IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
Noob question about Accuraterip
dmdevotee
post Oct 6 2013, 10:32
Post #1





Group: Members
Posts: 18
Joined: 29-May 08
Member No.: 53893



There is any chance that Accuraterip accept as "accurate" a lossy album converted to a lossless format (source -> lossy -> lossless)?
Does Accuraterip check the audio integrity bit by bit like foobar2000 does to check if 2 audio files are exactly the same?

Thanks!
Go to the top of the page
+Quote Post
Goratrix
post Oct 6 2013, 10:46
Post #2





Group: Members
Posts: 125
Joined: 5-August 08
Member No.: 56722



No, AR does not perform any checks, and yes, it's possible to "fake" AR entries, if someone burns a CD from lossy source and then rips that CD and submits the results. Note that I put "fake" into quotes, because technically, the CD is not fake. It's just another CD with its own AR signature, and there is no reason why AR should not accept it. There are also "real" CDs made from lossy sources (the recent Robert Palmer reissues from Edsel for example).

This post has been edited by Goratrix: Oct 6 2013, 10:46
Go to the top of the page
+Quote Post
dmdevotee
post Oct 6 2013, 11:23
Post #3





Group: Members
Posts: 18
Joined: 29-May 08
Member No.: 53893



thanks!
Go to the top of the page
+Quote Post
dmdevotee
post Oct 6 2013, 16:01
Post #4





Group: Members
Posts: 18
Joined: 29-May 08
Member No.: 53893



I thinked about it , and I have one doubt

If accuraterip does not check audio integrity, why it gives you "non accurate" result when a CD is scratched enough? It isn't the track lenght the same?
Go to the top of the page
+Quote Post
db1989
post Oct 6 2013, 16:10
Post #5





Group: Super Moderator
Posts: 5275
Joined: 23-June 06
Member No.: 32180



AccurateRip does check data by hashing the audio itself, as can be verified in many parts of its documentation.

I have no idea why Goratrix thinks it does not, or what purpose s/he thinks AccurateRip would have if all it did were to compare the lengths of tracks ripped by different users, something that will most likely remain constant while giving precisely no indications of quality.
Go to the top of the page
+Quote Post
Roseval
post Oct 6 2013, 16:15
Post #6





Group: Members
Posts: 482
Joined: 26-March 08
Member No.: 52303



QUOTE (dmdevotee @ Oct 6 2013, 10:32) *
Does Accuraterip check the audio integrity bit by bit like foobar2000 does to check if 2 audio files are exactly the same?


It doesn't compare a files directly but it calculates a checksum and submit this checksum to the AR database.
This checksum changes the moment the bits are different.
That will be the case if you rip a e.g. MP3 version of the original

There is a nice article in the Wiki: http://wiki.hydrogenaudio.org/index.php?title=AccurateRip


--------------------
TheWellTemperedComputer.com
Go to the top of the page
+Quote Post
Goratrix
post Oct 6 2013, 17:37
Post #7





Group: Members
Posts: 125
Joined: 5-August 08
Member No.: 56722



QUOTE (db1989 @ Oct 6 2013, 17:10) *
AccurateRip does check data by hashing the audio itself, as can be verified in many parts of its documentation.

I have no idea why Goratrix thinks it does not, or what purpose s/he thinks AccurateRip would have if all it did were to compare the lengths of tracks ripped by different users, something that will most likely remain constant while giving precisely no indications of quality.


It does NOT check the data in the way that OP was asking about (to determine if a disc came from a lossy source). Nor does it prevent the addition of a disc to the database if said disc has different data content. It just treats the new disc as a new disc or new pressing of a existing disc. Maybe I have misunderstood OP's question, but if the question was "can AR accept and verify a disc made from lossy (mp3) sources?" then the answer is yes, as all that is needed is to create a burned disc, rip it and submit it.

This post has been edited by Goratrix: Oct 6 2013, 17:45
Go to the top of the page
+Quote Post
greynol
post Oct 6 2013, 18:23
Post #8





Group: Super Moderator
Posts: 10000
Joined: 1-April 04
From: San Francisco
Member No.: 13167



Goratrix is right.

QUOTE (Roseval @ Oct 6 2013, 08:15) *
This checksum changes the moment the bits are different.

This sounds nice and all, but it's been shown how this is definitely not the case. The algorithm isn't so broken for it to be blamed as the likely culprit for mistaking a lossily encoded rip from the original source by providing positive verification, however.

This post has been edited by greynol: Oct 6 2013, 18:33


--------------------
I should publish a list of forum idiots.
Go to the top of the page
+Quote Post
Roseval
post Oct 6 2013, 19:25
Post #9





Group: Members
Posts: 482
Joined: 26-March 08
Member No.: 52303



QUOTE (greynol @ Oct 6 2013, 18:23) *
This sounds nice and all, but it's been shown how this is definitely not the case.


Care to elaborate?


--------------------
TheWellTemperedComputer.com
Go to the top of the page
+Quote Post
EricJ2190
post Oct 6 2013, 23:31
Post #10





Group: Members
Posts: 7
Joined: 17-April 13
Member No.: 107705



QUOTE (Roseval @ Oct 6 2013, 14:25) *
QUOTE (greynol @ Oct 6 2013, 18:23) *
This sounds nice and all, but it's been shown how this is definitely not the case.


Care to elaborate?


Because of the pigeonhole principle, as long as the checksum is smaller than the audio data, there must be multiple inputs that produce the same checksum. AccurateRip uses a CRC, which is very reliable at detecting accidental errors, but it would be fairly easy to produce a CRC collision intentionally. The AccurateRip CRC checksum in particular seems to ignore some parts of the audio data entirely, meaning it would definitely fail to detect errors in those areas.

This post has been edited by EricJ2190: Oct 7 2013, 00:03
Go to the top of the page
+Quote Post
greynol
post Oct 6 2013, 23:39
Post #11





Group: Super Moderator
Posts: 10000
Joined: 1-April 04
From: San Francisco
Member No.: 13167



Are you able to defend the idea that AR checksums are cyclic?


--------------------
I should publish a list of forum idiots.
Go to the top of the page
+Quote Post
EricJ2190
post Oct 7 2013, 00:01
Post #12





Group: Members
Posts: 7
Joined: 17-April 13
Member No.: 107705



QUOTE (greynol @ Oct 6 2013, 18:39) *
Are you able to defend the idea that AR checksums are cyclic?


I was going by this post which calls it a CRC, but looking closer, the code provided does not look like a CRC to me.
Go to the top of the page
+Quote Post
greynol
post Oct 7 2013, 00:08
Post #13





Group: Super Moderator
Posts: 10000
Joined: 1-April 04
From: San Francisco
Member No.: 13167



Here:
http://www.hydrogenaudio.org/forums/index....mp;#entry756260


--------------------
I should publish a list of forum idiots.
Go to the top of the page
+Quote Post

Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 30th August 2014 - 00:18