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: MP3 origin or source (Read 14485 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

MP3 origin or source

Reply #25 –
Quote from: Buckchoi on
2Bdecided, in that instance you'd be better off finding a lossless source and verifying it with AccurateRip, then transcoding it yourself.
No he wouldn't be better off.

The idea that one can use AccurateRip to verify the legitimacy of a download is ill-conceived!  There is nothing keeping me or anyone else from burning a download to CD-R, ripping this CD-R with EAC and then submitting my results back to the AR database.

Let me give you a simple example.  I recently copied some of my new Beatles remasters to be kept in my car.  As far as I know, the titles with enhanced content are not available without the enhanced content.  As far as the AR database is concerned a disc with the enhanced content and one without are two completely different discs.

Here are the AR results for the original Sgt. Pepper's with enhanced content:
Code: [Select]
[Disc ID: 00144057-00d47100-db0fe60e]
Track	[ CRC    ] Status
 01	[911982ca] (106/223) Accurately ripped as in pressing(s) #2
 02	[8e72bae1] (107/227) Accurately ripped as in pressing(s) #2
 03	[d237e69d] (106/227) Accurately ripped as in pressing(s) #2
 04	[8b3718c1] (107/228) Accurately ripped as in pressing(s) #2
 05	[022718c4] (107/229) Accurately ripped as in pressing(s) #2
 06	[5e22ab89] (107/228) Accurately ripped as in pressing(s) #2
 07	[38ba0f32] (107/229) Accurately ripped as in pressing(s) #2
 08	[b82e0d4a] (106/228) Accurately ripped as in pressing(s) #2
 09	[4066fe35] (107/229) Accurately ripped as in pressing(s) #2
 10	[0ea487e4] (107/229) Accurately ripped as in pressing(s) #2
 11	[ea6ca871] (107/229) Accurately ripped as in pressing(s) #2
 12	[14bc7f48] (106/227) Accurately ripped as in pressing(s) #2
 13	[3f81f413] (107/225) Accurately ripped as in pressing(s) #2
Here are the AR results for my CD-R copy without the enhanced content:
Code: [Select]
[Disc ID: 0012558a-00b999ca-c7095b0d]
Track	[ CRC    ] Status
 01	[911982ca] (09/14) Accurately ripped as in pressing(s) #1
 02	[8e72bae1] (09/14) Accurately ripped as in pressing(s) #1
 03	[d237e69d] (09/14) Accurately ripped as in pressing(s) #1
 04	[8b3718c1] (09/14) Accurately ripped as in pressing(s) #1
 05	[022718c4] (09/13) Accurately ripped as in pressing(s) #1
 06	[5e22ab89] (09/13) Accurately ripped as in pressing(s) #1
 07	[38ba0f32] (09/14) Accurately ripped as in pressing(s) #1
 08	[b82e0d4a] (09/13) Accurately ripped as in pressing(s) #1
 09	[4066fe35] (09/13) Accurately ripped as in pressing(s) #1
 10	[0ea487e4] (09/13) Accurately ripped as in pressing(s) #1
 11	[ea6ca871] (09/13) Accurately ripped as in pressing(s) #1
 12	[14bc7f48] (09/13) Accurately ripped as in pressing(s) #1
 13	[3f81f413] (09/13) Accurately ripped as in pressing(s) #1

Buckchoi, maybe you can explain to me how AR got this second record.

MP3 origin or source

Reply #26 –
I would say the kind of people who would burn downloaded mp3s as audio cds, are not the kind of people who would be using EAC with AR, let alone ripping these downloads back with it.
Sure it's not a foolproof system, but it's probably more accurate than the likes of aucdtect.

According to a local record store "All the albums come with a quick-time mini doco (FIRST PRINT RUN ONLY)" which would lead me to believe issues without this enhanced content exist.

MP3 origin or source

Reply #27 –
Quote from: Buckchoi on
it's probably more accurate than the likes of aucdtect.

I don't agree.

Anyhow, the very existence of this discussion demonstrates that people do in fact rip CDs that may be sourced from lossy material.

The point being is that AR has no mechanism to filter out rips coming from CD-R once there has been more than one submission.  If you search the forum, you will see I am not the only one to have ever voiced this concern.

MP3 origin or source

Reply #28 –
Quote from: greynol on
Quote from: Buckchoi on
2Bdecided, in that instance you'd be better off finding a lossless source and verifying it with AccurateRip, then transcoding it yourself.
No he wouldn't be better off.

The idea that one can use AccurateRip to verify the legitimacy of a download is ill-conceived!  There is nothing keeping me or anyone else from burning a download to CD-R, ripping this CD-R with EAC and then submitting my results back to the AR database.

Let me give you a simple example.  I recently copied some of my new Beatles remasters to be kept in my car.  As far as I know, the titles with enhanced content are not available without the enhanced content.  As far as the AR database is concerned a disc with the enhanced content and one without are two completely different discs.

Here are the AR results for the original Sgt. Pepper's with enhanced content:
Code: [Select]
[Disc ID: 00144057-00d47100-db0fe60e]
Track	[ CRC    ] Status
 01	[911982ca] (106/223) Accurately ripped as in pressing(s) #2
 02	[8e72bae1] (107/227) Accurately ripped as in pressing(s) #2
 03	[d237e69d] (106/227) Accurately ripped as in pressing(s) #2
 04	[8b3718c1] (107/228) Accurately ripped as in pressing(s) #2
 05	[022718c4] (107/229) Accurately ripped as in pressing(s) #2
 06	[5e22ab89] (107/228) Accurately ripped as in pressing(s) #2
 07	[38ba0f32] (107/229) Accurately ripped as in pressing(s) #2
 08	[b82e0d4a] (106/228) Accurately ripped as in pressing(s) #2
 09	[4066fe35] (107/229) Accurately ripped as in pressing(s) #2
 10	[0ea487e4] (107/229) Accurately ripped as in pressing(s) #2
 11	[ea6ca871] (107/229) Accurately ripped as in pressing(s) #2
 12	[14bc7f48] (106/227) Accurately ripped as in pressing(s) #2
 13	[3f81f413] (107/225) Accurately ripped as in pressing(s) #2
Here are the AR results for my CD-R copy without the enhanced content:
Code: [Select]
[Disc ID: 0012558a-00b999ca-c7095b0d]
Track	[ CRC    ] Status
 01	[911982ca] (09/14) Accurately ripped as in pressing(s) #1
 02	[8e72bae1] (09/14) Accurately ripped as in pressing(s) #1
 03	[d237e69d] (09/14) Accurately ripped as in pressing(s) #1
 04	[8b3718c1] (09/14) Accurately ripped as in pressing(s) #1
 05	[022718c4] (09/13) Accurately ripped as in pressing(s) #1
 06	[5e22ab89] (09/13) Accurately ripped as in pressing(s) #1
 07	[38ba0f32] (09/14) Accurately ripped as in pressing(s) #1
 08	[b82e0d4a] (09/13) Accurately ripped as in pressing(s) #1
 09	[4066fe35] (09/13) Accurately ripped as in pressing(s) #1
 10	[0ea487e4] (09/13) Accurately ripped as in pressing(s) #1
 11	[ea6ca871] (09/13) Accurately ripped as in pressing(s) #1
 12	[14bc7f48] (09/13) Accurately ripped as in pressing(s) #1
 13	[3f81f413] (09/13) Accurately ripped as in pressing(s) #1

Buckchoi, maybe you can explain to me how AR got this second record.

Some of the results from my testing

Beatles - A Hard Day's Night (2009 Remaster)
13 track audio only version of the enhanced disc has 2 pressings listed in the Accuraterip db.
A total of 14 submissions on file. That would mean that 14 people made an audio only cd-r from the enhanced disc, then ripped and reported the results to Accuraterip? More interesting is that people made a cd-r as only 1 of 2 offsets. With so many different drive read and write offsets you would think there would be a dozen.

Beatles - Beatles For Sale (2009 Remaster).cue
2 pressings. A total of 13 submissions on file.

Beatles - Let It Be (2009 Remaster)
Only 2 pressings listed 14 total submitions on file.

Beatles - Please Please Me (2009 Remaster)
3 pressings listed 15 total submissions on file.

... a lot more like this

MP3 origin or source

Reply #29 –
Quote from: alondon on
That would mean that 14 people made an audio only cd-r from the enhanced disc, then ripped and reported the results to Accuraterip?
Just in case you're not joking...

The most likely explanation is that people downloaded the torrent of FLAC+cue (audio only) and checked that against Accuraterip. Maybe they checked it using Cuetools, or maybe they burnt it, gave it to someone else, and that someone else ripped it.

I do like the idea of someone downloading a lossless torrent and using Accuraterip to verify that it's correct - and in doing so, only managing to prove that it's the same as every other copy downloaded from that torrnet(!) and hence not learning anything about how accurate a copy it is of a retail disc. It's really funny.

A little knowledge is a dangerous thing.


Let's assume at least some of those Accuraterip hits are from people re-ripping burnt discs - either their "own" (!), or ones given to them by friends. In this specific case, it's a verifiable lossless download - we can know it's both lossless, accurate, and downloaded. Yet it's equally likely (more likely?) that people are doing this with lossy files.

So that means anything that's widely distributed as a set of lossy files (and that would be basically any popular CD, wouldn't it?) will turn up on Accuraterip!

Blimey. That means Accuraterip is perfect for verifying that you've downloaded, burnt, and re-ripped some mp3s in the same way as someone else(!!!!) - and absolutely useless for checking whether an unknown disc consists of
1. a perfect lossless copy of a CD
2. some downloaded mp3s, or even
3. a widely distributed should-have-been-lossless-but-has-errors-in-it rip of a CD!


I'm glad we had this discussion! I've never used Accuraterip for this purpose, but I'm happy that I think I understand it far better now.

Cheers,
David.

MP3 origin or source

Reply #30 –
Quote from: alondon on
With so many different drive read and write offsets you would think there would be a dozen.

Read offsets are not a problem since ripping with EAC configured with AccurateRip is ubiquitous.  As for different write offsets, there isn't all that much diversity between different makes and models of drives (-30, -6, 0, 6, 18 make up the vast majority of all cases).  IOW, there are plenty of people in the world who would burn the exact same disc without having to configure a write offset.  There are also plenty of people in the world who are so obsessed with accuracy, they actually do compensate for their burner's write offset.  I doubt all those guides out there telling people how to burn exact copies are being ignored; certainly not if the types of requests being made at this very forum are representative of a cross section of those who download their music rather than purchase it.  Finally, for tracks in AR records with more than one CRC, only submissions with matches make it to the part of the database that is available.

It is extremely naive to dismiss the potential for positive hits on erroneous data as a result of file sharing.

Using AR to check the authenticity of your downloads: bad idea.

MP3 origin or source

Reply #31 –
On one side there are people who want to verify their rips are as accurate as possible. On the other side people like me are not that paranoid anymore. There was a time I listened to mono 'crystal' record players and vinyl, and to mono 'metallic' transistor radios. Then came the audiocassettes and fm-radio... Now there's cd, dvd and so on.

I rip my cd's with EAC, where I once diskcopied them. I don't care about pricise gaplenghts or play tracks gaplessly. Accuraterip is nice but can be misleading. What was the source ? Rips, originals, lossywavs, lossywav-mp3's. We aren't sure anymore and could we hear or notice differences in quality or exact gaplenght, and still be happy with fm radio and mp3 V4?

MP3 origin or source

Reply #32 –
Quote from: greynol on
As for different write offsets, there isn't all that much diversity between different makes and models of drives (-30, -6, 0, 6, 18 make up the vast majority of all cases).  IOW, there are plenty of people in the world who would burn the exact same disc without having to configure a write offset.

Finally, for tracks in AR records with more than one CRC, only submissions with matches make it to the part of the database that is available.


So it’s very common for 2 cd-rs to be made with the same offset but not common enough to be done 2 times so it shows up in the DB?

Read offsets will certainly come into play when rips are made with programs that don’t offset read and write. This would make a far greater diversity of offsets.

MP3 origin or source

Reply #33 –
Wow, you really don't seem to get it.

Did I ever say that you'll never see a record from AR with multiple entries due to different offsets?  No.

The comment about the selectivity of the AR database is to demonstrate why you don't see massive amounts of single entries due to a unique combination of offsets.

I've basically laid out undeniable facts that demonstrate a likelihood for there to be consistent CD-R/RW submissions from different individuals using a common downloaded source.  If you want to continue to keep your head in the sand about this, be my guest.

MP3 origin or source

Reply #34 –

Interesting
So that people reading this thread are completely clear on this.

You are saying that a typical Accuraterip DB record of any popular CD will exist and will consist of data collected from genuine factory made discs, the ripping of cd-r made from lossless files and the ripping of user made lossy source cd-r.
The Accuraterip data version(s) made from lossy material would then be reported back to anyone testing a download from the same distributed false lossless set.

Mechanism for inclusion of lossy versions in the Accuraterip record
User 1 downloads a lossy audio set.
He/she decodes and burns them to cd-r.
The cd-r is ripped, results reported to Accuraterip.
The lossy audio is then distributed on the internet in a lossless container.

Secondary users download this false lossless rip and burn to cd-r using proper write offset.
Then these 2nd generation cd-r discs are ripped, identical hash results reported to Accuraterip.
After the database has been updated a 3rd level user downloads the same files uploaded by user 1 and when tested Accuraterip reports a confidence level of at least 1 falsely verifying the audio.
There are variants to this procedure. Is this consistent with your assessment?

Quote from: 2Bdecided on
So that means anything that's widely distributed as a set of lossy files (and that would be basically any popular CD, wouldn't it?) will turn up on Accuraterip!


Would you please quantify this numerically? Your statement above suggests that with popular artists and discs it will happen with 100% of releases. What lag time can be expected for these lossy versions? It has been suggested that with the popular Beatles disc I sited that 14 cd-r rips made from a true lossless distribution are reported back within a very short time of its release. The actual number of cd-r hash values reported to Accuraterip should be double, triple or more due to the confirmation issue. So basically in your estimation how long will it take for a lossy hash entry for this record to show up?

It would be great if you would supply an example from the Accuraterip data. Something that is “common”, by definition should be easy to find so 2 or 3 examples will do.

Thanks

MP3 origin or source

Reply #35 –
Quote from: alondon on
Mechanism for inclusion of lossy versions in the Accuraterip record
User 1 downloads a lossy audio set.
He/she decodes and burns them to cd-r.
The cd-r is ripped, results reported to Accuraterip.
The lossy audio is then distributed on the internet in a lossless container.
I have to ask (and I can't think of a way of asking politely) - are you just not very clever, or are you trolling? Not just because of this mistake (which I'll explain below), but because of your whole attitude - what is your motivation, if you're not trolling and not stupid? You can't be looking to learn - you're spending all your time trying to convince people with far more experience than you that they are wrong. Anyway...

The lossy audio doesn't need to be distributed on the internet in a lossless container.

All it takes is for several people to download the same lossy files, and make CD-Rs from them.

People ripping these CDs with an Accuraterip capable ripper are going to put them in the database.


btw, you keep mentioning read and write offsets as confounding factors to make this less likely. Write offsets are (but it's already been said there are only a few common ones) - but read offsets are irrelevant: for the software to submit a disc to the Accuraterip database, the software must already have been configured to know the correct offset, using key discs by default. If this hasn't happened, it won't be submitting anything to the database.

(I realise there are several ways of messing this up, but not by "normal" use of Accuraterip in dBpowerAMP, EAC, etc)

Quote
Quote from: 2Bdecided on
So that means anything that's widely distributed as a set of lossy files (and that would be basically any popular CD, wouldn't it?) will turn up on Accuraterip!


Would you please quantify this numerically? Your statement above suggests that with popular artists and discs it will happen with 100% of releases. What lag time can be expected for these lossy versions? It has been suggested that with the popular Beatles disc I sited that 14 cd-r rips made from a true lossless distribution are reported back within a very short time of its release. The actual number of cd-r hash values reported to Accuraterip should be double, triple or more due to the confirmation issue. So basically in your estimation how long will it take for a lossy hash entry for this record to show up?

It would be great if you would supply an example from the Accuraterip data. Something that is “common”, by definition should be easy to find so 2 or 3 examples will do.
You're the one who thinks it can't happen - you look if you're that interested.

Cheers,
David.

MP3 origin or source

Reply #36 –
Quote from: alondon on
You are saying that a typical Accuraterip DB record of any popular CD will exist and will consist of data collected from genuine factory made discs, the ripping of cd-r made from lossless files and the ripping of user made lossy source cd-r.
To my knowledge, no one has claimed this; I certainly have not.

Quote from: alondon on
The Accuraterip data version(s) made from lossy material would then be reported back to anyone testing a download from the same distributed false lossless set.
It can, yes.

Quote from: alondon on
There are variants to this procedure. Is this consistent with your assessment?
More or less, sure.

Quote from: alondon on
Quote from: 2Bdecided on
So that means anything that's widely distributed as a set of lossy files (and that would be basically any popular CD, wouldn't it?) will turn up on Accuraterip!

Would you please quantify this numerically? Your statement above suggests that with popular artists and discs it will happen with 100% of releases.
I won't speak for David, but do you think that people who regularly send batches to the AR database don't ever burn cd-r from mp3 and then rip these discs at a later date with either EAC or dBpoweramp?  Even if the database never sees another match, it doesn't make the submissions disappear unless they are later purged.  In the event that such a submission generates a new record (due to a unique disc id), the record will be available for download showing a confidence of 1.

MP3 origin or source

Reply #37 –
Quote from: alondon on
It would be great if you would supply an example from the Accuraterip data. Something that is “common”, by definition should be easy to find so 2 or 3 examples will do.

Feel free to check some of your discs with CUETools taking note of the differences between the various offsets.  Compare these differences in offsets to the write offsets I've given you and see how many records contain these same differences.  You may also see a few offset differences that match read offsets for common drives (possibly demonstrating another concern to which you hinted: someone uploads a disc ripped without a read offset correction which is then downloaded, burned and re-submitted from the copy).

Are you at all familiar with Murphy's Law?

MP3 origin or source

Reply #38 –
Quote from: 2Bdecided on
The lossy audio doesn't need to be distributed on the internet in a lossless container.
All it takes is for several people to download the same lossy files, and make CD-Rs from them.
People ripping these CDs with an Accuraterip capable ripper are going to put them in the database.


So what!
If what they rip is never shared it is never downloaded by others.
If it is never downloaded then it can never be tested with Accuraterip by others.
Files that are never tested will never produce false results. 
Quote from: 2Bdecided on
.... but read offsets are irrelevant: for the software to submit a disc to the Accuraterip database, the software must already have been configured to know the correct offset


No, think back a few steps, this is why I included a general mechanism outline for you.
When ripping the cd-r the read offset will need to be set but not when ripping the cd-r source material. Sorry but this is where the lack of a read offset comes into play.
BTW how many people do you know that burn cd-rs from mp3 files set the read and write offset or even use an offset aware program?

While on the subject of mp3, are you under the impression that all mp3 files produce the same wave file? Many mp3 versions of popular discs are shared. Even when they are labeled as the same bit rate they are often from different discs, made with different settings and a different codec. Do you also know that different decoding programs will decode identical mp3 files differently? Adding write offsets to the permutations created here may be the least of the issue.
At best, what you may have is a few users sending orphan lossy data records to Accuraterip.
Orphan records that will not be used or reported to other people verifying other rips have no detrimental effect.

Quote from: 2Bdecided on
You're the one who thinks it can't happen.


On the contrary I know it can. I have deliberately added a number of false record sets to the Accuraterip database, as well as 2 fake drives. "Can it happen" is for fools to debate. What is relevant is how often it happens, what effect it has and if so what can be done to mitigate it.

No examples? What a surprise.
You should spend less time thinking up insults and more time validating your "experience".

MP3 origin or source

Reply #39 –
Quote from: greynol on
Feel free to check some of your discs with CUETools taking note of the differences between the various offsets.  Compare these differences in offsets to the write offsets I've given you and see how many records contain these same differences.  You may also see a few offset differences that match read offsets for common drives (possibly demonstrating another concern to which you hinted: someone uploads a disc ripped without a read offset correction which is then downloaded, burned and re-submitted from the copy).


The most common offset difference is 664, but I’ve know that for years.

CUETools? Way too slow and it does not show all there is to be seen.
I’ve done this type of testing a long time looking for just what we have been talking about and then some. I have examples that indicate the existence of cd-r submissions in various disc data sets. Extrapolating that to include that the testing of files by Accuraterip is not useful and not reliable is ill conceived.

MP3 origin or source

Reply #40 –
Quote from: alondon on
When ripping the cd-r the read offset will need to be set but not when ripping the cd-r source material.

Not when the CD-R is posted in a lossless format to be downloaded by hundreds of people, it isn't.

Quote from: alondon on
Extrapolating that to include that the testing of files by Accuraterip is not useful and not reliable is ill conceived.

Please ponder the meaning of the word reliable.  Testing downloads with AccurateRip is not reliable.

MP3 origin or source

Reply #41 –
Quote from: alondon on
Do you also know that different decoding programs will decode identical mp3 files differently?
No, I never knew that!
http://mp3decoders.mp3-tech.org/
 

Teaching your grandmother to suck eggs, dear boy!


The issue under discussion can happen. It's been shown that it's happened 14 times for the latest Beatles release! Now, that might be exceptional, but it shows there's also a good chance of other popular releases getting into the Accuraterip database from downloads. So Accuraterip is not a 100% reliable way of checking that a burnt disc is correct copy of a real pressed original CD - because there are entries for downloaded+burnt versions of the disc too.

I can't see any point arguing about what percentage of releases might exist in Accuraterip in this form. If you can, go and find someone else to argue with.

Cheers,
David.

MP3 origin or source

Reply #42 –
Quote from: 2Bdecided on
The issue under discussion can happen. It's been shown that it's happened 14 times for the latest Beatles release! Now, that might be exceptional, but it shows there's also a good chance of other popular releases getting into the Accuraterip database from downloads. So Accuraterip is not a 100% reliable way of checking that a burnt disc is correct copy of a real pressed original CD - because there are entries for downloaded+burnt versions of the disc too.

I can't see any point arguing about what percentage of releases might exist in Accuraterip in this form. If you can, go and find someone else to argue with.

Cheers,
David.


The only thing proven by the Beatles example shown is that it would accurately verify the files since the hash values are identical in both data records.

Some thing "exceptional" means there is a "good chance"?
hahahahhahahhahahahhaahha
 

MP3 origin or source

Reply #43 –
Quote from: greynol on
Not when the CD-R is posted in a lossless format to be downloaded by hundreds of people, it isn't.


There are thousands of rips made with non offset correcting programs. I can show you thousands of EAC logs that have the wrong offset or zero.  I can show you EAC logs where the uploader edited the offset.
Posted in a lossless format guarantees nothing. That's why people want to check in the first place.
When you download maybe you cherry pick only shared ones with proper offset logs so your stats are badly biased.

Quote from: alondon on
Extrapolating that to include that the testing of files by Accuraterip is not useful and not reliable is ill conceived.

Quote from: greynol on
Please ponder the meaning of the word reliable.  Testing downloads with AccurateRip is not reliable.


Reliable is a relative term. You must have something to compare its results to?