AccurateRip and UltraSecure Passes

[comp1mp]

Given that C2 accuracy cannot be known, I have a few questions.

First, I understand AR is based on cd pressings. How is it possible that my AR confidence scores are not the same across all songs on an album?

Second, technically speaking, why does comparing extracted data across multiple passes increase the security of my rip? Are reads against damaged sectors volatile?

Thanks,
Matthew

[bhoar]

Re: AccurateRip and UltraSecure Passes

It is possible that someone else submitted accuraterip results with some damaged tracks (or unreliable drives) and therefore only a subset of their tracks matched. Meaning their results contribute to the counts being different. Remember, EAC and dbpa both submit AccurateRip results even if using burst mode.

As to your second question: sometimes they are volatile...and sometimes they are not. Most older "secure ripping" software just assumed that the bit errors from damage are more likely to change than not across multiple passes.

But one cannot depend upon that...so...multiple passes of the data (in this case, on a track basis) is just one tool in the toolkit. Add C2 flagging for potential problem sub-frames, plus the frame re-rip strategy...and mitigate the issue to some degree. Some have theorized that drives may be more likely to give different bit errors for damage at different ripping speeds, so perhaps use the speed variance feature as well.

Note that a "Secure" rating on a track does not provide as much confidence in the results as "Accurate". It just means that, as far as the ripper can tell, without any knowledge of what the data should be, everything appeared to rip correctly. Only "Accurate" gives you more assurance.

Granted, "Secure" with dbpa can provide *as much or perhaps more* confidence than "Secure" with other tools (assuming you have a reliable drive with good C2 support and have set it up correctly).

-brendan

[comp1mp]

Re: AccurateRip and UltraSecure Passes

Thanks Brendan.

That clears things up. It should always be noted that the whole conversation about confidence in ripping accuracy becomes moot when you treat your CDs as pristine original archives. Fortunately, I made that decision long ago. I am almost done with my first 100 discs, all with outstanding AR confidence, all at almost burst speed .

I'll be purchasing a reference license.

[Spoon]

Re: AccurateRip and UltraSecure Passes

>AR confidence scores are not the same across all songs on an album?

Also not everyone rips the whole disc, I know myself I will only rip the tracks I want.

[llumy]

Re: AccurateRip and UltraSecure Passes

I appreciate that not everyone rips the whole CD, however based on observation, there is an odd anomaly going on also. A good example is Pink Floyd "The Wall". Setting aside my disbelief that only 28 people uploaded their results to Accuraterip, my only error (song is In The Flesh) claims I don't match 19 people. The only song on the album @ 19, every other song is 28. No matter how I rip it, I am out of luck.

This is very consistent that the song that isn't "perfect" doesn't match the other AR rip # on the CD. I understand that "not everyone rips the whole CD", but I find it very odd the 28 people ripped the entire CD but this 1 song. In my experience this is consistent across CDs and various artist.

Why would one song be significantly off (in terms of # of people who ripped it)? Additionally, I have re-ripped and have had the song go from "bad" to "good" and at the same time the # of previously ripped good CRC changes. How can it be bad (and not agree with X rips), then suddenly be good but match "y" rips"?

Regards,
Llumy

[Spoon]

Re: AccurateRip and UltraSecure Passes

>Setting aside my disbelief that only 28 people uploaded their results to Accuraterip,

For that particular pressing yes, there could be 20 pressings of the same CD.

>The only song on the album @ 19, every other song is 28.

When there is a miss-match any of the confidences from any pressing is used, after all AccurateRip is track based, there will be a 28 in the database but until you get a good rip it will not know to use that 28.

[llumy]

Re: AccurateRip and UltraSecure Passes

Spoon:

Thank you - That makes sense.

Happy Holidays!

Llumy

[Porcus]

Re: AccurateRip and UltraSecure Passes

[...] A good example is Pink Floyd "The Wall". Setting aside my disbelief that only 28 people uploaded their results to Accuraterip [...]

... the latter first: there are multiple pressings of The Wall around. Some are even with pre-emphasis ...

Oddly enough, it was also Pink Floyd who released the CD I mention in http://forum.dbpoweramp.com/showthread.php?t=17563 , where there is a single track with very different AR result. It probably is a manufacturing issue, where at some point, some drives read 1 and others 0. If I understand correctly, dBpoweramp will not invoke C2 error codes if a burst rip reveals "Accurate", and -- still with the reservation for whether I got it right -- such manufacturing errors may make their ways into the database without anyone but the two (?) first attempts ever checking whether there is a manufacturing issue. (I am not sure whether re-reading could have helped, though.)

Additionally, I have re-ripped and have had the song go from "bad" to "good" and at the same time the # of previously ripped good CRC changes. How can it be bad (and not agree with X rips), then suddenly be good but match "y" rips"?

This is because the X is the number of rips disagreeing with you, and y is the number of rips who agree with you. If you re-rip and you suddenly get it right, matching 3 others, then dBpoweramp will not report that 55 others disagree, because that is likely to be a different pressing.

And besides, the database is updated about every two or three months, so an "Inaccurate" CRC may become "Accurate" just because others who agree, have made their way into the database.

[iGowanus68]

Re: AccurateRip and UltraSecure Passes

It is possible that someone else submitted accuraterip results with some damaged tracks (or unreliable drives) and therefore only a subset of their tracks matched. Meaning their results contribute to the counts being different. Remember, EAC and dbpa both submit AccurateRip results even if using burst mode.

As to your second question: sometimes they are volatile...and sometimes they are not. Most older "secure ripping" software just assumed that the bit errors from damage are more likely to change than not across multiple passes.

But one cannot depend upon that...so...multiple passes of the data (in this case, on a track basis) is just one tool in the toolkit. Add C2 flagging for potential problem sub-frames, plus the frame re-rip strategy...and mitigate the issue to some degree. Some have theorized that drives may be more likely to give different bit errors for damage at different ripping speeds, so perhaps use the speed variance feature as well.

Note that a "Secure" rating on a track does not provide as much confidence in the results as "Accurate". It just means that, as far as the ripper can tell, without any knowledge of what the data should be, everything appeared to rip correctly. Only "Accurate" gives you more assurance.

Granted, "Secure" with dbpa can provide *as much or perhaps more* confidence than "Secure" with other tools (assuming you have a reliable drive with good C2 support and have set it up correctly).

-brendan

I am new to "ripping" & newer still to dBpa. Would you tell me what "C2 flagging" is? Also, is "speed variance" a feature of C2 or dBpa? Thanks.

[garym]

Re: AccurateRip and UltraSecure Passes

I am new to "ripping" & newer still to dBpa. Would you tell me what "C2 flagging" is? Also, is "speed variance" a feature of C2 or dBpa? Thanks.

C2 is an error indication scheme built into some drives.

https://en.wikipedia.org/wiki/C2_error

Wow....resurrecting an 8 year old thread! ripping speed is a function of several things (and not directly related to either C2 or dBpa). Different disks and different CDs will rip at different drive speeds (you can see the speed reported as your ripping a CD). In the ripping options, one can set a maximum ripping speed, although I personally see no need for using this option.

[iGowanus68]

Re: AccurateRip and UltraSecure Passes

C2 is an error indication scheme built into some drives.

https://en.wikipedia.org/wiki/C2_error

Wow....resurrecting an 8 year old thread! ripping speed is a function of several things (and not directly related to either C2 or dBpa). Different disks and different CDs will rip at different drive speeds (you can see the speed reported as your ripping a CD). In the ripping options, one can set a maximum ripping speed, although I personally see no need for using this option.

Thank you.