2. Local Correct
Prev  Chapter 6. Correcting Data Automatically  Next

2. Local Correct

Local correct groups together some powerful tasks that can be performed without an Internet connection, each task attempts to improve the data in one column. Most of the local correct tasks work in a similar fashion, this is explained fully in the Cross Referencing Correct section, the other local correct tasks are discussed next

2.1. Correct Artwork

Often artwork is not embedded into the metadata but held as a separate file, sometimes it is not available at the same time and has to be retrieved later. But cover art is much more useful if it is embedded into the metadata because only then will it be displayed in Music players such as iTunes and iPods. Correct Artwork checks the folders of the songs you are correcting and adds any image files it finds to the audio file unless they have already been added. The following formats are supported JPEG,PNG and GIF

There are options you can make that apply to Correct Artwork and when getting artwork from a Musicbrainz or Discogs match

Many applications only show the first image, so you may decide to only add the first image by setting the value of Maximum artwork images that can be added to song.

You can resize large images using Resize Artwork if dimensions larger than (pixels), very large images will make your songs correspondingly large.

You can ignore artwork smaller than a certain size using Ignore artwork smaller than this (pixels)

By default Local Artwork Correct ignores folders containing more than fifty songs because it is unlikely that the folder contains only a single album, but sometimes this is the case and you can remove this restriction by checking Preferences/Local Correct/Artwork Correct/Allow Artwork to be loaded from folders containing more than 50 songs

2.2. Delete Duplicates

Song is a duplicate if allows you to decide how Jaikoz find duplicates. You can detect duplicates based on Musicbrainz Id, AmpliFIND Id or MusicBrainz Id/AmpliFIND Id

The When you have duplicate songs option defines what to do once duplicates have been found. They can either be deleted , moved to a special Duplicates folder or linked to the duplicate file that is not being deleted. If you want them to be moved you must specify a Duplicates Folder. If you want to link them use the Alias should be named so that option to specify if they should keep their name, be renamed to match the file they link to, or have a combination of both names.

Delete from iTunes if moving duplicates searches for these songs in iTunes and removes them from the iTunes library as well.

2.3. Correct Track Nos

If Remove Undisplayable Characters is selected control characters can be removed.

2.4. Cross Referencing Correct - Correct Artists/Albums/Titles/Genres/Recording Times/Comments/Composers

If you have a number of songs that you wish to adjust the data for the chances are at least some of them are by the same Artist and from the same Album. However if the source for these files are different the Metadata may be subtletly different,this means that when you try and play all tracks by a particular group or artist they do not all appear. For example the following artists are all different,but only slightly

  1. Flatuella Blurb - All words start with a capital letter

  2. Flatuella blurb- First words start with a capital letter

  3. The Flatuella Blurb - Additional definite article

  4. Blurb Flatuella - Words in different order

  5. Flatuella Blorb - Misspelling of second word

  6. Flatuella Blurb - Extra white space character at end of word

  7. Flatuella_Blurb - Words separated by underscore character

  8. Flatuella Blurbs - Extra punctuation

Which is the correct answer?

Item five is a misspelling so it is not this,item four is not that easy to read but is the accepted format for some audio encoders, any of the others could conceivably be ok. The correct answer really depends upon your personal preference. AutoFormat allows you to automatically modify values to to your personal taste.Are all these artists the same band, as a human reading this document it should be obvious that they are but to a computer it is not. But AutoMatch can detect these similar values, again the similarity is defined by you.

Cross Referencing Correct trys to work out which is the correct value to use by a combination of comparing similar values, and user options. Local Correct is available for the following fields

  1. Artist

  2. Album

  3. Listitem

  4. Genre

  5. Recording Time

  6. Comments

  7. Composer

2.4.1. How it Works

Cross Referencing Correct works on a column by column basis,it identifies similar values and converts all similar values to the most common value. The procedure is as follows

  1. Create an empty new list

  2. For each record retrieve the Field Value relating to that column

  3. Using the options defined in AutoMatch Preferences calculate a Match Value

  4. Using the options defined in AutoFormat Preferences calculate a Format Value

  5. Has this field been found before, if not add it to the list with Count of zero

  6. If it has just increment the Count of this field in the list

  7. Sort the list by their Match Value and then Count,highest count first.

  8. If a record has the same Match Value as its preceding record set its Format Value to the same as the previous records Format Value.

  9. Go through each Audio File record and find the record in the list with the same Field Value and set the name in the Audio record to the Format Value.

2.4.2. AutoMatch Preferences

Local correct tasks like Local Artist Correct only modify songs that do not have a Musicbrainz Id. The reason for this is that Local Correcters are primarily provided to improve the quaility of your metadata in preparation for Musicbrainz matching, so having run Autocorrecter on a group of songs you may run it again after making some manual edits to match the remaining unmatched songs, and would not expect songs that have been successfully matched to Musicbrainz to be modified further. You can make it run against all songs by checking Preferences/Local Correct/Auto Match/Do local correct on songs that already have a Musicbrainz Id

You can decide on what the Correcter deems is similar when matching by selecting options from the Auto Match Preferences . The more options you enable the more matches the correcter will make but more false positives will also occur.

2.4.2.1. Match words that appear misspelt

Jaikoz can look for values that sound the same when spoken (based on English) but are spelt differently. For example Audio Bullies and Audio Bullys

2.4.2.2. Ignore Word Order when matching

If a field contains multiple words this option will ignore the order. This is particulary useful for Solo Artists because some Audio Encoders specify surname then firstname. For example 'Kate Bush' and 'Bush Kate' would match.

2.4.2.3. Ignore Case when matching

Ignore case when matching values. For example 'Flaming Lips' and 'Flaming lips' would match.

2.4.2.4. Ignore words in this list when matching

Certain common words in a Language are often used without much thought especially in group names. In English the biggest culprit is 'The'. This option allows to specify a Word Remover to use. For example 'Flaming Lips' and 'The Flaming Lips' would match.

2.4.2.5. Split into Words using these values

Earlier we looked at word order, but what defines a word. The Space character separates words but other characters such as '.' and ',' also separate words. Here you can specify a Word Separator which define show to break up words

2.4.3. AutoFormat Preferences

AutoFormat formats fields to save you from tedious typing , there are a number of options

2.4.3.1. Remove whitespace at start or end of value

This removes whitespace at the start or end of field values, this extra whitespace is marked pink in the field values

2.4.3.2. Remove Widespace

Widespace is when there is more than one space between words, this option removes the additional spaces so that there is only one space between words

2.4.3.3. Remove undisplayable characters

This option removes any characters found that do not relate to a character set, this works for all languages as defined in the Unicode standard, it is very useful for removing control characters

2.4.3.4. Capitalization

This option changes the case of the field, the options are as follows

  • No capitalization

  • Capitalize all words

  • Capitalize first word

  • Capitalize all letters

  • Lowercase all letters

If you select to Capitalize all words or Capitalize first words there a few improvements Jaikoz makes to simple capitalizing. The rules listed below are processed before the simple capitalization method of making the first letter upper case and the remaining letters of a word lowercase

  • O'Reilly rule

  • Mc rule

  • Acronym rule,words between 3 and 5 characters that contain only uppercase letters and numbers are not modified

2.4.3.5. Replace words by another word

Some words have acronyms or shorthands that can be used instead, this option allows you specify a Word Replacer to always use the one you want. For example 'ft' and 'Feat' are common shorthands for 'Featuring'. And '-' and '_' are often used instead of spaces.

2.4.3.6. Remove these Punctuation Characters

Specify a Punctuation Remover which will remove the list of characters you have specified from the resulting output.

Prev  Up  Next
Chapter 6. Correcting Data Automatically  Home  3. File and Folder Correct