Someone contacted me via email wanting to know how I created the deck.
Here is a link to the utility that I made (including a readme):
http://www.mediafire.com/download.php?ynk2j14zmyb
You will need the .Net framework installed.
Note: Don't try to break it, it has rather poor exception handling (reason being that I created it for personal use).
--------------------------------------------------------------------------------
From the readme:
Overview of tools used
subs2srs – A little utility that I created. You input 1 or 2 subtitle files (in SRT,ASS, or SSA format). All it does is parse through subtitles and generate text file(s) in a desired format to be used as input for other tools. If the subtitle files aren’t 1:1, the closest line (chronologically) will be used.
These tools are just the ones I used. Feel free to come up with a better toolchain.
Mp3splt – The tool I used to make the audio clips.
Avs2avi+AviSynth+ImageWriter plugin – Tools that I used to take the snapshots.
Kakasi – Tool that I used to generate the readings.
MKVextractGui – Tool that I used to extract the audio from the MKV video file.
Anki – My SRS.
If you need Japanese subtitles, you might find some here:
http://kitsunekko.net/JapaneseAnimeSubtitles.aspx
Overview of usage
Run subs2srs. subs2srs is basically a utility that will parse the subtitles and allow you to output the parsed info to a file in a one-line format of your choosing.
Read the help file to get an idea of what kinds of things you can do (Help->Formatting…).
You must enter at least one subtitle file.
If you have multiple episodes, use the batch feature. For example, if you have a 23 episode show, enter “1-23” in the Batch text box. Use $batch_num to substitute the current episode number in the subtitle file name or the output format or output file name. $batch_num always outputs 2 digits with a leading zero if necessary.
Enter the output format (you will have to refer to the help file for the tokens you may use.
The output format must look like this:
“<output format>” <token1> <token2> …
The quotes are required.
Enter the output file name (you can use the $batch_num token here).
Click Go.
Example
(This example does not create images or readings – see the advanced section below for that)
That probably did not make too much sense. So here is an example that you can follow:
Assume that we want to run the program for all 23 episodes of the anime Clannad. We will then split the audio files and create the Anki import files.
In a directory called I:/Temp2, create a directory structure like this:
./Clannad
./Anki
./Media
./Clannad_01_Audio
./Clannad_02_Audio
./Clannad_03_Audio
.
.
.
./Clannad_23_Audio
./Subs
./TempAudio
Put all the subtitle files in I:/Temp2/Clannad/Subs (there should be 23 Japanese subtitle files and 23 English subtitle files). They are named as follows:
Clannad_01_jap.ass
Clannad_01_eng.ass
Clannad_02_jap.ass
Clannad_02_eng.ass
Clannad_03_jap.ass
Clannad_03_eng.ass
Etc.
Put all the audio files in I:/Temp2/Clannad/TempAudio (there should be 23 files, one for each episode): They are named as follows:
01 - Clannad.mp3
02 - Clannad.mp3
03 - Clannad.mp3
Etc.
The Anki folder will hold all anki import files.
The Media/Audio folders will hold all of the split audio files.
Run subs2Srs
In Subs1, enter I:/Temp2/Clannad/Subs/Clannad_$batch_num_jap.ass
In Subs2, enter I:/Temp2/Clannad/Subs/Clannad_$batch_num_eng.ass
In Frame rate, enter 23.976
In the Replace… with … enter ; and , respectively. This will remove any instance of ; in the dialog. It to prevent Anki import from thinking it’s a separator.
In Batch, enter 1-23
Make sure the Include styled lines is unchecked so that we don’t parse any signs or karaoke effects.
In Output 1 Format we will put a format that will generate a .bat file that will split each audio file with mp3splt (assuming mp3splt is located in C:/WINDOWS/mp3splt.exe). Enter this:
"C:/WINDOWS/mp3splt.exe -q -o Clannad_{0}_{1:00.}.{2:00.}.{3:00.}-{4:00.}.{5:00.}.{6:00.}.mp3 -d I:/Temp2/Clannad/Media/Clannad_{0}_Audio "I:/Temp2/Clannad/TempAudio/{0} - Clannad.mp3" {1:00.}.{2:00.}.{3:00.} {4:00.}.{5:00.}.{6:00.}" $batch_num $s_min_t $s_sec $s_hsec $e_min_t $e_sec $e_hsec
Yes, that is one big line. {0} corresponds to the $batch_num token. {1} corresponds to the $s_min_t token. Etc. The “:00.” Means make it two digits with a leading zero and no decimal (Keep in mind that this is always the case for $batch_num so you don’t have to use this format when using $batch_num).
This is what one of the lines of the .bat files might look like when its generated:
C:/WINDOWS/mp3splt.exe -q -o Clannad_01_00.07.09-00.08.41.mp3 -d I:/Temp2/Clannad/Media/Clannad_01_Audio "I:/Temp2/Clannad/TempAudio/01 - Clannad.mp3" 00.07.09 00.08.41
In the Output 1 Output, enter this:
I:/Temp2/Clannad/TempAudio/Clannad_$batch_num_mp3splt.bat
In Output 2 Format we will put a format that will generate the files that we will use for importing into anki.
Enter this:
"[sound:Clannad_{0}_{1:00.}.{2:00.}.{3:00.}-{4:00.}.{5:00.}.{6:00.}.mp3];{7};{8}" $batch_num $s_min_t $s_sec $s_hsec $e_min_t $e_sec $e_hsec $sub1_line $sub2_line
This is what one of the lines of the semicolon separated anki import files might look like when its generated:
[sound:Clannad_01_00.07.09-00.08.41.mp3];この町は嫌いだ;I hate this city.
In the Output 2 Output, enter this:
I:/Temp2/Clannad/Anki/Clannad_$batch_num_anki.txt
Click Go!
The .bat files were created in I:\Temp2\Clannad\TempAudio
You can double click each one, or just create another batch file that will run each one. They will split the audio and put the split files in the appropriate folder. Move all the files to your Anki media directory.
Import the Anki files one by one.
You are done.
Advanced Topic – Generate Snapshots
For snapshots use avisynth and avs2avi.
Use something like the following format and file names:
"ImageWriter("Clannad_{0}_", {1}, {1}, "jpg")" $batch_num $m_frame
I:/Temp2/Clannad/Media/Clannad_$batch_num_Images/Clannad_$batch_num_snapshot.avs
Your should also change the format of the anki file:
"[sound:Clannad_{0}_{1:00.}.{2:00.}.{3:00.}-{4:00.}.{5:00.}.{6:00.}.mp3];<img src="Clannad_{0}_{7:000000}.jpg">;{8};{9}" $batch_num $s_min_t $s_sec $s_hsec $e_min_t $e_sec $e_hsec $m_frame $sub1_line $sub2_line
Be sure to prepend something like this to each AVS file:
DirectShowSource("I:/Temp2/Clannad/Media/Clannad_01_Images/Clannad_01.mkv").LanczosResize(256,144)
ConvertToRgb
You can run a command like this for each AVS file (takes about 5 minutes per file):
I:/Dev/Utils/AVS2AVI/avs2avi.exe I:/Temp2/Clannad/Media/Clannad_01_Images/Clannad_01_snapshot.avs -o n -c null
Advanced Topic – Generate Readings
For the reading I use something like the following format and file names (this one is for toki wo kakeru shoujo):
";{0}" $sub1_line
I:/Temp2/Toki_wo_Kakeru_Shoujo/Anki/Toki_wo_Kakeru_Shoujo_01_kakasi.txt
Then generated the readings with kakasi and appended them to the Anki file:
cd /cygdrive/i/Temp2/Toki_wo_Kakeru_Shoujo/Anki
cat Toki_wo_Kakeru_Shoujo_01_kakasi.txt | iconv -f UTF-8 -t EUCJP | /cygdrive/c/kakasi/bin/kakasi.exe -JH -p | iconv --unicode-subst=? --byte-subst=? --widechar-subst=? -f EUCJP -t UTF-8 | paste -d '\0' Toki_wo_Kakeru_Shoujo_01_anki.txt - | tr -d '\r' > Toki_wo_Kakeru_Shoujo_01_anki_reading.txt
Last edited by cb4960 (2009 January 25, 1:31 pm)
