I use it for listening.

I recently stated using to just get all the talking parts from the drama episodes.and usually the length is at least half the episode, and obviously it depends on how much you filter like my filtering text right now is:

yoro;um.;ano;yes!;yes.;thank;please;ah..;what?;eh?;me?;but.;he?' what?;my fault;what in the world;that's right.;is that true;eh!;ouch; sensei.; no.; um.; ah,; ah.; excuse me.; Excuse me.; yes!; yeah!; Um.; Um,; Um,; Please!; Got it!; Yes!; Everyone!; Yes?; Eh?; Oh.; Oh!
=== then I join the mp3 files using mp3albummaker. opening it by folder works fine, opening by files make it crash though.

So in the time it takes you listen to the whole episode 4 times, you can listen to pure dialogue part like 10 times,etc.

It's just awesome forme cause I'm not into re-watchign the ep but now i can do it efficiently (just listening than watching) so I love it.

For this use obviously it doesn't matter what language the subs are in ( except for filtering) as long it matches the video.

Now that's just ingenuity there. Obviously I'd have to take out timing buffers, but it would get rid of the annoying theme songs they put in to stretch the episode out. Great idea.

The program you linked sounds good to use with the iKnow audio files to make shadowing files (well, except you won't have the space buffer).

So, subs2srs -- How are you using it these days?

I'm about to try a variation of my current method--well, old method now, so I'm pretty excited.

Originally: I was going through a 'video clip deck' I made from an episode, selecting new vocabulary in batches and either finding that vocabulary in the smart.fm Core 6000 deck, or making single word cards for them in an 'extemporaneous' deck I use for making cards out of the list of interesting words and phrases that I keep in Ultraedit.

I did those Core 6000 cards the way I did Core 2000--Recognition, where I'd read and subvocalize and write stuff I was fuzzy on, then flip over and listen/speak, correcting my subvocalization.

Anyway, when those cards became mature, I would go through the original video deck (as 'video active recall' with no dictation unless I was feeling masochistic), focusing just on parsing/grammar and listening to sentences, as if watching the episode.

At any rate, that sounds complicated, but it's actually a really comfortable process of distributing the burden. However, it definitely needed some changing up. Going to test out new methods and get back to you in a month or something. ;p

I'm not sure why I was so excited (see above), as it was really a minimal change I made, but thought I'd post a minor update. Now, with all the subtitles qop123 provides in timed/untimed form at d-addicts (be sure to go and thank them if you use them, they seemed quite appreciative of that kind of verbal support), as I mentioned a while ago in another thread, I've taken a different tact, more of a 'sampling' approach. This was the original goal from the beginning--all 'extemporaneity', but at the time jsubs were rare enough that I put it off, focusing on SRSing entire shows.

I simply grab an episode of a given program, with Japanese and/or English subtitles (for the videos I use direct download sites found via Google, rather than torrents), and either go straight to making a deck, or watch it and make (mental) notes of the most interesting Japanese in the episode, usually relatively small spans, and just create a deck from that with subs2srs when it's finished.

The goal for this is to minimize the amount of cards to essentially just small samples from a large corpus (the many episodes with subtitles now available). So I end up with 25-50 sentences per episode. I generally remove duplicates, only lines with kanji, each line must be between 800ms-2500ms in length (via newest features in subs2srs), and I make use of spans and cropping where it's necessary to remove the extraneous stuff like opening/ending songs, hardcoded subs, etc.

This whole process, including processing and importing, takes less than 10 minutes, methinks, total. My goal is to keep up a steady deck-creating pace until I've reached 100 of them, to build a decent-sized corpus of 3000-5000 cards. Other than that, for doing actual study, I just use the method I described in above comment or here.

I'm still working on developing a similar strategy for raw novels... I have some ideas for that, but, busy busy.... (but not too busy to comment here ;p)

nest0r Wrote:I'm still working on developing a similar strategy for raw novels... I have some ideas for that, but, busy busy.... (but not too busy to comment here ;p)

Regarding that, take a look at my latest post in this thread http://forum.koohii.com/showthread.php?p...7#pid92257 . It seems like that is the sort of thing you would be wanting?

EDIT: huh, seems you already had!

blackmacros Wrote:

nest0r Wrote:I'm still working on developing a similar strategy for raw novels... I have some ideas for that, but, busy busy.... (but not too busy to comment here ;p)

Regarding that, take a look at my latest post in this thread http://forum.koohii.com/showthread.php?p...7#pid92257 . It seems like that is the sort of thing you would be wanting?

EDIT: huh, seems you already had!

I think I'll actually be shying away from doing 'sentences' per se, more like swaths of 1-3000 words, per some ideas based on 'quantfifying semantic information'/judging information value (see comment on first or second page of my book thread). Although I do think there's probably a lot of potential streamlining to do with regards to going from 'unknown word' to 'Anki card' automation.

from my previous post about using it to listen to drama/anime more efficiently.
i wanted to share these audio files since I'm going to delete them off my computer. haven't listened to them since last year or something?

http://www.mediafire.com/myfiles.php?r=niaip#2,1
The mediafire folder holds
Contains bakemonogatari episode 1 & 2 (15/16 minutes each)
Suzumiya Haruhi no yuutsu: Episode 21 to 24
JIN episodes 1-7
London Hearts - I think this ep is from 2010... it's the one where they rank the geinin by their wives (YOME...dakaretai). there was memorable part where chihara junior said to inoue 美容師が途中で死んだ?
GONZO episode 4
Liar Game II episode 1-7
Samurai High school episode 02 04 3
MOP GIRL episode 6-9
Saito-san episode 10 (the last ep?)
Baptista team season 1 episode 01


When I made them I tried filter out phrases and words that I don't need to hear since they're used so often so the audio file does not contain every single line from the episode ... i tried to filter out all the easy/heard a billion times stuff but some of it still went through but yeah it's more efficient than listening to the whole episode.

I have more audio files I made but I can't upload everything right now.　

that mediafile like isn't working?

oh sorry, got the link:
http://www.mediafire.com/?4laew456eg2qi
happy downloading

Are these audio files purely for listening immersion or are there any scripts/decks to accompany them?

thanks

well i just used them for immersion since it's not the whole episode but just the parts with the lines.

I did make a deck for JIN (ep3-7) I'm making decks for the other eps right now.
http://learnanylanguage.wikia.com/wiki/Sub2srs_decks

If you want the Japanese script for JIN, samurai high school , liar game II
it's on d-addicts

so is london hearts: http://www.d-addicts.com/forum/viewtopic_87376.htm
transcript for haruhi: http://www1.atwiki.com/animetranscripts/pages/719.html

for mop girl and gonzo there's entries on dramanote　＆ you an try googling for lines if you want to (like blogs that quote from the show while blogging about it).
GONZO episode 4
MOP GIRL episode 6-9

or you could google translate the korean fansub script to Japanese if your japanese isn't that beginner since it's obviously not going to be exactly what they say. i find it helpful as opposed to like no script whatsoever or the english script
http://forum.koohii.com/showthread.php?p...#pid117636

ADDED MORE:
CHASE episode 1
Haken no hinkaku 3-5
Propose Daisakusen Episode 10.
Watashitachi no Kyoukasho 1 & 2
Nodame Cantabile (DRAMA) Episode 5
Yume t oAme no ato ni episode 1
2 random honma dekka episodes.
JIN 9-11

the files in the folder I made called Japanese TV shows are just japanese tv shows' audio... i didn't do anything with sub2srs.

Id like to revisit this topic to see how ppl have progressed with their Subs2srs review since last.

I have been into it for about a month now and Id like to get some feedback on few things.

- I have my card set up with the picture + text + audio on the question side and I have to repeat the sentence after hearing it while reading it at the same time. What I find though is that its really hard to shadow "natural" Japanese. They often speak way too fast and mumble the words, many times words becoming almost entirely inaudible.

- I realized at the beginning I was spending way too much time on each card, trying to figure out the grammar being used and the roots of all of the contracted words. Trying to distinguish between what is gobi and what is not was driving me mad. I have switched to more of a quantity over quality approach now, meaning I just want the exposure to spoken Japanese and to pick out a few new words here and there instead of having a deep understanding of it. Thoughts on this?

Looking forward to the feedback.