IPB

Welcome Guest ( Log In | Register )

> Notice!

Please take a minute to check our Frequently Asked Questions. Use Search to reveal possible related topics.

Also make sure you've read the Forum Guidelines before posting in this forum.

 
Reply to this topicStart new topic
> Second instance or alternative, Comparing similar files and coping tags
Zerow
post Mar 17 2015, 21:16
Post #1


Member


Group: Full Members
Posts: 703
Joined: 31-January 15
From: East of Eurolandia
Member No.: 20375
Mp3tag Version: 2.86



Hello

I have this problem, that I will simplify and describe in points:

1] Some of the old songs have LYRICIST tag with names written down in them

2] I want to add new songs

3] I do not remember what old song have a lyricist

4] I do not remember if a new song is some other version of an old song or a completely new song

5] So: I have to check every single new song [new added file] and compare it with already existing old songs [files]



So you can see that if I could open a second instance of Mp3tag, I could just switch between windows [one with old files and a second one with the new files]. It would be still a drag but a much less of a drag than my current one

Alternatively: if I could highlight [with lets say bright green] a set of files with a given characteristic [for example a creation date from today or by a name of a new album], then I could load the old and the new songs [all of my files] and just simply scroll throughout the whole list and see which files [the new ones] should be dealt with [if they lay next to a old not highlighted file bearing the same or similar name]. It would be still a manual job with a possibility of making errors [skipping], but it would be much more efficient and safer that my current possibilities
Go to the top of the page
 
+Quote Post
DetlevD
post Mar 17 2015, 21:47
Post #2


Member


Group: Full Members
Posts: 6729
Joined: 26-May 06
From: Wuppertal, Germany, Planet Earth
Member No.: 3194
Mp3tag Version: 2.85



QUOTE (zerow @ Mar 17 2015, 21:16) *
... I have to check every single new song [new added file] and compare it with already existing old songs [files] ... if I could highlight ... then I could load the old and the new songs ... and see which files ... should be dealt with ...

In Mp3tag list view do create a new column for a 'highlight' helper tag-field, which you can give a name of your wish.
Load some media files, select them all, fill the helper tag-field with one or more characters, for example '*' or '#' or 'X' or 'new' or ... what you want.
Then additonally load the other set of media files.
Sort all files by ARTIST or whatever sort criteria you want.
Have a look at the 'highlight' column.

DD.20150317.2147.CET


--------------------
* Beyond that, don't ask, when you don't know what to do with the answer. *
♥ home is where the heart is ♥
Go to the top of the page
 
+Quote Post
ohrenkino
post Mar 18 2015, 08:28
Post #3


Member


Group: Full Members
Posts: 9319
Joined: 9-December 09
From: Norddeutschland / Northern Germany
Member No.: 11458
Mp3tag Version: 2.85h



QUOTE (zerow @ Mar 17 2015, 21:16) *
...
3] I do not remember what old song have a lyricist

4] I do not remember if a new song is some other version of an old song or a completely new song

5] So: I have to check every single new song [new added file] and compare it with already existing old songs [files]
So you can see that if I could open a second instance of Mp3tag, I could just switch between windows ...

The main problem is: synchronizing.
And this especially with larger collections.
You would have to do every action twice: once you have modified something in one instance, the other one would have to reload the data and/or update the loaded data. As you would probably load one instance with the whole collection and the other instance only with a subset, it would always be the window with the huge amount of data that needs reloading.
So, in short: working with mp3tag would become very tiring, I guess, as you keep waiting for the synchronization.

But as DetlevD described one alternative, I would like to add that the main approach is:
sorting and filtering.
All those conditions that you have described can be expressed with filter expressions. A second instance is not really necessary.


--------------------
42 - wie war die Frage / what was the question / quelle était la question
Go to the top of the page
 
+Quote Post
DetlevD
post Mar 18 2015, 12:47
Post #4


Member


Group: Full Members
Posts: 6729
Joined: 26-May 06
From: Wuppertal, Germany, Planet Earth
Member No.: 3194
Mp3tag Version: 2.85



QUOTE (zerow @ Mar 17 2015, 21:16) *
... I have to check every single new song [new added file] and compare it with already existing old songs ...

You may take into account, to compare two text files, ...
having each file the tag dump from one or more media files, ...
made by the help of the functions $list and $listtec.
https://www.google.de/search?q=site%3Aforum...ag+Tag+Dump+Col

DD.20150318.1247.CET


--------------------
* Beyond that, don't ask, when you don't know what to do with the answer. *
♥ home is where the heart is ♥
Go to the top of the page
 
+Quote Post
Zerow
post Mar 18 2015, 18:40
Post #5


Member


Group: Full Members
Posts: 703
Joined: 31-January 15
From: East of Eurolandia
Member No.: 20375
Mp3tag Version: 2.86



QUOTE (zerow @ Mar 17 2015, 21:16) *
... I have to check every single new song [new added file] and compare it with already existing old songs [files] ... if I could highlight ... then I could load the old and the new songs ... and see which files ... should be dealt with ...

QUOTE (DetlevD @ Mar 17 2015, 21:47) *
In Mp3tag list view do create a new column for a 'highlight' helper tag-field, which you can give a name of your wish.
Load some media files, select them all, fill the helper tag-field with one or more characters, for example '*' or '#' or 'X' or 'new' or ... what you want.

I have already used that method may times while clearing my collection and setting the new corrected nomenclature [system, order]

I must have been working too much, to not to come up by myself with adapting this trick to those ordinary purposes

QUOTE (DetlevD @ Mar 17 2015, 21:47) *
Then additonally load the other set of media files.
Sort all files by ARTIST or whatever sort criteria you want.
Have a look at the 'highlight' column.

Yes, that works, I've tested it on my latest addition of new songs

But could work better. Because the problem is the amount of old files [20 000+ and counting]. Scrolling throughout them with attention and thinking is time consuming and simply exhausting. And taking into account the frequency of this task it will be choosing between a drag and between disregarding potential information [data in tags]



The way to make it better would be if I could filter titles. I've already have been provided with codes for filtering while searching dis-similarities between TITLE and ALBUM tag, in one file:
CODE
"$ifgreater($strstr($lower($left(%TITLE%,6)),$lower($left(%ALBUM%,6))),0,1,0)" IS 1

"$strstr($lower($left(%TITLE%,6)),$lower($left(%ALBUM%,6)))" GREATER 0

"$ifgreater($strstr($left(%TITLE%,6),$left(%ALBUM%,6)),0,1,0)" IS 1

"$strstr($left(%TITLE%,6),$left(%ALBUM%,6))" GREATER 0

But this new filter would have to work in such a way: show only files with titles similar to titles of other files

And the key would be how much similar?


It should for example show

I Love You.mp3
I Love you.flac
I'll Love You.mp3
I'll Love You Part 2.mp3
I'll Love You Part .mp3
I'll Love, You Part 2.mp3


But take out from the list things like

I do not love you.mp3
I will always love you.mp3
I never loved you - you asshole!.mp3


Because if I make a mistake and put a coma where it shouldn't be or accidentally delete the number of part, then it won't show up at all. Such a code should look at signs, disregard pauses between them or some extra letters [because this will be the latter job of a human being, to manually evaluate the results]

So the code should return results with are similar in [lets say] 80%. And if there would be to many unwanted positions [not similar enough], or the other way, then there would be made a correction in code [that would show less or more files]. Just like in those codes that I presented above, where you can change the number "6". But as I explained, this particular code can't look at a "unchopped" string but would have to count the percentage of signs

Is it doable? I can guess that a filtering of that kind on that many files would take a very long time


QUOTE (zerow @ Mar 17 2015, 21:16) *
...
3] I do not remember what old song have a lyricist

4] I do not remember if a new song is some other version of an old song or a completely new song

5] So: I have to check every single new song [new added file] and compare it with already existing old songs [files]
So you can see that if I could open a second instance of Mp3tag, I could just switch between windows...

QUOTE (ohrenkino @ Mar 18 2015, 08:28) *
The main problem is: synchronizing.
And this especially with larger collections.
You would have to do every action twice: once you have modified something in one instance, the other one would have to reload the data and/or update the loaded data.
...

No, that is incorrect; sort of

Do this test:
1] Load a bunch of files into Mp3tag
2] Go to explorer / file handler
3] Delete some of those loaded files
4] Go back to Mp3tag
5] Select deleted file
6] See how Mp3tag reacts


Also, please do this other test:
1] Load a file into Mp3tag
2] Copy it's title to filter
3] change it's title outside Mp3tag [in Winamp or whatever]
4] Select changed file
5] See how Mp3tag reacts; especially the filter



Although my problem with "LYRYCIST" is quite solved [at the top], I'm still opting for the second instance. Because rarely I filter a lot, remove unwanted files from list, do whatever I need- and then it often comes a moment that I need to change a tag in a single file. But most likely I need / should use all the tags from some other existing file which is not on the current list. So it's drag, no matter how I do it [use the method from the top or save TITLE to clipboard]; because at the end I have to load files once again. And that takes time and / or disrupts my working [filtering] process

And as those two above test show, I do not have to change data once it have been changed outside [first instance of] Mp3tag. The file either disappears or updates itself when highlighted one more; and the already made changes simply do not show up. So It may be that it should be reloaded [and when it really neends to I do it], but it doesn't necessarily have to be reloaded- because it works to some point. At that "point" is all I need, when I'm working in the "point mode"

If you open a JPG photo in Photoshop and the same JPG file in PaintShop Pro and then make some [different] changes in both of those programs- do you need a synchronization for that? Or just a basic knowledge and common sense, that whatever you will save as latter will be the result left on a hard drive? And that the changes that are being made on first JPG won't automatically appear on the second version of it [at least not until they are saved in first program and detected in the second program]?

[So I do don't understand, why would I need to repeat each action twice? Especially if in one intance I would have only old files and in the second only new files?]



QUOTE (zerow @ Mar 17 2015, 21:16) *
... I have to check every single new song [new added file] and compare it with already existing old songs ...

QUOTE (DetlevD @ Mar 18 2015, 12:47) *
You may take into account, to compare two text files, ...
having each file the tag dump from one or more media files, ...
made by the help of the functions $list and $listtec.
https://www.google.de/search?q=site%3Aforum...ag+Tag+Dump+Col

DD.20150318.1247.CET

Thank you for this alternative

But creating some external lists is just to many steps. I also do no wish to potentially bury myself under a text file lists. A simpler way would be just using a second tag program [but that is obsolete question as the main problem is solved at the top of this post]
Go to the top of the page
 
+Quote Post
ohrenkino
post Mar 18 2015, 19:22
Post #6


Member


Group: Full Members
Posts: 9319
Joined: 9-December 09
From: Norddeutschland / Northern Germany
Member No.: 11458
Mp3tag Version: 2.85h



QUOTE (zerow @ Mar 18 2015, 18:40) *
...I can guess that a filtering of that kind on that many files would take a very long time

80000 files take some 20 seconds depending on computing power.
QUOTE
...If you open a JPG photo in Photoshop and the same JPG file in PaintShop Pro and then make some [different] changes in both of those programs- do you need a synchronization for that? ...... Especially if in one intance I would have only old files and in the second only new files?]

Most programs that allow multiple instances lock files so that you would have to write a copy. Mp3tag does not lock files.

What you do here is a so-called "dirty read" which gives you a snapshot of the data at the time of reading. Esp. if you try to get your data more consistent I would not tolerate outdated data in one of the instances.

If you know which old and new files you want to edit, you can d&d them into MP3tag while pressing CTRL. This adds the files to the already loaded ones and you have all those files there that you want to edit.

Just along these lines: iTunes and MP3tag work fairly well together as you can drag&drop files from iTunes into MP3tag for editing (I wouldn't do it in iTunes as you can never be sure that the modifications really end up in the files and not just the database). So iTunes could act as the second instance of Mp3tag.


--------------------
42 - wie war die Frage / what was the question / quelle était la question
Go to the top of the page
 
+Quote Post
DetlevD
post Mar 18 2015, 19:42
Post #7


Member


Group: Full Members
Posts: 6729
Joined: 26-May 06
From: Wuppertal, Germany, Planet Earth
Member No.: 3194
Mp3tag Version: 2.85



QUOTE (zerow @ Mar 18 2015, 18:40) *
... I do not have to change data once it have been changed outside [first instance of] Mp3tag. The file either disappears or updates itself when highlighted one more; and the already made changes simply do not show up. ...

Did you ever ... press [Ctrl+T] ... for re-reading the tag of a single file?

DD.20150318.1942.CET


--------------------
* Beyond that, don't ask, when you don't know what to do with the answer. *
♥ home is where the heart is ♥
Go to the top of the page
 
+Quote Post
Zerow
post Mar 18 2015, 21:12
Post #8


Member


Group: Full Members
Posts: 703
Joined: 31-January 15
From: East of Eurolandia
Member No.: 20375
Mp3tag Version: 2.86



QUOTE (ohrenkino @ Mar 18 2015, 19:22) *
80000 files take some 20 seconds depending on computing power.
[...]


And it the future the cost of terra-tetra-tera-tyranozarusbyte will drop to one dollar



Of course I will patentielly wait for the Moore's Law to kick in even futher. But until then I have to wait around 15 seconds- when reloading. But when loading for the first time I can go play with a cat

QUOTE (ohrenkino @ Mar 18 2015, 19:22) *
[...]
If you know which old and new files you want to edit, you can d&d them into MP3tag while pressing CTRL. This adds the files to the already loaded ones and you have all those files there that you want to edit.
[...]

Now that's a new one to me

And quite useful

Thank you for this info. And maybe I should read all of the on-line help
Go to the top of the page
 
+Quote Post
Zerow
post Mar 18 2015, 21:14
Post #9


Member


Group: Full Members
Posts: 703
Joined: 31-January 15
From: East of Eurolandia
Member No.: 20375
Mp3tag Version: 2.86



QUOTE (DetlevD @ Mar 18 2015, 19:42) *
Did you ever ... press [Ctrl+T] ... for re-reading the tag of a single file?


Yes, I should differently read the manual
Go to the top of the page
 
+Quote Post
Zerow
post Mar 25 2015, 11:09
Post #10


Member


Group: Full Members
Posts: 703
Joined: 31-January 15
From: East of Eurolandia
Member No.: 20375
Mp3tag Version: 2.86



QUOTE (zerow @ Mar 18 2015, 18:40) *
The way to make it better would be if I could filter titles. I've already have been provided with codes for filtering while searching dis-similarities between TITLE and ALBUM tag, in one file:

CODE
"$ifgreater($strstr($lower($left(%TITLE%,6)),$lower($left(%ALBUM%,6))),0,1,0)" IS 1

"$strstr($lower($left(%TITLE%,6)),$lower($left(%ALBUM%,6)))" GREATER 0

"$ifgreater($strstr($left(%TITLE%,6),$left(%ALBUM%,6)),0,1,0)" IS 1

"$strstr($left(%TITLE%,6),$left(%ALBUM%,6))" GREATER 0

But this new filter would have to work in such a way: show only files with titles similar to titles of other files

And the key would be how much similar?


It should for example show

I Love You.mp3
I Love you.flac
I'll Love You.mp3
I'll Love You Part 2.mp3
I'll Love You Part .mp3
I'll Love, You Part 2.mp3


But take out from the list things like

I do not love you.mp3
I will always love you.mp3
I never loved you - you asshole!.mp3


Because if I make a mistake and put a coma where it shouldn't be or accidentally delete the number of part, then it won't show up at all. Such a code should look at signs, disregard pauses between them or some extra letters [because this will be the latter job of a human being, to manually evaluate the results]

So the code should return results with are similar in [lets say] 80%. And if there would be to many unwanted positions [not similar enough], or the other way, then there would be made a correction in code [that would show less or more files]. Just like in those codes that I presented above, where you can change the number "6". But as I explained, this particular code can't look at a "unchopped" string but would have to count the percentage of signs

Is it doable? I can guess that a filtering of that kind on that many files would take a very long time

What about that improvement?

Is it doable?
Go to the top of the page
 
+Quote Post
DetlevD
post Mar 25 2015, 20:26
Post #11


Member


Group: Full Members
Posts: 6729
Joined: 26-May 06
From: Wuppertal, Germany, Planet Earth
Member No.: 3194
Mp3tag Version: 2.85



QUOTE (zerow @ Mar 25 2015, 11:09) *
But this new filter would have to work in such a way: show only files with titles similar to titles of other files
And the key would be how much similar? ... Is it doable?

Once you have defined the rules for similarity, then it might be done with Mp3tag too.

Have a look there ...
http://forums.mp3tag.de/index.php?showtopi...ost&p=74337
I have coded a well known algorithm into Mp3tag Export Scripting Language.

DD.20150325.2030.CET

This post has been edited by DetlevD: Mar 25 2015, 20:30


--------------------
* Beyond that, don't ask, when you don't know what to do with the answer. *
♥ home is where the heart is ♥
Go to the top of the page
 
+Quote Post
Zerow
post Mar 26 2015, 18:33
Post #12


Member


Group: Full Members
Posts: 703
Joined: 31-January 15
From: East of Eurolandia
Member No.: 20375
Mp3tag Version: 2.86



QUOTE (zerow @ Mar 18 2015, 21:14) *
But this new filter would have to work in such a way: show only files with titles similar to titles of other files
And the key would be how much similar? ... Is it doable?

QUOTE (DetlevD @ Mar 25 2015, 20:26) *
Once you have defined the rules for similarity, then it might be done with Mp3tag too.

Have a look there ...
http://forums.mp3tag.de/index.php?showtopi...ost&p=74337
I have coded a well known algorithm into Mp3tag Export Scripting Language.

That just gave me a list, not very way helpful. If I am suppose to change it somehow, that this is way out of my league

Instead I came up with this method:


I went manually through my files, using some filters [like showing only those files with "feat." artists]. As I have an order and well defined naming system, I was able to display certain groups, one at a time

The most tricky part was deleting only "copies" [from additional folders] and not "originals" [from the main artist catalog]; and also noticing, if I made an error in the past and forgot to put a "copy" in "original" artist [and thus making it the original]. And as there were some other glitches, it had to be done manually [but I'be managed to speed up the process with some tricks]



And now I can check for leftovers with this export code:


CODE
$filename(txt,utf-16)
$loop(%TITLE%,1)
$repeat('#',185)

$loop(%TITLE%,0)%YEAR%        %TITLE%        %ARTIST%



$loopend()$loopend()

This will give me still the list of all the titles [files], but in such a graphical way, than I can very easily spot repetitions. [And what's more, seeing a YEAR and ARTIST will pretty much allow me to evaluate the records, making me to go back to Mp3tag only then when a real case to be dealt with shows up]



But my question is: can similar list be generated via filter? With TITLE that is repeated at least once [leaving out all the single uses of song titles]?

[So now I'm not looking for similarities in tags enymore, but for copies of files based on the data from tag]

This post has been edited by zerow: Mar 26 2015, 18:34
Go to the top of the page
 
+Quote Post
ohrenkino
post Mar 26 2015, 18:44
Post #13


Member


Group: Full Members
Posts: 9319
Joined: 9-December 09
From: Norddeutschland / Northern Germany
Member No.: 11458
Mp3tag Version: 2.85h



QUOTE (zerow @ Mar 26 2015, 18:33) *
...
[So now I'm not looking for similarities in tags enymore, but for copies of files based on the data from tag]

No, Mp3tag does not know anything about the next file. It compares only data of the current file.
If you want to go through your collection, you might be interested in the program
similiarity
This scans through given folders and detects similar files in respect to tag and/or contents.


--------------------
42 - wie war die Frage / what was the question / quelle était la question
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: 16th December 2017 - 23:43