Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: [EAC] Test & Copy, Burst vs. Secure and Accuracy (Read 18536 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

[EAC] Test & Copy, Burst vs. Secure and Accuracy

I use AccurateRip where available, so that's a given. I also don't really mind 30-40 minutes per CD rip time. I also convert to FLAC straight from EAC during extraction. Now..

1. Is secure mode Test & Copy the best for accuracy?

2. How much better is it than plain secure Copy?

3. Secure Copy + a separate Test [still in secure mode] to check CRCs is the same thing as T&C, right?

4. Would secure Copy + a separate Test in burst mode be better? Or how about burst Copy and secure Test?


I kinda gave up on using burst mode T&C alone, because the last time I did I got an inaccurate rip (confidence 200) whereas with secure mode it was all accurate.

[EAC] Test & Copy, Burst vs. Secure and Accuracy

Reply #1 –
1. No, not necessarily.  

2. Test & Copy is redundant when using secure mode because secure mode already reads each sector at least twice. If you get different CRCs with T&C secure then something is seriously wrong and you shouldn't trust anything at that point. Try a different drive/extraction mode/ripping software.

3. Yes.

4. Yes, since it would combine different extraction modes for additional confidence.




Don't give up on burst mode because of the occasional problem rip, read this:

http://www.digital-inn.de/129988-post4.html

http://www.digital-inn.de/exact-audio-copy...ive-answer.html (the second poster, Andre, is EAC's creator)

http://www.hydrogenaudio.org/forums/index....mp;#entry715370


[EAC] Test & Copy, Burst vs. Secure and Accuracy

Reply #2 –
Whether it be in burst mode or secure mode [which is nothing more than burst reads that are repeated (depending on whether EAC is configured to use C2 pointers) and occasionally partially overlapped], T&C is nothing more than a demonstration of precision.  It does not provide provide any guarantee of accuracy.

Quote
Secure Copy + a separate Test [still in secure mode] to check CRCs is the same thing as T&C, right?
Yes and being in secure mode does not matter; configuring EAC's ripping mode is completely separate from telling EAC to rip to  your drive (F5), perform a test rip in order to generate a second CRC that is not written to the drive (F8), or do both one after the other (F6).

Quote
inaccurate rip (confidence 200)
When you see a rip labeled inaccurate, the only confidences that have meaning are those that are attached to tracks that are verified as accurate.

Regarding this idea generating one CRC in burst mode and another in secure mode, the point is to help catch errors that may have otherwise gone undetected.  When ripping this way and getting matching CRCs, the result provides a higher degree of confidence than just ripping in burst mode and getting matching CRCs which provides a higher degree of confidence than getting matching CRCs using only secure mode (without C2 pointers).  The highest degree of confidence that a rip is Accurate is one that is verified by AccurateRip, regardless of whether T&C is used, provided that the results came from someone else's disc.

Please take some time and read some of the discussions in this sub-forum.  You will find the answers you're looking for.

[EAC] Test & Copy, Burst vs. Secure and Accuracy

Reply #3 –
Quote from: Cynic on
Test & Copy is redundant when using secure mode because secure mode already reads each sector at least twice.
This is only true when C2 pointers aren't being used, otherwise double reading of sectors is only done periodically in order to help detect possible issues with synchronization.

Quote from: Cynic on
If you get different CRCs with T&C secure then something is seriously wrong and you shouldn't trust anything at that point.
You can still trust tracks that are verified by AccurateRip.

Thank you for digging up those posts.  In his research, hopefully Brand will find others like them.

[EAC] Test & Copy, Burst vs. Secure and Accuracy

Reply #4 –
OK, I guess I'll stick to matching the CRCs from both secure and burst modes. (Besides AccurateRip, of course.)
Seems fast enough for me and still better than just one of the modes.

I've read some of the older discussions and some were relatively old (2004, 2005..), so I wanted to make sure that the info still applies today.

What got me a bit confused is that while many people (including EAC's author in that post) use burst mode T&C, EAC (through its wizard) still recommends secure mode for accuracy.
Now maybe this is just a matter of burst T&C not being as well integrated as a setting in the UI, but it's a bit confusing.

Or, the way I see it now, burst is perfectly OK if there's AccurateRip to confirm the rip.


Quote from: greynol on
Quote
inaccurate rip (confidence 200)
When you see a rip labeled inaccurate, the only confidences that have meaning are those that are attached to tracks that are verified as accurate.

So, in other words: the only inaccurate track reports that have meaning are those where other tracks on the CD are reported as accurate?
(I know about different pressings and therefore the whole CD being "inaccurate".)

[EAC] Test & Copy, Burst vs. Secure and Accuracy

Reply #5 –
Quote from: greynol on
being in secure mode does not matter; configuring EAC's ripping mode is completely separate from telling EAC to rip to  your drive (F5), perform a test rip in order to generate a second CRC that is not written to the drive (F8), or do both one after the other (F6).

Maybe I misunderstood what you wrote but, just to make it clear...
From what I've seen the Testing of tracks (F8) will definitely be different (different speed) depending on whether the ripping mode is set to secure or burst.
Or did you mean something else?

[EAC] Test & Copy, Burst vs. Secure and Accuracy

Reply #6 –
Quote from: Brand on
Now maybe this is just a matter of burst T&C not being as well integrated as a setting in the UI, but it's a bit confusing.
T&C can be found under the Action menu.  The ripping mode is selected in the Extraction Method tab in the Drive Options (F8).  They have nothing to do with one another.  Burst T&C just means T&C done in burst mode.  Secure T&C just means T&C done in secure mode.

Quote from: Brand on
So, in other words: the only inaccurate track reports that have meaning are those where other tracks on the CD are reported as accurate?
Yes.  I worked with Andre to get the word "inaccurate" removed from the EAC vernacular.  As of V0.99pb2, you will not see it used.  "Cannot be verified" is a more accurate description that is never misleading.  I regret that the confidence of non-verified tracks is still being reported, but that's just the way it goes sometimes.

[EAC] Test & Copy, Burst vs. Secure and Accuracy

Reply #7 –
Quote from: greynol on
Quote
Secure Copy + a separate Test [still in secure mode] to check CRCs is the same thing as T&C, right?
Yes and being in secure mode does not matter; configuring EAC's ripping mode is completely separate from telling EAC to rip to  your drive (F5), perform a test rip in order to generate a second CRC that is not written to the drive (F8), or do both one after the other (F6).

You have the ability to change the mode in between copying and testing when you perform them as separate steps, whereas you don't have this ability when telling EAC to perform test and copy as a single operation; it will use whichever mode is configured at the time.

[EAC] Test & Copy, Burst vs. Secure and Accuracy

Reply #8 –
Quote from: Brand on
I use AccurateRip where available, so that's a given. I also don't really mind 30-40 minutes per CD rip time. I also convert to FLAC straight from EAC during extraction. Now..

1. Is secure mode Test & Copy the best for accuracy?

2. How much better is it than plain secure Copy?

3. Secure Copy + a separate Test [still in secure mode] to check CRCs is the same thing as T&C, right?

4. Would secure Copy + a separate Test in burst mode be better? Or how about burst Copy and secure Test?


I kinda gave up on using burst mode T&C alone, because the last time I did I got an inaccurate rip (confidence 200) whereas with secure mode it was all accurate.



I used to be a real settings Nazi until I started doing some thinking and testing on my own. The most important thing with EAC is to be in secure ripping Mode and have your read offset configured. Either manually or by using Accurate rip (easiest)

I'm going to tell you this from experience. There is no reason to test AND copy a brand new CD. There is also no reason to defeat audio cache on a drive that doesn't cache audio. I've never had a drive that cached audio, checked with cachex and feurio. There's no reason for EAC to be copying scratch free discs at 5-7x speed. You should be able to set it up to copy at 15-20x speed reliably and quickly.  All I do anymore is just a quick image copy, then I use Cue Tools to re-encode the wav image into Flac tracks with a non compliant gaps appended cue sheet.

I just know too much now to spend an hour ripping a disc and wearing out another burner and ripping with EAC will wear out your burner too. I'm not saying I'm expert but it is what it is. Check out my new thread on debunking EAC myths.

[EAC] Test & Copy, Burst vs. Secure and Accuracy

Reply #9 –
Quote from: RendoR on
There is no reason to test AND copy a brand new CD.

One can make a similar case suggesting that secure mode is not necessary.

[EAC] Test & Copy, Burst vs. Secure and Accuracy

Reply #10 –
A few more questions:

1. What is the chance of a ripped track being different than another, even tho their AccurateRip CRCs (or what they are called?) are the same? In other words: how thoroughly does AccurateRip check a file? Does it check the entire file?

2. If I rip a CD with EAC, submit that info to the AccurateRip database, then - at a later date - rip that CD again, will it check the accuracy against my own previous rip? Can I see any info (like date) of the entry(s) that the rip is being checked against?

3. Is there a way to tell how many AccurateRip entries there are for a CD, before ripping it?

[EAC] Test & Copy, Burst vs. Secure and Accuracy

Reply #11 –
Quote from: Brand on
2. If I rip a CD with EAC, submit that info to the AccurateRip database, then - at a later date - rip that CD again, will it check the accuracy against my own previous rip?

If you rerip on the same system as the first time then no, it will not check against your previous rip.

[EAC] Test & Copy, Burst vs. Secure and Accuracy

Reply #12 –
I don't think that's right.  It will simply prevent you from submitting it a second time.

For what you say to be true, AR records would have to contain user IDs.  I'm pretty certain that they do not.  Do you have a reference that states otherwise?

[EAC] Test & Copy, Burst vs. Secure and Accuracy

Reply #13 –
Quote from: greynol on
Quote from: RendoR on
There is no reason to test AND copy a brand new CD.

One can make a similar case suggesting that secure mode is not necessary.



I read this topic and now i'm confused about secure rip with EAC.

In the Drive option, I cheched "Secure Mode" and "Accurate Stream" feature but no drive caches audio data or C2 error information. The offset correction of my drive is configured and i always use "AccurateRip".

When I rip my cd (most them are brand new), I has always used "Test & Copy". But in this topic, you say that T&C is useless on brand new cd ? Is it that means that all I do is wearing out my own burner for nothing ?
Windows 10 Pro
fb2k v.1.3.16

[EAC] Test & Copy, Burst vs. Secure and Accuracy

Reply #14 –
Quote from: Brand on
1. What is the chance of a ripped track being different than another, even tho their AccurateRip CRCs (or what they are called?) are the same? In other words: how thoroughly does AccurateRip check a file? Does it check the entire file?

I don't think anyone will be able to give you a concrete answer.  IIRC, 3% of the data is not covered by the original version of AR.  The way the 3% is distributed may still be quite robust against typical ripping errors.  The new version of AR (AR2) uses a more robust checksum algorithm, which should cover all the data except the edges of the disc to compensate for drives that require different offset corrections which cannot overread (as is also the case with the original version of AR).

Quote from: Brand on
2. If I rip a CD with EAC, submit that info to the AccurateRip database, then - at a later date - rip that CD again, will it check the accuracy against my own previous rip?
I believe the correct answer is yes.

Quote from: Brand on
Can I see any info (like date) of the entry(s) that the rip is being checked against?
No.

Quote from: Brand on
3. Is there a way to tell how many AccurateRip entries there are for a CD, before ripping it?
With EAC or dBpoweramp the answer is no, though it could easily be possible.  An AR disc ID is generated from your disc's table of contents.  Using this disc ID, a record is downloaded from the database from which the number of entries can easily be counted.

[EAC] Test & Copy, Burst vs. Secure and Accuracy

Reply #15 –
Quote from: Yuna on
Is it that means that all I do is wearing out my own burner for nothing ?

Pretty much, yes.  For discs in excellent condition, a simple burst rip is usually adequate.  If the burst rip is verified by AR then it is secure.

When you use secure mode without C2 pointers, everything is read at least twice.  When you use T&C, everything is now read four times.  If it turns out that the first read was correct, then all the rest were unnecessary.  Of course you won't have any confidence that the first read was correct unless you have AR information or have read the data another time.

The problem is that there are no definitive answers.  T&C in secure mode can sometimes catch errors that would have otherwise gone unreported, but it is also true that T&C can match even though there are errors, so what can you do?

[EAC] Test & Copy, Burst vs. Secure and Accuracy

Reply #16 –
Quote from: pdq on
Quote from: Brand on
2. If I rip a CD with EAC, submit that info to the AccurateRip database, then - at a later date - rip that CD again, will it check the accuracy against my own previous rip?

If you rerip on the same system as the first time then no, it will not check against your previous rip.

What counts as same system?

Same hardware or same hardware+OS? Or even same hardware+OS+same EAC installation?

In my case, between the two rips I reinstalled Windows on the same hardware, copied the EAC config files and, IIRC, updated the DVD drive's firmware.. And I'm not sure if the one and only AccurateRip entry is the one from my own rip from a few weeks ago or somebody else's.


Quote from: greynol on
it could easily be possible.  An AR disc ID is generated from your disc's table of contents.  Using this disc ID, a record is downloaded from the database from which the number of entries can easily be counted.

Yeah, that makes sense to me and would be great to have in EAC.


For those wondering about this: the number of submitted CDs could be helpful for rare CDs, when deciding on whether to go for a slower/more secure method or for a fast extraction that can be checked against many AR entries.

[EAC] Test & Copy, Burst vs. Secure and Accuracy

Reply #18 –
Good idea!

[EAC] Test & Copy, Burst vs. Secure and Accuracy

Reply #19 –
Quote from: greynol on
I don't think that's right.  It will simply prevent you from submitting it a second time.

For what you say to be true, AR records would have to contain user IDs.  I'm pretty certain that they do not.  Do you have a reference that states otherwise?

My mistake. Initially each submission has a unique identifier to, as you say, protect against multiple submissions. Once there are multiple matching submissions with multiple identifiers then those are combined and the identifiers are lost.

[EAC] Test & Copy, Burst vs. Secure and Accuracy

Reply #20 –
Everyone who submits to AccurateRip has a unqiue ID, if a track is submitted again later on the newer rip replaces the older one. If a user (perhaps 3 months later after they have been submitting) is found to have a drive with an incorrect offset, they are removed and all their submissions are removed.

The above is in the offline database, the online database is made up from combining and removing all the IDs.

Quote
I don't think anyone will be able to give you a concrete answer. IIRC, 3% of the data is not covered by the original version of AR. The way the 3% is distributed may still be quite robust against typical ripping errors. The new version of AR (AR2) uses a more robust checksum algorithm, which should cover all the data except the edges of the disc to compensate for drives that require different offset corrections which cannot overread (as is also the case with the original version of AR).


In theory even AR1 can be improved by also checking the different offsets for pressings, as these offsets are well offset the missing data might not be missing at that exact point, if there were 4 offset matches that for me would strengthen the crc.

[EAC] Test & Copy, Burst vs. Secure and Accuracy

Reply #21 –
Just one alternate offset would essentially remove any and all concerns about collisions resulting from data that was not covered by the AR1 hash.  FWIW, the announcement made about the hash calculation never caused me to worry.