Which offset detection is more accurate? I get different results using EAC integrated detector and AccurateRip detector.

I believe the AccurateRip may be off by 30 samples according to another thread here. That being said, its the standard. Also, EAC and AR should give the same results. The reason AR is off is it was based on EAC which is apparently off!?

Yes! I see 30 samples difference.

However there are three types of offset correction: 'combined', 'write' and 'read'.

* When I use a CD listed in EAC database I get +1167.
* Write offset of my drive is -511
* When I create an Offset Checking CD (using EAC) and test it I get 656 (combined offset = [+1167] - [-511] = [656]
* AccurateRip shows me 686 (686 - 30 = 656).

Anyway, if we even close eyes on 30 samples difference, offsets we get using AccurateRip are 'COMBINED' offsets = WRONG offsets.

Am I wrong?

First of all, that "off by 30 samples" thing is completely off-topic and should not even have been brought into this conversation. AccurateRip uses the same reference as EAC, and I think the 30 sample difference you mention is only a coincidence due to error.


AccurateRip most certainly does not give combined offsets. But people often confuse read offset correction with combined offset correction when performing tests with EAC.

What is your drive? (The easiest thing to do would be to find known results.)

My drive is TEAC E-540W (burner). And I'am pretty sure I know my drive both combined and read offsets. Look - there is create 'offset test CD' feature in EAC. Which type of offset should I get with that test CD? I see no way to get 'read' and 'write' offset values separately using test CD on one drive. And the test CD brings me +656.

If you wrote the test CD with "0" entered for "write samples offset", the test result is your combined read/write offset.

There are many drives in the AccurateRip and daefeatures.co.uk db`s that show a read offset correction of +686 (including some TEAC drives, though there is no "E-540W" listed). None that show +656. This isn't conclusive, but I think the evidence leans heavily toward +686 being the correct value.

Unless you have a few CDs that return consistent results your finding here is most likely unreliable.

Cosmo is right, this 30 sample skew shouldn't have been mentioned here as it will only help to confuse. If Accuraterip reports 686 this will almost certainly be the correct result.

everux and Cosmo are right about making mention of the new supposed "correct" reference.

Back to the original question, between AccurateRip and EAC which is more accurate at determining the offset, the answer is that they are both accurate and both produce the same result if done correctly. The major difference is that AccurateRip has a much larger database of discs to use. The EAC database gives you exact pressing information to follow, however. This is an advantage over AccurateRip since many of the discs in the AR database have multiple pressings which cannot be used to measure drive models which don't have entries in the database.

Regarding the +1167 measurement, did you make sure that you matched the pressing information with that on your CD?

http://users.pandora.be/satcp/eacoffsets02.htm#-

Also, the equation you used is not correct.

Combined read/write sample offset correction =
Read samples offset correction + write samples offset

But you're really suppoed to use the equation this way:
Write samples offset = Combined offset correction - Read offset correction

You measure the combined using the test disc burned with a write samples offset of 0 (just like what Cosmo said). You determine the Read offset correction using either AccurateRip or EAC's test. Using these two numbers, you calculate the write samples offset.

http://users.pandora.be/satcp/eacoffsets03.htm#-