Version: amarok-2.2.1-46.2 (using KDE 4.3.4) Installed from: openSUSE RPMs I discussed it already in http://forum.kde.org/viewtopic.php?f=116&t=83916 and I think it can be described best with photos, like I published in http://forum.kde.org/viewtopic.php?f=116&t=83916&start=10#p139675 This is the fastest way to see what I mean. You see, it doesn't work in Amarok 2, but it works with Amarok 1.4 The problem is in short, that there are artists, titles, a.s.o. which are displayed with questionmarks. At least it is not a simple characterset problem, because it can be that parts/songs of the same cd are ok, while others are not. I have mp3s where everything is ok with Greek letters only and with mixed English / Greek characters too. The mp3s are created from the original cd with uptodate applications: opensuse 11.1 amarok-2.2.1-46.2 Kernel 2.6.27.39-0.2-default. abcde-2.3.99.6-58.20 lame-3.98.2-1.pm.4.5 python-eyeD3-0.6.17-0.pm.3.1 (for the embedded coverimages) Maybe there is a connection with this: man lame: Lame will smartly choose which tags to use. It will add ID3 v2 tags only if the input comments won't fit in v1 or v1.1 tags, i.e. if they are more than 30 characters. In this case, both v1 and v2 tags will be added, to ensure reading of tags by MP3 players which are unable to read ID3 v2 tags. example metadata: exiftool -G -H -a 06-maroko-elli_paspala-sto_megaro_mousikis.mp3 [ExifTool] - ExifTool Version Number : 8.00 [File] - File Name : 06-maroko-elli_paspala-sto_megaro_mousikis.mp3 [File] - Directory : . [File] - File Size : 4.6 MB [File] - File Modification Date/Time : 2009:12:03 17:31:37+01:00 [File] - File Type : MP3 [File] - MIME Type : audio/mpeg [File] - ID3 Size : 9837 [MPEG] - MPEG Audio Version : 1 [MPEG] - Audio Layer : 3 [MPEG] - Audio Bitrate : 128000 [MPEG] - Sample Rate : 44100 [MPEG] - Channel Mode : Joint Stereo [MPEG] - MS Stereo : On [MPEG] - Intensity Stereo : Off [MPEG] - Copyright Flag : False [MPEG] - Original Media : True [MPEG] - Emphasis : None [MPEG] 0x0001 VBR Frames : 8164 [MPEG] 0x0002 VBR Bytes : 4775447 [MPEG] 0x0003 VBR Scale : 77 [ID3] - Album : Στο Μέγαρο μουσικής [ID3] - Artist : Έλλη Πασπαλά [ID3] - Title : Μαρόκο [ID3] - Recording Time : 1993 [ID3] - Genre : Jazz [ID3] - Track : 06 [ID3] - Comment (eng) : 6 | maroko | elli paspala | sto megaro mousikis | [ID3] - Picture Mime Type : image/jpeg [ID3] - Picture Type : Front Cover [ID3] - Picture Description : [ID3] - Picture : (Binary data 8526 bytes, use -b option to extract) [ID3] - Tagging Time : 2009:12:03 16:31:37 [Composite] - Date/Time Original : 1993 [Composite] - Duration : 03:33 (approx)
What system encoding do you use? And what encoding is used in the tags of these tracks?
locale LANG=de_DE.UTF-8 LC_CTYPE="de_DE.UTF-8" LC_NUMERIC="de_DE.UTF-8" LC_TIME="de_DE.UTF-8" LC_COLLATE="de_DE.UTF-8" LC_MONETARY="de_DE.UTF-8" LC_MESSAGES="de_DE.UTF-8" LC_PAPER="de_DE.UTF-8" LC_NAME="de_DE.UTF-8" LC_ADDRESS="de_DE.UTF-8" LC_TELEPHONE="de_DE.UTF-8" LC_MEASUREMENT="de_DE.UTF-8" LC_IDENTIFICATION="de_DE.UTF-8" LC_ALL= eyeD3 --set-encoding=utf8 ... Please note: It is _1_ cd, the artist is defined _once_ by the cddb and amarok2 doesn't show the artist _sometimes_ correctly. I understand, that UTF8-charsets can be tricky, but I have a lot files, where everything is ok.
Thanks for the feedback. You did try do a full rescan of the collection with and without the character set detection activated (in Settings -> Configure Amarok -> Collection), then restart Amarok, as I suggested in the forum, right?
In this case I created a test folder and did _not_ add this folder to the collection. I used drag and drop and moved the test-folder over the amarok-icon. I access the same files with amarok 2 and 1.4 via nfs. Should I send you a sample file? A lot of things changed since the start of the thread in the forum here. I remember I saw an option "character set detection" somewhere, but I cannot find it now under "Settings -> Configure Amarok -> Collection". The collection at the moment was collected by default values. But as mentioned above, it is not a file from the collection.
*** Bug 208601 has been marked as a duplicate of this bug. ***
Changing status, this slipped from my attention
Ability to turn off the charset detector is in 2.2.2, not 2.2.1.
There is a misunderstanding, sorry, I can confirm this with current 2.2-git. Using the charset detector or not doesn't change the problem at all. The two files still don't show correctly in Amarok, but do so in VLC.
Myriam, I didn't forget it to write to the mailinglist. I am still trying to find out, what is the problem. In short, IMHO picard doesn't write the tags with v1 correctly and this _incorrect_ version is shown with Amarok _correctly_. http://www.statemaster.com/encyclopedia/ID3 "ID3v1 also lacked support for internationalization. While nominally the text was supposed to be encoded in ISO-8859-1, in practice the user's local encoding was usually used, and so mojibake are common in ID3v1 tags." IMHO Amarok should use ISO-8859-1 for ID3v1 as default.
Al, why are you using ID3v1 tags?
Simply, because a lot of hardware (eg. car-audio) doesn't support v2. man:/lame "Lame will smartly choose which tags to use. It will add ID3 v2 tags only if the input comments won't fit in v1 or v1.1 tags, i.e. if they are more than 30 characters. In this case, both v1 and v2 tags will be added, to ensure reading of tags by MP3 players which are unable to read ID3 v2 tags." A lot of my files are tagged with lame, others with eyeD3. If a developer needs an example send me a PM.
So write both ID3v2.4 and ID3v1 tags to the file. There's no problem with having both -- ID3v2 lives at the front of the file and ID3v1 at the end. Amarok will use the UTF-8 encoded ID3v2.4 tag and your hardware player will use your whatever-encoded ID3v1 tag. In case you aren't aware, there is no actual standard for ID3v1, which is why things don't interoperate well and there's no standard encoding.
Charset detector, which could cause this with ID3v2 tags, is now disabled unless explicitly enabled.