IPB

Welcome Guest ( Log In | Register )

Understanding Rippers and Accuraterip Submissions
jayess
post Sep 10 2012, 15:48
Post #1





Group: Members
Posts: 84
Joined: 18-August 12
Member No.: 102432



I'm trying to learn more about the process of ripping and how it works in regards to Accuraterip.

"For example, audio CDs contain 16-bit, 44.1 KHz LPCM-encoded audio samples interleaved with secondary data streams and synchronization and error correction info. The ripping software tells the CD drive's firmware to read this data and parse out just the LPCM samples. The software then dumps them into a WAV or AIFF file, or feeds them to another codec to produce, for example, a FLAC or MP3 file."

http://en.wikipedia.org/wiki/Ripping

If I insert a CD from say Led Zepplin's box set of complete studio recordings, or an Elton John hybrid SACD, or a CD that's been remastered, how does a ripper and Accuraterip differentiate and correctly identify which of the recordings I have? If the selection the ripper returns in identifying the media is wrong (the media is a 2009 remaster and the ripper claims it is the original 1973 disc) and I don't change/correct it, how does that affect the database?
Go to the top of the page
+Quote Post
 
Start new topic
Replies (1 - 19)
greynol
post Sep 10 2012, 15:54
Post #2





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



AR records are based on the lengths of tracks (including data when applicable) and their start positions; nothing more.

Please search the forum in order to avoid asking redundant questions.


--------------------
I should publish a list of forum idiots.
Go to the top of the page
+Quote Post
pdq
post Sep 10 2012, 16:04
Post #3





Group: Members
Posts: 3392
Joined: 1-September 05
From: SE Pennsylvania
Member No.: 24233



There are multiple online databases of CD metadata that the ripper has access to. The accuracy of these databases is dependent on the submitters to these databases and can range from good to poor, depending on the CD. The ripper has no way of verifying the accuracy of these databases other than by comparing one against another (as dBpoweramp's ripper does).

All of this has nothing whatever to do with the AR database.
Go to the top of the page
+Quote Post
greynol
post Sep 10 2012, 16:10
Post #4





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



That wikipedia article has nothing to do with AR either.

The term interleaved reminded me that I forgot to address its misuse in a previous discussion. I won't bother with it here since it really belongs in that discussion. Unfortunately I had to close it because TOS #5 was not being honored.

This post has been edited by greynol: Sep 10 2012, 16:15


--------------------
I should publish a list of forum idiots.
Go to the top of the page
+Quote Post
Tahnru
post Sep 10 2012, 16:27
Post #5





Group: Members
Posts: 325
Joined: 17-October 05
From: United States
Member No.: 25178



HA has a very nice wiki.

http://wiki.hydrogenaudio.org/index.php?title=AccurateRip
Go to the top of the page
+Quote Post
greynol
post Sep 10 2012, 16:52
Post #6





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



Not the best article I've seen but not the worst either. Does it even address the basic types of questions such as the one specifically raised in this discussion?

Regarding some of the content as an aside, attributing cross-pressing verification to ARv2 is incorrect, as an example. There is also a lot of superfluous technical information as well as relevant but confusing technical information because it is not presented in a cogent way. Lastly, someone please tell me how an AR checksum is cyclically redundant!

This post has been edited by greynol: Sep 10 2012, 17:23


--------------------
I should publish a list of forum idiots.
Go to the top of the page
+Quote Post
mjb2006
post Sep 10 2012, 18:52
Post #7





Group: Members
Posts: 786
Joined: 12-May 06
From: Colorado, USA
Member No.: 30694



It would be quite impressive to find a CD from 1973, considering they didn't exist until the mid-1980s, but anyway... audio CDs generally don't contain IDs or metadata. When an ID is needed, such as to look up metadata (artist/title/date/etc.) or look up pressings in the AccurateRip database, an ID is generated by the software that needs it. There are different ways of generating a disc ID, but they are basically just different ways of summarizing the info encoded in the disc's Table Of Contents (TOC), which gets read as soon the disc is inserted in a player or drive. The TOC contains a list of start positions (precision: 1/75th of a second) for each track. The number of tracks and their durations can be derived from the this. As you can imagine, there are occasional collisions, but for purposes of AccurateRip, it doesn't matter; the odds of two albums having the same TOC and the same AR checksum are pretty slim! The remastered Led Zep album has a different track layout...maybe the same number of tracks, but the boundaries are not going to match up with the original editions at all.

Regarding minor errors in our wiki article on AccurateRip, although I wrote most of it, it was Spoon(!) who added the 'Pressings' section about ARv2. And I was careful to use the word checksum, but accidentally used the term CRC when linking to Spoon's explanation of the algorithm; I fixed it just now, but it's an understandable oversight since Spoon himself characterizes the checksum as a CRC. smile.gif

As for superfluous info... not to say that using the forum's search tools isn't a good idea, but I feel the wiki is there to make it so people don't have to keep searching for and referring to half-remembered posts scattered throughout a decade-plus archive of the forums. Most people aren't very good at searching; the key in this case would be to search for AccurateRip disc identification, which would lead to the following thread: http://www.hydrogenaudio.org/forums/index....showtopic=79082 (jayess, please read that thread).

The wiki is also there to explain—hopefully simplifying, without compromising—technical subjects that aren't documented very well elsewhere. In this instance, we really needed to get redundant AR and related technical info out of the EAC docs and ripping guides, and we needed to aggregate and summarize AR-related info that was scattered throughout forum postings. We could stand to move the general drive offset info to a separate article, sure, but there are only so many hours in the day. If anything, the AR article needs to be expanded to give some examples of the database in action, interpreting verification results, and answering any other questions that people have about the subject, such as jayess's question about how discs are identified.
Go to the top of the page
+Quote Post
greynol
post Sep 10 2012, 19:05
Post #8





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



I don't care who added it. From the standpoint of someone viewing the database irrespective of the limitations of any single piece of software, ARv2 was just a silly and futile patch to a half-baked checksum algorithm in order to attempt to fix a problem that was never shown to exist in the real-world. To suggest that it is required for cross-pressing verification is categorically incorrect. Long prior to the advent of ARv2, the AR database has been successfully used for cross-pressing verification. I don't think it's a good idea to rewrite history for the sake of convenience.

A good wiki article will answer the basic questions asked rather than dive into a convoluted explanation of offsets and overreading.

What does AR do?
How is AR organized?
What do AR results mean?
What are pressings according to AR and how does this differ from what pressings are in real-life?
What steps are taken to curb errant or redundant submissions as well as vandalism?

Another question on a different point:
How does a drive know where the disc ends?

This post has been edited by greynol: Sep 11 2012, 16:46


--------------------
I should publish a list of forum idiots.
Go to the top of the page
+Quote Post
jayess
post Sep 10 2012, 19:16
Post #9





Group: Members
Posts: 84
Joined: 18-August 12
Member No.: 102432



To put this into perspective, when I insert this disk:

http://www.amazon.com/Goodbye-Yellow-Brick...brick+road+sacd

I get this from 1973:

http://www.flickr.com/photos/86975222@N06/.../in/photostream
Go to the top of the page
+Quote Post
db1989
post Sep 10 2012, 19:40
Post #10





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



Wow…

Just wow.


The album was first released in 1973, but not on CD.

I am ashamed for everyone involved that I have to explain this.



As for the other chasms in your knowledge, AccurateRip does not serve metadata. Rippers use other services, such as freeDB, for that.
Go to the top of the page
+Quote Post
jayess
post Sep 10 2012, 19:46
Post #11





Group: Members
Posts: 84
Joined: 18-August 12
Member No.: 102432



QUOTE (db1989 @ Sep 10 2012, 13:40) *
Wow…

Just wow.


The album was first released in 1973, but not on CD.

I am ashamed for everyone involved that I have to explain this.



As for the other chasms in your knowledge, AccurateRip does not serve metadata. Rippers use other services, such as freeDB, for that.


Thanks, but it was remastered in 2003, and I'd rather know that date.

So what happens if I click "rip" and don't fix it?
Go to the top of the page
+Quote Post
pdq
post Sep 10 2012, 19:58
Post #12





Group: Members
Posts: 3392
Joined: 1-September 05
From: SE Pennsylvania
Member No.: 24233



Many different people submit CD metadata to the databases, and there is wide variation in how they all interpret the meaning of the different fields (as well as outright errors). Unless you check every time you are about to rip, you will often get files with tags that are not what you wanted.

For someone who has ripped as many CDs as you say you have, I am surprised that this is all news to you.
Go to the top of the page
+Quote Post
jayess
post Sep 10 2012, 20:04
Post #13





Group: Members
Posts: 84
Joined: 18-August 12
Member No.: 102432



QUOTE (pdq @ Sep 10 2012, 13:58) *
Many different people submit CD metadata to the databases, and there is wide variation in how they all interpret the meaning of the different fields (as well as outright errors). Unless you check every time you are about to rip, you will often get files with tags that are not what you wanted.

For someone who has ripped as many CDs as you say you have, I am surprised that this is all news to you.


I fix the errors before I rip and submit, but I await the wiki with these questions:

What does AR do?
How is AR organized?
What do AR results mean?
What are pressings according to AR and how does this differ from what pressings are in real-life?
What steps are taken to curb errant or redundant submissions as well as vandalism?

Another question on a different point:
How does a drive know where the disc ends?

This post has been edited by jayess: Sep 10 2012, 20:05
Go to the top of the page
+Quote Post
jayess
post Sep 10 2012, 20:19
Post #14





Group: Members
Posts: 84
Joined: 18-August 12
Member No.: 102432



QUOTE (pdq @ Sep 10 2012, 13:58) *
Many different people submit CD metadata to the databases, and there is wide variation in how they all interpret the meaning of the different fields (as well as outright errors). Unless you check every time you are about to rip, you will often get files with tags that are not what you wanted.

For someone who has ripped as many CDs as you say you have, I am surprised that this is all news to you.


On an added note, I really like the drop-down box in CueTools that has you select and match your pressing with a list of different ones created. It creates a handshake that tells me you and I both understand what disc I'm trying to rip. Inserting a disc and sending out the data without me having any clue about which pressing it's getting matched to doesn't give me that same warm, fuzzy feeling.
Go to the top of the page
+Quote Post
greynol
post Sep 10 2012, 20:23
Post #15





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



The last question was there to address an explanation about how the time of the last track on a disc is derived from what is in its TOC. I appreciate the additional information that was given in that post, however since I was slightly incorrect in my first reply as I forgot about the number of tracks.


--------------------
I should publish a list of forum idiots.
Go to the top of the page
+Quote Post
greynol
post Sep 10 2012, 20:32
Post #16





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



QUOTE (jayess @ Sep 10 2012, 12:04) *
I fix the errors before I rip and submit

Submit to what? AR doesn't keep that information; (again!) it comes from a different database. I would be surprised if CTDB was any different.

This post has been edited by greynol: Sep 10 2012, 20:32


--------------------
I should publish a list of forum idiots.
Go to the top of the page
+Quote Post
greynol
post Sep 10 2012, 20:48
Post #17





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



QUOTE (greynol)
What are pressings according to AR and how does this differ from what pressings are in real-life?

AR keeps individual records for each unique disc id. The pertinent information used to generate this id has already been given. Within each record exists one or more checksums for each track and how many times it was submitted. It is assumed that each unique checksum for any given track came from a unique pressing. In the real world, pressings of the same title can have tracks with different lengths and start times. It is even possible for the tracks to be in a different order or have a different total because there were bonus tracks or some tracks on one pressing may be combined into a larger track on another pressing. Because discs with different numbers, track lengths or start times will generate a new record in the AR database, they are not considered to be different pressings of the same title.

Ancillary cases like remasters are already covered in the above explanation. Please read it a second time before asking about them. wink.gif

EDIT: The key point here is that AR submissions are organized into different records (records as in a dental or criminal, not records as in sides one and two of "The Song Remains the Same" on vinyl). Each record may contain information for multiple "pressings" (scary quotes intentional; they may or may not reflect different pressings in the real world) and will do so on a track by track basis.

This post has been edited by greynol: Sep 11 2012, 15:02


--------------------
I should publish a list of forum idiots.
Go to the top of the page
+Quote Post
godrick
post Sep 10 2012, 22:29
Post #18





Group: Members
Posts: 306
Joined: 31-December 10
Member No.: 86948



mjb2006, I heartily agree a single reference wiki page for all things AccurateRip is helpful, and I've been to that wiki page many times, so thanks for your efforts. I also agree with Greynol that it could be much improved by taking a top-down approach of starting with the basic intent and structure and then drilling down. An FAQ approach (again, including all of Greynol's specific questions) could be very appropriate. If you can make the time for such an improvement effort, I think many would appreciate it.

Building on Greynol's question of "what do AR results mean", to anticipate obvious and frequently-discussed questions in HA and elsewhere, the specifics of results from the major AR checkers should be described, especially dBpoweramp, EAC and CUETtools, without getting into controversies of what is really needed or which approach is better. I could easily be wrong, but think that CUETools checks ARv1 and ARv2 but does cross pressing checks on ARv1 values only, while the latest version of dBpoweramp will first look for cross pressing verification using ARv2 and if found stops there and if not then looks at ARv1 including a cross pressing check. Whether my understanding is correct or not (or perhaps especially if it is not), I think clarifying this with examples in a wiki would be of considerable value.
Go to the top of the page
+Quote Post
jayess
post Sep 11 2012, 13:49
Post #19





Group: Members
Posts: 84
Joined: 18-August 12
Member No.: 102432



This stuff does make you go h'mmm. I find out I've got two copies of another disc. I start ripping the first last night and go to bed, because Accurate rip says it's not in the database and rips its secure. This morning I rip the second copy and it gets mainly 6's and the CRC's exactly match the first disc and remain green to indicate they did match.

And I had some brand new CD's arrive yesterday and witnessed another strange one I see from time to time; you put the disc in and it reports being in Accuraterip before you start ripping, then as soon as you do start ripping every single tracks gets ripped secure to lead you to believe it wasn't actually a match to anything.
Go to the top of the page
+Quote Post
pdq
post Sep 11 2012, 14:22
Post #20





Group: Members
Posts: 3392
Joined: 1-September 05
From: SE Pennsylvania
Member No.: 24233



When you first put the CD in, it reads the TOC and generates the ID that matches one or more entries in the AR database. At this point the offset has not been determined, so you don't know if your pressing is in the database.
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: 28th August 2014 - 04:34