Understanding the files on the DVD of Arirang Karaoke (hk2 & midi0)

classic Classic list List threaded Threaded
76 messages Options
1234
Reply | Threaded
Open this post in threaded view
|

Re: Arirang Vol. 31 Song List (Chinese sorted)

bigboss97
Administrator
Used the Advanced Sorting in the GREAT "ASAP Utilities":

cn_xls_list.zip
Reply | Threaded
Open this post in threaded view
|

Re: English song book on DVD of Arirang Vol 31

cl163
In reply to this post by bigboss97
Here is the English song book for Arirang DVD Vol 31 in Excel format :

http://rapidshare.de/files/40038513/English_Song_Book.zip.html

By the way, does anyone know what program is used to produce a Midi Karaoke DVD like that?
Reply | Threaded
Open this post in threaded view
|

Re: Understanding the file MULTAK.DAT (only few readable songs)

bigboss97
Administrator
In reply to this post by Pminh
When I searched for "your" I found:


For "madonna":


But strange, that I can't find anything for "beatles","elvis".
Why only those few songs? If they are compressed they all shouldn't be readable.

Speculations :-)
1) here's a dynamic mask which "rotates" and plain text only appears accidentally at certain rotation.
2) The mask is based on the song number or title. All these readable songs have a common charater.
3) They are not compressed. Otherwise all should be unreadable.
Reply | Threaded
Open this post in threaded view
|

Re: Understanding the file MULTAK.DAT (Swiss File Knife)

bigboss97
Administrator
In reply to this post by Pminh
Reply | Threaded
Open this post in threaded view
|

Re: Understanding the file MULTAK.DAT (Mask found!)

bigboss97
Administrator
In reply to this post by bigboss97
Based on the fact, that some lyrics have upper and lower case reversed. In order to reverse this, a 0x20 mask is required. After analysis of some blocks (based on "OK") I found the mask  It's 4 bytes before the text starts:


It's only a single byte mask which is much simpler than California Electronics.

Do to:
Find the offset of a song. It could be in the index file.

Phuoc Can HUA
Reply | Threaded
Open this post in threaded view
|

Re: Understanding the file MULTAK.DAT (length of lyrics)

bigboss97
Administrator
In reply to this post by bigboss97
2 of those 4 bytes could the the length of the lyrics:




Reply | Threaded
Open this post in threaded view
|

Re: Understanding the file MULTAK.DAT (Help! addressing problem)

bigboss97
Administrator
In reply to this post by bigboss97
I still can't figure out how to associate the song ID from MASECOS4.IDX to the data in MULTAK. Any one can help?
I also tried in SEARCH1.IDX, no success  :-(

Alternatively, we can search for "OK" (start of a data block). I know that the first song in MULTAK is 20001. I simply count from there. But there's a risk that there are jumps in song numbers. So, the original song ID's won't match my count. Also, I need to identify the first song of each language because there's a jump in song number.

That's a very "dirty" solution. But if we can't read the address table I think that's the only way. Any suggestion?



bigboss97 wrote
Do to:
Find the offset of a song. It could be in the index file.
Reply | Threaded
Open this post in threaded view
|

Re: Understanding the file MULTAK.DAT (Offset table found, but...)

bigboss97
Administrator
It's obvious that from 0xd20 there's a table. But there's no direct link to the offset of each song. Since we know the size of a song block (based on "OK") I calculated the difference between 2 values. I took the result and times 8. Then we can see it's almost exactly the size of the blocks:



1) Ignore the first 2 songs. This table must be referring to the offset of each song.
2) There are recurring huge jumps, e.g. 20009. When do they occur?
3) 20013 & 20014 look more like corrupt data. But if that's the case these 2 songs should fail on the player?

To do:
Find out the pattern of the jumps. Then we can get to the offset of each song.
Reply | Threaded
Open this post in threaded view
|

Re: Understanding the file MULTAK.DAT (Offset table found, almost)

bigboss97
Administrator
I use the 2 MSB bytes seperately and multiply with 0x4b00. Besides the first 2 songs, the new algorithm seems to be handling the jumps well

Reply | Threaded
Open this post in threaded view
|

Re: Understanding the file MULTAK.DAT (Help! addressing problem)

bigboss97
Administrator
In reply to this post by bigboss97
I think that I've got the table read correctly. The list is created by reading the header information of each data block in MULTAK. Although there are craps in the middle it has read up to 6000 records without crash.

I guess that the address table also refers to the next big file. But I'll start looking into details in each block. I think that 6000 songs should be enough for experimenting.

I'll post the source codes later. Unless someone want to see the current codes.

lyrics.zip

bigboss97 wrote
I still can't figure out how to associate the song ID from MASECOS4.IDX to the data in MULTAK. Any one can help?
Reply | Threaded
Open this post in threaded view
|

Re: Understanding the file MULTAK.DAT (Read Header of each Record)

bigboss97
Administrator
In reply to this post by bigboss97
Ari2Midi.zip

The lyrics should be read in the same way.

I still haven't found a way to read the MIDI data. I've tried to read the section after the lyrics according to:
http://board.midibuddy.net/showpost.php?p=533722&postcount=31

which was used in California as well. I can't see any 5 bytes pattern:
Reply | Threaded
Open this post in threaded view
|

Re: Understanding the file MULTAK.DAT (use the shortest song)

bigboss97
Administrator
In reply to this post by bigboss97
Since we don't know the start of the MIDI part we need to do some experiments. I'm going to run the CalElect MIDI routine on the data. In order to get the best chance to read the MIDI correctly, I've picked the shortest song (from 5000).

mask.zip
Reply | Threaded
Open this post in threaded view
|

Re: Understanding the file MULTAK.DAT (Cut a song)

bigboss97
Administrator
In reply to this post by bigboss97
This routine will "cut" the shortest song from 1000 songs (starting at the given song number). Two files are written, one raw and one masked. Hopefully, this can help further analysis.

P.S. The number in the output file names is not the song number.
Ari2Midi.zip
Reply | Threaded
Open this post in threaded view
|

Re: Understanding the file MULTAK.DAT (Anyone see MIDI?)

bigboss97
Administrator
Can anyone see any MIDI data?
30093_raw.zip
Reply | Threaded
Open this post in threaded view
|

Re: Arirang Vol. 31 Song List (Chinese sorted)

violet2110
In reply to this post by bigboss97
Hi, can you help me to get the Korean list song also please, I really have no idea how to work with the DVD. I really appreciate your help. Thanks
Reply | Threaded
Open this post in threaded view
|

Re: Arirang Vol. 31 Song List (Korean)

bigboss97
Administrator
See:
http://www.nabble.com/Understanding-the-files-on-the-DVD-of-Arirang-Karaoke-%28hk2---midi0%29-tp16219545p18180984.html

I don't know what code is used. It's not unicode, i.e. you won't be able to read it on English XP. You might be able to read it on Korean XP.
Reply | Threaded
Open this post in threaded view
|

Re: Arirang Song List Generator - Chinese Song Lists vol.31 & vol.34

bigboss97
Administrator
In reply to this post by bigboss97
I have updated my list generator and created two Chinese song lists. It seems that vol.34 has less Chinese songs than vol.31, is that true?

Arirang_CN_31_34.zip
Reply | Threaded
Open this post in threaded view
|

Re: Arirang Vol. 31 Song List (Chinese sorted)

X Digital
In reply to this post by bigboss97
Found this forum after googled HOTDOG30.dat

Registered just to say: Your work is impressive!!!
Now trying to understand your code, especially the midi and lyrics parts :)
Reply | Threaded
Open this post in threaded view
|

Re: Arirang Vol. 31 Song List (Chinese sorted)

bigboss97
Administrator
Welcome to forum.

Since you were searching for HOTDOG file I assume that you are using California player, right? That one is basically done, at least for 2 DVD's of different versions.

In Arirang, I couldn't progress further. The lyrics part is half way done. But no progress in MIDI at all although (I believe) the MIDI is not that hidden as the lyrics, because normal eyes won't be able to read MIDI data. So, I guess the same mask is applied to the MIDI part. Unfortunately, I still can't read it.

Currently, I don't have access to Arirang player. Without patching the data I have no chance  :-( Even the analysis of patching MIDI data would be hard enough. If you can help with the MIDI part that would be great. Get the latest source code. It allows you to extract a section of data of one song.


X Digital wrote
Found this forum after googled HOTDOG30.dat

Registered just to say: Your work is impressive!!!
Now trying to understand your code, especially the midi and lyrics parts :)
Reply | Threaded
Open this post in threaded view
|

Re: Arirang Vol. 31 Song List (Chinese sorted)

henry123
This post was updated on .
In reply to this post by bigboss97
i was spending about 3 days to get all of Arriang DVD KARAOK' original midi files,  if anyone can help me to arrange the midi source of vol 34 (all of arriange  Karaoke DVD midi and mp3 files : order number from 00001   ) . I'll give you alll the midi files as well as uploading for dowload to all for you.
i got a trouble in extracting the resource of LISTSONG.DAT FILE (I THOUHG IT A DATABASE FILE NOT A DVD FILE)
midi.move.JPG
mp3.JPG

picofkaraoke.JPG
ok.JPG
ok2.JPG
ok.JPG

1234