Hydrogenaudio Forums > using auCDtect with individual or single file

Full Version: using auCDtect with individual or single file

Hydrogenaudio Forums > Hydrogenaudio Forum > General Audio

Jan S.

Sep 9 2004, 02:42

First I analysed a bootleg cd track by track:

QUOTE

------------------------------------------------------------
Processing file: [01.wav]

------------------------------------------------------------
This track looks like MPEG with probability 89%

------------------------------------------------------------
Processing file: [02.wav]

------------------------------------------------------------
This track looks like MPEG with probability 89%

------------------------------------------------------------
Processing file: [03.wav]

------------------------------------------------------------
This track looks like CDDA with probability 100%

------------------------------------------------------------
Processing file: [04.wav]

------------------------------------------------------------
This track looks like MPEG with probability 90%

------------------------------------------------------------
Processing file: [05.wav]

------------------------------------------------------------
This track looks like CDDA with probability 100%

------------------------------------------------------------
Processing file: [06..wav]

------------------------------------------------------------
This track looks like CDDA with probability 100%

------------------------------------------------------------
Processing file: [07..wav]

------------------------------------------------------------
This track looks like CDDA with probability 100%

------------------------------------------------------------
Processing file: [08..wav]

------------------------------------------------------------
This track looks like CDDA with probability 44%

------------------------------------------------------------
Processing file: [09..wav]

------------------------------------------------------------
This track looks like CDDA with probability 100%

------------------------------------------------------------
Processing file: [10.wav]

------------------------------------------------------------
This track looks like MPEG with probability 90%

------------------------------------------------------------
Processing file: [11.wav]

------------------------------------------------------------
This track looks like CDDA with probability 100%

------------------------------------------------------------
Processing file: [12.wav]

------------------------------------------------------------
This track looks like MPEG with probability 43%

------------------------------------------------------------
Processing file: [13.wav]

------------------------------------------------------------
This track looks like CDDA with probability 100%

------------------------------------------------------------
Processing file: [14.wav]

------------------------------------------------------------
This track looks like CDDA with probability 100%

------------------------------------------------------------
Processing file: [15.wav]

------------------------------------------------------------
This track looks like CDDA with probability 40%

------------------------------------------------------------
Processing file: [16.wav]

------------------------------------------------------------
This track looks like MPEG with probability 100%

------------------------------------------------------------
Processing file: [17.wav]

------------------------------------------------------------
This track looks like CDDA with probability 99%

------------------------------------------------------------
Processing file: [18.wav]

------------------------------------------------------------
This track looks like CDDA with probability 99%

------------------------------------------------------------
Processing file: [19.wav]

------------------------------------------------------------
This track looks like CDDA with probability 99%

------------------------------------------------------------
Processing file: [20.wav]

------------------------------------------------------------
This track looks like CDDA with probability 99%

------------------------------------------------------------
Processing file: [21.wav]

------------------------------------------------------------
This track looks like CDDA with probability 99%

------------------------------------------------------------
Processing file: [22.wav]

------------------------------------------------------------
This track looks like CDDA with probability 99%

------------------------------------------------------------
Processing file: [23.wav]

------------------------------------------------------------
This track looks like CDDA with probability 99%

------------------------------------------------------------
Processing file: [24.wav]

------------------------------------------------------------
This track looks like CDDA with probability 99%

------------------------------------------------------------
Processing file: [25.wav]

------------------------------------------------------------
This track looks like CDDA with probability 99%

------------------------------------------------------------
Processing file: [26.wav]

------------------------------------------------------------
This track looks like CDDA with probability 99%

------------------------------------------------------------
Processing file: [27.wav]

------------------------------------------------------------
This track looks like CDDA with probability 99%

------------------------------------------------------------

And I got very inconsistent results. Some even very likely to be mpeg.

Then I used foobar to make 1 single file with all the tracks:

QUOTE

Processing file: [whole.wav]

------------------------------------------------------------
This track looks like CDDA with probability 99%

This time it seems there is no doubt that it is CDDA even though each individual piece showed something entirely different!
How can this be?
Should this happen?
What it most reliable?
How reliable is it at all when you can get results like this?

Bern

Sep 11 2004, 21:07

QUOTE(Jan S. @ Sep 9 2004, 12:42 AM)

First I analysed a bootleg cd track by track:
.....
And I got very inconsistent results. Some even very likely to be mpeg.
Then I used foobar to make 1 single file with all the tracks:
...
This track looks like CDDA with probability 99%

This time it seems there is no doubt that it is CDDA even though each individual piece showed something entirely different!
How can this be?
Should this happen?
What it most reliable?
How reliable is it at all when you can get results like this?

Jan S.,
1.General The current algorithm is statistical one, supposing that you have only one composition per wav file. So the most accurate way to use it is to scan each track separately and then to make conclusion by yourself. By the way - we have some disks with MP3s and CDDAs-soureced tracks at single disk (4-5MP3, 6-5CDDAs, for example), so we can not make solution instead of you about the whole disk;)

2. How this can be?
Generally: The 80% of songs from your disks are probable CDDA according to the first case, and all the disk most probable CDDA according to the second case. So the results obtained in both cases are close. It is more or less expected - most probably your disk is CDDA.
Details:
When you have a number of results, that shows you 5 MPEGs and 22 CDDAs per disk, you could make only subjective solution about the entire disk. One man will say that it is mostly CDDA, another - that it is MP3, next one - that it is UNKNOWN. The v.0.7 program is not intended to make a solution about entire disk, it could make only solution about single track/song/composition.
Current statistics predicts 99% CDDA for this disk if you treat it as an single composition.
But it is not single composition, so the result, obtaned in the last case is not very correct, but only more/less...

3. Should this happen?
Yes, this could happen - current v0.7. algorithm is not intended to analyze a whole disk, only a single songs.

4. What is the most reliable?
For sure, the first case.

5. How reliable is it at all when you can get results like this?
We have a number of such questions from other people too.
For this very case we have build-in into a new version of auCDtect (TauAnalyzer - windows-only GUI version with build-in ripper) the solution algorithm for entire disk, the most correct from our mathematical point of view. But now it is in testing mode and we plan to release it only in couple days.

Best Regards,
Oleg.

This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.