to insert a CD in the drive in order to run some of the drive tests. It needs to be a CD that can be found in the database, so bring a pile of reasonably well-known CDs with you to the computer.
- Insert a CD that is in the AccurateRip key disc database.. There will be a popup offering you to take advantage of the opportunity to have some of your drive features automatically set by AccurateRip:
- Click "Configure". You will be greeted with a success message after a few seconds:
If AccurateRip does not start automatically, and you do have a CD that is in the AccurateRip key disc database in the drive you are about to configure, make sure that the "Use AccurateRip with this drive" option in the Offset/Speed tab is checked (see 5c. Offset/Speed). If the option is greyed out, so that AccurateRip is not available to you, see 5f. Appendix: If "Use AccurateRip with this drive" is greyed out, but note that it will be greyed out unless/until you have a CD that is in the database in the drive. Be prepared that you may have to try with a number of CDs until you find one that is in the database.
- Leave the CD in the drive, it will be needed for some of the following tests.
5a. Extraction Method: Secure
Since this is a "radio button" choice, it follows that "Paranoid Mode", "Synchronized Modes" and "Burst Modes" should all be left unchecked. (But see Alternative Extraction Method: Burst for using burst mode instead of secure mode after all.)
- Next, click "Detect Read Features" in order to let EAC check if your drive caches audio data, and if it has Accurate Stream (in which case EAC has no need to perform additional synchronization). It will also check for C2 error info, but we will leave this option unchecked no matter what EAC says about it.
We will trust what EAC says about the Accurate Stream feature. We will also trust EAC if it says that the drive caches audio data, like this:
- After the detection is done, click "Apply" to have the detected features automatically applied as settings. Note that in most cases you will change those detected settings, though, so do not leave it like this.
- Uncheck "Drive is capable of retrieving C2 error information" - if the feature is left enabled, some errors may in fact go undetected.
Quote from EAC's FAQ page:
"What is C2?
On all CD-ROM media are at least two levels of error correction, called C1 and C2. If both fail, the output is probably not correct anymore. Most drives are not able to report if audio reads failed or not, so each block had to be read twice and be compared to make sure that everything is fine. But some newer drives are able to report if C1/C2 failed on specific samples on a read, making it possible to read only once and see if a read error occured. But there is still a problem, as some drives do not report these errors correctly, so you should test it thoroughly before trusting the results.
It seems that C2 is not correctly implemented in some drives. To be on the safe side, you should turn off the C2 error correction."
In other words: C2 could be utilised for speeding up the extraction process if you dared trust it with EAC. It would make (a lot of) the re-reads that EAC performs unnecessary. However, testing your drive is a bit... involved. See this page about DAE Quality (scroll down a bit). So the recommendation is to simply not use C2 with EAC rips, even if it would be OK in some cases. If you use it with EAC and it does not perform as expected, errors may go undetected and thus uncorrected.
EAC may show that your drive does not cache audio data, like this:
You can still choose to have "Drive caches audio data" checked as shown below. It will do no harm to the rip quality, but the rip may take a bit longer. So:
- Check "Drive caches audio data" if it did not get automatically checked after detecting drive features. (See below about "Testing cache with CacheX and Feurio!", though.)
It is very important that "Drive caches audio data" is checked if your drive does cache the data. EAC needs to read from the disc, not from a cache, each and every time. This is the basis for error detection and correction, and thus for establishing that the rip is a (near) perfect copy of the disc: it is almost impossible that the data would read the same more than once if the drive did not read it correctly from the disc. However, if the data is cached it is no surprise that it reads the same the next time it is accessed.
This is how the settings should be. When you have got it right, feel free to go straight to 5b. Drive (<- CLICK).
If you wish, you can return and read the following sections about testing the cache, and perhaps ripping in burst mode, some other time...
So, a drive that caches audio data defeats the purpose of making secure rips, unless you tell EAC that the drive caches audio data. If EAC "knows" that a drive caches audio data, it will flush the drive's cache in order to be able to read from the disc again.
There are doubts about whether EAC really detects drive caching properly. This is why many guides recommend (or even demand) that you always disable/defeat cache (that is, check "Drive caches audio data"). As far as rip quality is concerned, it is never wrong to have "Drive caches audio data" checked.
Testing cache with CacheX and Feurio!
Note: This is not compulsory. Only do this if EAC states that your drive does not cache audio and you are convinced that you want to do the extra work of testing your drive and, perhaps, rip without defeating cache. This is by no means necessary to do. As stated above, letting EAC defeat a non-existing cache will not harm the rip quality at all.
However, it may seem pointless to defeat an audio cache that does not take place in the first place (and some sources indicate that it puts unnecessary strain on the drive, shortening its lifetime). There are other tools for testing whether the drive caches audio data or not, such as Spath's Cache Explorer and Feurio!. If EAC reports that your drive does not cache audio data, you have to test with one or both of them in order to determine if your drive really does not cache audio data. If you do not perform further tests on your drive, or if any of the tests show that your drive does in fact cache audio data, you have to check "Drive caches audio data" as shown above.
N. B. EAC reads data in 64 kB chunks, so a cache smaller than 64 kB will actually not affect the ripping result. See this Hydrogenaudio Wiki article.
Spath's Cache Explorer (shortened to CacheX from now on) is a small command line tool that can actually be used to investigate some other drive features, too, but here we will only look at the drive cache detection features. Test with this tool first!
Download CacheX from here. It is a .zip file that needs to be extracted before you can use the program. Extract it to a location of your choice; if you extract it to C:\ it will be easier for you to access the program in the command prompt, as described below. I recommend you to do so if you are not used to working with command line tools.
(Since I like to have my files logically structured, I created a CacheX folder in "Program Files" for it. This is why the screenshots show a slight difference compared to the instructions in the text. Just ignore it. Do as the text says.)
- Insert an audio CD in the drive you want to test, then click Start - Run and type "cmd" (without the quotes), hit Enter or click "OK". This will open the command prompt.
Here is how to perform the test, assuming that you extracted CacheEx to C:\ as recommended above:
- In the command prompt, go to C: by typing "cd\" (without the quotes), hit Enter.
If you are curious about what the program can do, you can get a list of the command line options by typing "cachex --help" (without the quotes).
- For a "standard" test for audio caching, type "cachex -i -c -n 10 [drive letter]:" (without the quotes), then hit Enter. Of course you have to substitute [drive letter] with the actual drive letter for your drive.
In the examples below, I have tested the drives on M and D, respectively. The -i command makes CacheX show information about the drive. Note that the drive may report a buffer (that can be used to cache data), but still not cache audio data. The -c command makes CacheX test if the drive actually caches audio data. The -n command tells CacheX how many times it should test if the drive caches audio data. Let it test at least 10 times in order to make sure that the result is reliable. The cache will not be detected in each and every test, so running it just a few times may give you a misleading result.
(The user name has been edited out from the home directory C:\Documents and Settings\*USERNAME*, creating an abnormal empty space in that line in the screenshots. Just ignore it.)
This drive caches audio data according to CacheX - with a result like this, you have to check "Drive caches audio data" in the drive settings for EAC:
This drive does not cache audio data according to CacheX:
If CacheX shows that your drive does not cache audio data, you may run yet another test. Feurio! is an "ordinary" Windows application with installer and graphical user interface. Download it from here and install the trial version. It is pretty straight-forward to use it for testing your drive cache. Insert an audio CD into the drive. Click the "Program" menu and choose "Program Parameters", select the drive you want to test in the drop-down menu, click the "Test Device" button (bottom right), check only "Cache test" for "Do the following tests", then click "Start test". When it is done, save the results. Here is an example of a report for a drive that does not cache audio data:
Informations from INQUIRY command:
Manufacturer: PLEXTOR, Product: DVDR PX-800A, Version: 1.00
Synchronous data transfer: Not supported
Reading device capabilities: OK
Maximum speed: 7056 kByte / second (40.0 times)
Cache size: 2048 kByte
Read CD-RW: Yes
Read Bar code: No
Read UPC code: Yes
Read ISRC code: Yes
Return C2 error pointers: Yes
Read R-W subcodes: Yes
R-W subcode de-interleaved: No
Read CD-DA: Yes
Read CD-DA correctly: Yes
++ Cache test
Number of sectors: 1 (=2 kByte) -> 1.438 MBytes / second
Number of sectors: 2 (=4 kByte) -> 0.125 MBytes / second
Number of sectors: 3 (=7 kByte) -> 0.217 MBytes / second
Number of sectors: 4 (=9 kByte) -> 0.290 MBytes / second
Number of sectors: 5 (=11 kByte) -> 0.363 MBytes / second
Number of sectors: 6 (=14 kByte) -> 0.435 MBytes / second
Number of sectors: 7 (=16 kByte) -> 0.508 MBytes / second
Number of sectors: 8 (=18 kByte) -> 0.578 MBytes / second
Number of sectors: 9 (=21 kByte) -> 0.569 MBytes / second
Number of sectors: 10 (=23 kByte) -> 0.544 MBytes / second
Number of sectors: 15 (=35 kByte) -> 0.817 MBytes / second
Number of sectors: 22 (=51 kByte) -> 0.802 MBytes / second
Number of sectors: 15 (=35 kByte) -> 0.816 MBytes / second
Number of sectors: 16 (=37 kByte) -> 0.871 MBytes / second
Number of sectors: 17 (=39 kByte) -> 0.739 MBytes / second
Number of sectors: 18 (=42 kByte) -> 0.784 MBytes / second
Number of sectors: 19 (=44 kByte) -> 0.826 MBytes / second
Number of sectors: 20 (=47 kByte) -> 0.866 MBytes / second
Number of sectors: 21 (=49 kByte) -> 0.808 MBytes / second
Number of sectors: 31 (=72 kByte) -> 0.966 MBytes / second
Number of sectors: 46 (=108 kByte) -> 1.114 MBytes / second
Number of sectors: 69 (=162 kByte) -> 1.363 MBytes / second
Number of sectors: 103 (=242 kByte) -> 1.496 MBytes / second
Number of sectors: 154 (=362 kByte) -> 1.598 MBytes / second
Number of sectors: 231 (=543 kByte) -> 1.623 MBytes / second
Number of sectors: 346 (=813 kByte) -> 1.692 MBytes / second
Number of sectors: 519 (=1220 kByte) -> 1.729 MBytes / second
Number of sectors: 778 (=1829 kByte) -> 1.741 MBytes / second
Number of sectors: 1167 (=2744 kByte) -> 2.631 MBytes / second
Number of sectors: 1750 (=4116 kByte) -> 2.669 MBytes / second
Number of sectors: 2625 (=6174 kByte) -> 2.669 MBytes / second
Number of sectors: 3937 (=9259 kByte) -> 2.718 MBytes / second
Number of sectors: 5905 (=13888 kByte) -> 2.751 MBytes / second
Number of sectors: 8857 (=20831 kByte) -> 2.806 MBytes / second
Number of sectors: 13285 (=31246 kByte) -> 2.856 MBytes / second
Number of sectors: 19927 (=46868 kByte) -> 2.926 MBytes / second
Maximum transfer rate: 2926 kBytes/Second
Cache size for audio data: 0 kByte
It seems that your device isn't able to cache audio data!
If you have tested your drive with CacheX and Feurio!, and if neither EAC nor CacheX nor Feurio! say that your drive caches audio data, then you may apply the following settings:
5a. Alternative Extraction Method: Burst
Here is a controversial alternative: using burst mode for ripping. If you want to be on the safe side (not least in discussions over proper rips), always follow the above instructions for configuring EAC for ripping in secure mode and skip this alternative section.
However, it can be argued that all the re-reads that EAC performs in secure mode are not really necessary for CDs that are in good condition. In burst mode, EAC just reads the track once, which makes burst rips go faster than secure rips. Now, just one read is of course not reliable, but if you combine it with the option to "test and copy" the tracks, and use AccurateRip to boot, I believe that you can trust that your rip is as perfect a copy as it is possible to get.
"Test and copy" is explained here and here and AccurateRip results here.
If you rip a CD in burst mode it is extra important, indeed necessary, to check the log afterwards. If all the other EAC settings are correct and if there are...
- No errors
- Matching test and copy CRC values
- AccurateRip confidence at or above 2
...then you may be quite confident that your burst rip is fine.
If there are errors, mismatching CRC values, or no/low Accurate rip verification, re-rip the CD in secure mode.
Note that there is no actual error detection in burst mode; the "timing errors" that may show up may or may not indicate actual problems with the rip.
In other words, do not interpret this section as saying that you do not need to configure your drive for secure ripping as detailed above. On the contrary: I recommend that you set your drive up for secure ripping, and save your settings in the profile for FLAC rips (see 6. Saving Profiles). If and when you want to test if a CD can be properly ripped in burst mode, all you need to do is to check "Burst mode" (and click "OK") before you start ripping. You can revert to secure ripping at any time by loading the profile you have saved with all the correct settings for secure ripping.
Since AccurateRip has configured your drive settings, the top half of this window will be greyed out:
When the settings from AccurateRip are enabled, you cannot run EAC's own drive test. Therefore, you have to temporarily disable the AccurateRip setting in order to test if your drive can overread or not.
- Uncheck "Use AccurateRip with this drive", then press the button that says "Detect read sample offset correction". This is just for being able to run the overreading test, so after the test is completed, re-check "Use AccurateRip with this drive".
- Overread Lead-In and Lead-Out: Unchecked in most cases.
Check "Overread Lead-In and Lead-Out" only if the test result says that your drive can overread from both the Lead-In and Lead-Out, or if it says Lead-Out and your offset correction is positive ,or if it says Lead-In and your offset correction is negative. Otherwise disable (uncheck) it.
- Allow speed reduction during extraction: Checked.
EAC needs to be allowed to slow down during error correction.
- CD-Text Read capable drive: Checked for most drives, unless quite old.
If you want to check if your drive can read CD-Text, check this option, then (with a not-too-old CD in the drive) click the Database menu and choose "Clear Actual CD information...". (Click OK when you are warned that the information will indeed be erased.) Click the Database menu again, choose "Get CD information from... CD-Text". If no information appears, your drive can not read CD Text. You can get the information from freedb back by clicking the Database menu once again, then choose "Get CD informations from... Remote freedb".
- Remember to check "Use AccurateRip with this drive" again. The correct read offset correction will then be automatically re-entered and the top half of the window will be greyed out.
The resulting settings:
Old EAC 0.95 configuration, for informational purposes only - skip when you are setting up EAC 0.99.
- Use read sample offset correction: Checked.
Do not trust what EAC says when reporting the result of the "Detect read sample offset correction" test; it may or may not detect your drive's offset correctly.
- Check for the correct offset correction value for your drive in the AccurateRip database. (You know the name of your drive from the drop-down menu top left in EAC's main window.) You can also look for offset correction values in the DAE Drive Features Database, but it is considered somewhat less reliable than AccurateRip.
The resulting settings:
5d. Gap Detection
- Gap/Index retrieval method: Try method A, change to B or C if detection is slow, or is not working at all. You will also need to change method if the gap lengths are obviously wrong.
Sometimes the retrieval method will have to be changed depending on the CD you rip: with the Plextor drive used in most of the examples here, I can usually use any of the methods, but with some CDs only method A will work, while B and C both freeze at a certain track. This setting does not influence the ripping quality as such: A, B and C are merely different methods for retrieving indices (such as gaps). If they all work on your drive, A will usually be faster than B, which is usually faster than C.
A few drives will show slightly different gap lengths depending on which retrieval method you use. (See discussion in this Hydrogenaudio Forum thread.) Note that the gap length does not affect the audio. Properly ripped tracks will have the same CRC values even if the gap lengths differ depending of the retrieval method. But, if you want to be able to re-create and burn a perfect copy of the audio CD (or, as perfect as is possible), take the time to check if your drive detects the same gap lengths for all the methods. If it does not, use the same CD and test with another drive (if you have one). Presumably the gap lengths that match across methods and drives are the correct ones.
Occasionally, one method will give results that are obviously wrong. I have copied and slightly edited this list of what to look out for from a thread at the EAC Support Forums:
* Gaps are all the same length (apart from the first pre-gap, which may, for example, be the the standard 2 seconds while the rest of the gaps are all equal, see a log file example here). Note, however, that some retail CDs and many CD-Rs will have standard 2 second gaps between all the tracks. If they stay at 2 seconds no matter which method you use, you may conclude that all the gaps really are 2 seconds.
* Gaps lengths are all very short: 00.00.01, 00.00.03.
* Gaps lengths are very long. Over 1 minute.
* Gaps lengths get progressively longer (see a log example here, and the correct gaps in another log for the same CD here).
* Gaps lengths are equal to the length of the track.
* The total exceeds 30 seconds (2,250 sectors).
- Detection accuracy: Secure.
Change it (try "Accurate" before "Inaccurate") if the gap detection does not work at all. "Secure" and "Accurate" both retrieve the position of the gaps several times, "Inaccurate" only once.
5f. Appendix: If "Use AccurateRip with this drive" is greyed out
Note: these instructions are only for the problem with the greyed out option in EAC 0.99
- if AccurateRip works, you don't have to perform this step! Also, note that the option will be greyed out until/unless you have a CD that is in the AccurateRip key disc database
in the drive.
If you have installed EAC 0.99, but AccurateRip does not start even though you have a CD that is in the AccurateRip database in the drive, and you can't check the option "Use AccurateRip with this drive" because it is greyed out...
... you need to edit the registry. N.B. be cautious, do NOT make any other changes than the one described here. There is no "undo" option as you edit the registry, and improper changes can have severe consequences.
- Click "Start" - "Run", type "regedit" (without the quotes) and hit Enter (or click "OK").
- In the Registry Editor that thus opens, navigate to "HKEY_CURRENT_USER\Software\Illustrate\dBpowerAMP\CD-NoAccurateRip". If you click the dBpowerAMP folder in the left pane, you will see the "CD-NoAccurateRip" key in the right pane:
If "CD-NoAccurateRip" has any other value than 0 (zero), the option to use AccurateRip with EAC will be disabled, so you will have to edit it back to 0 (zero).
- Right-click "CD-NoAccurateRip", choose "Modify":
- A box for editing the value pops up. Leave it at hexadecimal and type "0" without the quotes, then click "OK":
This is what the value for "CD-NoAccurateRip" should look like when you are done:
Then just exit the Registry Editor. You may need to restart EAC in order to have the new setting recognized so that AccurateRip becomes available to you.
Fix originally found in a thread at the EAC support forums.