So, is there any support for linux? It seems I can run it, but that it doesn't function properly (something about a reference error). I saw in the changelog that support was dropped?

stephenmac7 Wrote:So, is there any support for linux? It seems I can run it, but that it doesn't function properly (something about a reference error). I saw in the changelog that support was dropped?

Linux is no longer a supported platform.

Why was support dropped for linux?

Also, if you're not supporting it, is this at least open source?

夜空

The last show I imported took 2 hours to run, mostly on the snapshots step. Has anyone else seen it take this long? All the previous shows took more like 10 minutes. This show was a .mkv (previously I had .mp4) is that the reason?

Could be, depending what the encoding was, what codecs you use, and what hardware you have, it's possible those mp4s were able to decode faster, perhaps with hardware decoding or something.

Can't really say more without you trying one of the mp4s again, and giving us more information about your hardware and software.

1. does the audio normalization not work or am i doing something wrong? it doesn't seem to do anything when i check "Normalize Audio". clips from the same batch come out with some quiet and some loud. though it's not much of an issue because using mp3gain separately is easy enough.

2. has anyone noticed that the left & right audio channels seem to get swapped when making cards? or is it only happening to me? (i notice because i always use earphones). it probably doesn't really make a difference though, i was just wondering.

3. i'm wondering how everyone else is specifically creating cards. if i just run the program automatically, it seems that i'll end up with a lot of truncated sentences, since naturally the program doesn't know where sentences begin and end. i've been using aegisub to manually join short parts together (parts of the same sentence and also short sentences that make more sense when kept together). is this what everyone else does? or am i missing something?

just as a simple example:
(ユキノ）27歳の私は
15歳の頃の私より少しも賢くない
these two appear on separate lines and wouldn't make sense unless joined.

part of the reason i'm asking is, i was using Nayr's Core5000 deck, and i'm just wondering why no one has yet created a massive deck of assorted sentences. if you were looking for example sentences for a particular vocab word, you could then just search the deck, and maybe find a few examples if the deck were big enough. i mean, the sentences in Core5000 are one sentence per vocab word, which is certainly not ideal. and it's not like they've been carefully constructed to try to illustrate the meaning of the words or anything like that (at least i don't think they are?), they're just sentences that use the word. it seems to me that a lot of effort would be saved if some kind of pre-made deck could be created. or maybe i'm totally mistaken? my own subs2srs deck is still small right now (only like 350 cards, some of which are not useful) so maybe there's something i'm not getting because i haven't used the program enough yet.

qweasdzxc Wrote:1. does the audio normalization not work or am i doing something wrong? it doesn't seem to do anything when i check "Normalize Audio". clips from the same batch come out with some quiet and some loud. though it's not much of an issue because using mp3gain separately is easy enough.

2. has anyone noticed that the left & right audio channels seem to get swapped when making cards? or is it only happening to me? (i notice because i always use earphones). it probably doesn't really make a difference though, i was just wondering.

3. i'm wondering how everyone else is specifically creating cards. if i just run the program automatically, it seems that i'll end up with a lot of truncated sentences, since naturally the program doesn't know where sentences begin and end. i've been using aegisub to manually join short parts together (parts of the same sentence and also short sentences that make more sense when kept together). is this what everyone else does? or am i missing something?

just as a simple example:
(ユキノ）27歳の私は
15歳の頃の私より少しも賢くない
these two appear on separate lines and wouldn't make sense unless joined.

part of the reason i'm asking is, i was using Nayr's Core5000 deck, and i'm just wondering why no one has yet created a massive deck of assorted sentences. if you were looking for example sentences for a particular vocab word, you could then just search the deck, and maybe find a few examples if the deck were big enough. i mean, the sentences in Core5000 are one sentence per vocab word, which is certainly not ideal. and it's not like they've been carefully constructed to try to illustrate the meaning of the words or anything like that (at least i don't think they are?), they're just sentences that use the word. it seems to me that a lot of effort would be saved if some kind of pre-made deck could be created. or maybe i'm totally mistaken? my own subs2srs deck is still small right now (only like 350 cards, some of which are not useful) so maybe there's something i'm not getting because i haven't used the program enough yet.

Thanks for reporting possible issues. Haven't looked into them yet.

>3. i'm wondering how everyone else is specifically creating cards. if i just run the program automatically, it seems that i'll end up with a lot of truncated sentences, since naturally the program doesn't know where sentences begin and end. i've been using aegisub to manually join short parts together (parts of the same sentence and also short sentences that make more sense when kept together). is this what everyone else does? or am i missing something?

I manually retime all my subs to be perfectly aligned with the audio of their dialogue. It takes hours, but my OCD means that can't stop me, and it makes them more effective as listening cards. Also gets me lots of time rewatching each line with essentially a timed transcript for me to read, so I don't think the time goes to waste either.

qweasdzxc Wrote:2. has anyone noticed that the left & right audio channels seem to get swapped when making cards? or is it only happening to me? (i notice because i always use earphones). it probably doesn't really make a difference though, i was just wondering.

Never mind this one. It only happens when the audio is played through Anki (possibly this is only happening for me), it's normal when played outside with a separate player. (I noticed just now because one card had audio coming mostly from the right side. But when I hit "r" to replay the audio, the sound then switched to the left side. I guess it's just an issue with Anki)
---
Aside from that, I've noticed before that I'd sometimes get an error saying the number of fields is wrong when importing the .tsv file. It says "had 7 fields, expected 6" or something. I figured out that it's just because the subtitle file sometimes contains tabs inside the text. Assuming the tabs aren't actually being used anywhere important in the subtitle file, you can just use notepad to automatically replace all tabs with a space or two instead (you have to copy&paste a "tab" into the "find what" box). That's what I did anyway. Possibly the subs2srs program could check to make sure there aren't any stray tabs inside the subtitle text, though I don't know how often this occurs. It appears easy enough to remove the unwanted tabs manually, if you know they're there. I didn't know before so I was confused when that error kept coming up.

Does anyone have backups of the Death Note decks? links are broken on the wiki page.
Also, i reckon the wiki page isn't updated anyway, does anyone have decks of any of the ff:
Angel beats, Bakemonogatari, Code geass, Darker than Black, Death Note, LoGH, Hayate, Howls, Madoka, NGE, Spice & Wolf, Pandora Hearts, Spirited Away, S.E. Lain, S;G, Toaru Majutsu/Kagaku, Totoro, Tatami Galaxy
Well... I might be asking too much but I'll upload my own decks when I become sufficiently advanced

polyturn Wrote:Does anyone have backups of the Death Note decks? links are broken on the wiki page.

I don't have decks for any of those, but I just made a deck for all episodes of Shingetsutan Tsukihime, if anyone wants it: https://www.mediafire.com/?d1m0hhwm4z2h8s4
I joined some subtitle lines together and what not (some may have ended up a bit too long though). Deleted some of the easier/useless stuff, so 544 cards. copy&pasted definitions onto some of them (there might be a better way i'm unaware of?). you might want to edit the card template also.

i haven't used subs2srs that much yet. perhaps others who've been learning longer have something more/better to share.

Any chance you could add a .flac function to the audio ripper? Would be much appreciated.

polyturn Wrote:does anyone have decks of any of the ff:

Anki deck for となりのトトロ。
http://www.mediafire.com/download/v34k7m...otoro.apkg
99 cards. Mostly only kept cards that might be useful for vocabulary. Added definitions to most. Also ignored some lines from the old woman as she seemed hard to understand (for me anyway). All of the audio is short, unlike the Tsukihime deck I uploaded above, where some clips were way too long.

I used the jp subs from kitsunekko.net, which were retimed to remove the commercials. Some of the lines are doubled up or overlapping in spots, but otherwise it should be useable if anyone wants it.
http://www.mediafire.com/download/xc7yb0...etimed.srt

Thanks guys.

Might seem like an odd feature request, but it'd be great if there was an option for "only grab first/last X millisecond of audio for each clip" or something. This can be aped by using negative pad timings, provided the subtitles cover a reliable period of time, every time. But most don't.

Again, I realise this might be an odd request, but it would come in handy in instances where there may consistently be extraneous audio, such as vocabulary lesson videos and the like, where the vocabulary and explanation are featured on the same subtitle line.

I realise subs2srs isn't designed for digesting video lessons though, so I understand if this isn't implemented.

nvm

Dialogue: 0,0:01:35.56,0:01:38.26,Default,,0000,0000,0000,,昔むかし　山奥の村に　仲のいい→\N
Dialogue: 0,0:01:38.26,0:01:41.62,Default,,0000,0000,0000,,爺さんと婆さんが\N暮らしていました。\N

With the default settings, these lines wont get combined even with the newest patch. Any idea how to get the new feature to combine these?

qweasdzxc Wrote:

polyturn Wrote:does anyone have decks of any of the ff:

Anki deck for となりのトトロ。
http://www.mediafire.com/download/v34k7m...otoro.apkg
99 cards. Mostly only kept cards that might be useful for vocabulary. Added definitions to most. Also ignored some lines from the old woman as she seemed hard to understand (for me anyway). All of the audio is short, unlike the Tsukihime deck I uploaded above, where some clips were way too long.

I used the jp subs from kitsunekko.net, which were retimed to remove the commercials. Some of the lines are doubled up or overlapping in spots, but otherwise it should be useable if anyone wants it.
http://www.mediafire.com/download/xc7yb0...etimed.srt

Thanks for the deck, how did you make the notes? Manually or generated with something?

I have just released version 29.4 of subs2srs.

Download subs2srs v29.4 via SourceForge

What Changed?

● Fixed bug where .ass parser did not trim text after deleting \N. This caused the "join lines" feature to not function as expected.

● Subtitle parsers now remove tabs from the subtitle text.

cb4960

xtreme1 Wrote:Thanks for the deck, how did you make the notes? Manually or generated with something?

manually, if you mean the stuff in the "notes" field.

Weird feature request but you never know.

I have 2 trs files for the same piece of audio.

One is for each individual word. The other is for each sentence.

I'd love it if I could use these two files to generate Vocab + Sentence pairings.

That is, I could generate a file where each individual word is then linked to the sentence within who's timeframe it occured, which also has its audio extracted based on the sentence level timings, and all these are used to generate notes with the vocab + vocab audio and the sentence + sentence audio all in one note.

Then I could have each individual word in the piece on its own card, and then in the same note, have the sentence text and audio to go with it on the back.

If that makes sense.

Just an idea, no idea how feasible it is, or how willing you are. But it'd be interesting, and very helpful to me.

This tool is simply amazing, with or without this feature.

NinKenDo Wrote:Weird feature request but you never know.

I have 2 trs files for the same piece of audio.

One is for each individual word. The other is for each sentence.

I'd love it if I could use these two files to generate Vocab + Sentence pairings.

That is, I could generate a file where each individual word is then linked to the sentence within who's timeframe it occured, which also has its audio extracted based on the sentence level timings, and all these are used to generate notes with the vocab + vocab audio and the sentence + sentence audio all in one note.

Then I could have each individual word in the piece on its own card, and then in the same note, have the sentence text and audio to go with it on the back.

If that makes sense.

Just an idea, no idea how feasible it is, or how willing you are. But it'd be interesting, and very helpful to me.

This tool is simply amazing, with or without this feature.

Thanks for the suggestion. I imagine that feature would not get much usage though do to the effort involved in timing individual words.

Can anyone share some non children anime decks?


I have huge problem. I find files (anime, jap and eng sub files) and try to import it, everything get s completed messed up. Subs doesn't match each other at all.

Ok. I managed to get my hand on files I wanted but now I have another problem.

What do you do if audio is out of synchronization. Everything is fine. Japanese and English subs match but audio matches only for the first few cards.

For example, I am trying to make Code Geass episode 1 deck. First few combinations are ok, but after that some cards only have background music or text from different part.