PDA

View Full Version : Secure mode and Accuraterip question



Veovis
10-01-2015, 04:10 AM
A couple of OCD audiophile questions...;)

1. When ripping in secure mode but with Accuraterip turned off, does the ripper use CRC32 to compare the different readings of the CD and determine whether or not they are secure?
2. Do the CRC32 calculation include every single frame read by the CD-drive?
3. How common is it that the few frames at the beginning and end of a CD that Accuraterip ignores actually contain audio and not only silence?

I'm asking mostly because I'm curious, but I have been thinking about changing my ripping work flow to use only the ultra secure mode and then check the rips against the AccurateRip database with PerfectTunes afterwards. In theory that should be more "secure" than using only Accuraterip to decide whether or not a rip is accurate, right?

Spoon
10-01-2015, 04:44 AM
1. Yes, but this method is no where near 100% as there are consistent errors
2. Yes
3. Not that common

Veovis
10-01-2015, 05:16 AM
Thanks!

garym
10-01-2015, 06:39 AM
I'm asking mostly because I'm curious, but I have been thinking about changing my ripping work flow to use only the ultra secure mode and then check the rips against the AccurateRip database with PerfectTunes afterwards. In theory that should be more "secure" than using only Accuraterip to decide whether or not a rip is accurate, right?

I disagree with your statement. Doing your own multiple rips (same physical CD, same drive, etc.) can never be as good as the evidence provided by even a single AccurateRip match (assuming that single match is not a previous rip of the CD reported by you). Do you realize that the beauty of AR is that an AR match (of even 1) is equivalent to you buying two copies of the CD, ripping each CD on two different drives, then comparing the resulting rips for bit by bit match. And of course a match of > 1 is the same as above mulitipled by X (number of AR matches).

Veovis
10-01-2015, 07:00 AM
I disagree with your statement. Doing your own multiple rips (same physical CD, same drive, etc.) can never be as good as the evidence provided by even a single AccurateRip match (assuming that single match is not a previous rip of the CD reported by you). Do you realize that the beauty of AR is that an AR match (of even 1) is equivalent to you buying two copies of the CD, ripping each CD on two different drives, then comparing the resulting rips for bit by bit match. And of course a match of > 1 is the same as above mulitipled by X (number of AR matches).

I am aware of the beauty of AR and wouldn't dream of taking it out of the process altogether. My idea is to rip in ultra-secure mode and then AR-control the result with PerfectTunes. If I understand correctly that would at least mean a bigger chance (but not 100% certainty) that the first and last frames of the CD, in the event they actually contain audio, are accurate, compared with only relying on Accuraterip in the ripping process. Sort of the best of two worlds. With the downside that ripping will be more time consuming and put more workload on the drive of course.

I warned you the questions was a bit OCD...:)

garym
10-01-2015, 07:06 AM
I am aware of the beauty of AR and wouldn't dream of taking it out of the process altogether. My idea is to rip in ultra-secure mode and then AR-control the result with PerfectTunes. If I understand correctly that would at least mean a bigger chance (but not 100% certainty) that the first and last frames of the CD, in the event they actually contain audio, are accurate, compared with only relying on Accuraterip in the ripping process. Sort of the best of two worlds. With the downside that ripping will be more time consuming and put more workload on the drive of course.

I warned you the questions was a bit OCD...:)

OK. I see now that you'd rip with ultrasecure, with AR turned OFF, then compare resulting rip to AR database with PerfectTunes. Yes, that would certainly work. And I'd say it is more than a "bit" OCD....more like major OCD. But then again virtually everyone using dbpa (including me) is at least a bit OCD regarding their digital music! :)

Veovis
10-01-2015, 07:14 AM
OK. I see now that you'd rip with ultrasecure, with AR turned OFF, then compare resulting rip to AR database with PerfectTunes. Yes, that would certainly work. And I'd say it is more than a "bit" OCD....more like major OCD. But then again virtually everyone using dbpa (including me) is at least a bit OCD regarding their digital music! :)

Ok. Major OCD. I admit it.:)

schmidj
10-01-2015, 12:19 PM
Ripping in ultrasecure without Accuraterip and then checking in PerfectTunes will also slow the ripping down significantly for any CD in Accuraterip. Instead of ripping once presumably at a high speed, if you have your secure set up as two fast rips, then at least one variable speed "ultrasecure" rip, you are now ripping the CD three times, once relatively slowly, instead of once.

From my experience, a clean rip of a CD in Accuraterip takes 3 to 5 minutes. An "ultrasecure" rip of a CD not in Accuraterip takes about 15 minutes.

Veovis
10-02-2015, 06:48 AM
Ripping in ultrasecure without Accuraterip and then checking in PerfectTunes will also slow the ripping down significantly for any CD in Accuraterip. Instead of ripping once presumably at a high speed, if you have your secure set up as two fast rips, then at least one variable speed "ultrasecure" rip, you are now ripping the CD three times, once relatively slowly, instead of once.

From my experience, a clean rip of a CD in Accuraterip takes 3 to 5 minutes. An "ultrasecure" rip of a CD not in Accuraterip takes about 15 minutes.

Yes, I am aware of this. I was asking more out of curiousity of how dbpa works. But maybe there should be an option to actually rip a couple of passes first and let dbpa check the rip against AR as the last step for those who think it's worth the extra time? Kind of how EAC does it a guess.

Spoon
10-02-2015, 07:41 AM
Each pass ripped is checked against AccurateRip, even cross checking bad frame sections from each pass (interchanging). EAC is no where near as advanced in this area.

PepsiCan
10-02-2015, 05:01 PM
Yes, I am aware of this. I was asking more out of curiousity of how dbpa works. But maybe there should be an option to actually rip a couple of passes first and let dbpa check the rip against AR as the last step for those who think it's worth the extra time? Kind of how EAC does it a guess.

Then here is another one for you on how DBPA works.
1) Rip track
2) Compare to AccurateRip DB
3) Apply DSP effects

--> and some DSP effects alter the bits in the file, meaning you will lose any link to AccurateRip as the checksum of your rip does not compare. In my case, this happened with the DSP effects to delete silences between tracks. I did that to isolate any hidden tracks better. But all these CDs subsequently did not match with AccurateRip anymore when I checked them in PerfectTunes. If you have the AccurateRip checksum in your metadata, at least you know the file was ripped accurate. But using ultra secure won't give you that checksum and therefore you'll have to do a lot of exploring to figure out why PerfectTunes tells you the file is not accurate. So, if you go this route, pick your DSP with care.

garym
10-02-2015, 05:37 PM
Then here is another one for you on how DBPA works.
1) Rip track
2) Compare to AccurateRip DB
3) Apply DSP effects

--> and some DSP effects alter the bits in the file, meaning you will lose any link to AccurateRip as the checksum of your rip does not compare. In my case, this happened with the DSP effects to delete silences between tracks. I did that to isolate any hidden tracks better. But all these CDs subsequently did not match with AccurateRip anymore when I checked them in PerfectTunes. If you have the AccurateRip checksum in your metadata, at least you know the file was ripped accurate. But using ultra secure won't give you that checksum and therefore you'll have to do a lot of exploring to figure out why PerfectTunes tells you the file is not accurate. So, if you go this route, pick your DSP with care.

If I were to use any DSPs that modify the actual audio bits (remove silence, volume normalize, RG(Apply), HDCD, etc.) I'd do this to a separate copy of my lossless files. I want my archive version of CD rips to be bit perfect back to the original CD.

Veovis
10-06-2015, 06:25 AM
Each pass ripped is checked against AccurateRip, even cross checking bad frame sections from each pass (interchanging). EAC is no where near as advanced in this area.

I understand.

I've actually started to use the following workflow with dbpa when ripping CD:s:

1. Uncheck the first and last tracks and rip with AccurateRip activated.
2. Invert the checked tracks and rip the first and last track without AccurateRip. Ultra-secure with variable drive speed and write file after two matching passes.
2,5. If problems with the first and/or last tracks; check CRC32 with another drive (rarely needed)
3. Check complete album with PerfetcTunes.

Doesn't take that much longer and I usually spend more time fixing meta data and finding the best album art than ripping anyway, so this works for me. Should mean a another layer of "security", which is probably not really necessary but cannot hurt, so...

mville
10-06-2015, 07:02 AM
3. How common is it that the few frames at the beginning and end of a CD that Accuraterip ignores actually contain audio and not only silence?

Isn't this an issue with DAE software and optical drives rather than that of AccuraterRip, so may occur regardless of whether AccurateRip is ON or OFF?

Veovis
10-06-2015, 08:05 AM
Isn't this an issue with DAE software and optical drives rather than that of AccuraterRip, so may occur regardless of whether AccurateRip is ON or OFF?

Others can give you a precise technical explanation why, but as far as I understand it Accuraterip ignores (by design) a tiny amount of data at the beginning and end of a CD. The Cuetools equivalent (CTDB) does the same.

Edit: The above doesn't mean that the data isn't read, only that it is not checked for accuracy against the AR database. Whether or not Accuraterip is on or off does not affect how the drive itself reads the CD.

mville
10-06-2015, 09:01 AM
Others can give you a precise technical explanation why, but as far as I understand it Accuraterip ignores (by design) a tiny amount of data at the beginning and end of a CD. The Cuetools equivalent (CTDB) does the same.

Edit: The above doesn't mean that the data isn't read, only that it is not checked for accuracy against the AR database. Whether or not Accuraterip is on or off does not affect how the drive itself reads the CD.

So, from what you have said and what I understand, there are potentially two separate issues here.

For a reason I am unsure about, a few frames of the first and last tracks are not checked against the AccurateRip database, so these tracks may be inaccurate,

AND

due to optical drive technology, data may not be read from the first and last tracks and will be padded with digital silence. As such these tracks may also be inaccurate?

Veovis
10-07-2015, 11:00 AM
So, from what you have said and what I understand, there are potentially two separate issues here.

For a reason I am unsure about, a few frames of the first and last tracks are not checked against the AccurateRip database, so these tracks may be inaccurate,

AND

due to optical drive technology, data may not be read from the first and last tracks and will be padded with digital silence. As such these tracks may also be inaccurate?

The first statement is true. As for the second one I'm not quite sure, but I think all data on the CD is read apart from, if your drive does not support it, the lead-in and lead-out area of the CD (and which should only consist of silence anyway). Another factor is the off-set of different drives and I think it's because of this fact that AR ignores the first and last (very few) frames of the CD.

There's a wiki about these things if you google it. And someone else here should be able to correct or elaborate.