harsh-harlequin•15mo ago
Trouble setting up custom cards / Mining workflow
I have a custom card layout that goes like this:
1. English word
2. That word in japanese
3. The sentence it was in
4. Word Audio
5. Sentence Audio
When I watch my show to mine vocabulary, I want the workflow to go like this:
Say I have the word 来る in the sentence 田中さんが来たの?.
Then from the parsed dialogue i choose the whole sentence and the word and send it to the card creator in Migaku Dictionary.
Then what I basically want it to do is:
1. Add the word I chose to line 2, with pitch graph and audio playable when clicking it.
2. Add the sentence to line 3, with pitch graphs applied and audio.
3. Add forvo audio of the single word to line 4
4. Add the recorded sentence audio from the show to line 5.
I will take care of the english translation (line 1) myself. I also want to do a card type where line 1 and 2 are swapped.
Is this kind of workflow possible? And if so, is anybody willing to sit down with me and get it up and ready? I am really amazed by this software and am intending to subscribe to a premium plan, but I need to know how to do what I want to do.
Thanks in advance for your help, I hope I am not asking too much here. I tried piecing it all together with tutorials but I couldnt really do it.
33 Replies
foreign-sapphire•15mo ago
The card is just a representation of the note
you can achieve this by cloning the migaku Japanese card and shuffling things around
the note won't change
or you can clone both note and card, then you could even add extra fields
so basically in Anki's Browser, click notes, manage note types
Click add, select Clone: Migaku Japanese
give it a cool name
then click on the cool name you've given and press Cards
a card has a front and a back... usually the front has the "question" and the back has the "answer"
if you're testing your recall of the Japanese word, you could have the Japanese word in isolation on the front and the rest on the back
The migaku card type supports that out of the box... you just need to click the "vocabulary" selector at the bottom
then that note will be shown as a vocabulary card. You can then shuffle the fields on the back of the card to have them how you want
harsh-harlequinOP•15mo ago
thanks for the explanation, i will try this as soon as i get home.
there is still the issue of extracting the single word into a section of the note though. so far i only managed to get the whole sentence into a section.
but as described i also want a singular word in its own section. its hard to explain without being able to show it unfortunately. on another note (no pun intended), is there a way to decide which sections are on the front (or the back) of the card?
but as described i also want a singular word in its own section. its hard to explain without being able to show it unfortunately. on another note (no pun intended), is there a way to decide which sections are on the front (or the back) of the card?
foreign-sapphire•15mo ago
yes, but as I said normally front is the piece of info you're testing on
harsh-harlequinOP•15mo ago
i see. so how about the single word thingy? is there a way to do that?
in the migaku dictionary there is a way to apply certain mining data to fields via a drop down menu like "sentence", "forvo audio" etc.
what i mean by single word is something like the "sentence" option but only the single word i picked.
for example i watch my show and come across the sentence 今日は何曜日. so i choose the full sentence and 今日 in the tick menu.
i want one section to contain the whole sentence and another the single word
foreign-sapphire•15mo ago
you could override it manually, but the idea is that normally it picks the unknown word automatically based on your anki known db
you can get it to fully mine an episode for unique i+1 sentences
I have a slightly different workflow
of course, if you don't have a known db... trying to create a card based on the sentence will flag both 今日 and 何曜日 as unknown and you could unselect the one you don't want
in an ideal world, you want only 1 unkown component in the sentence you mine
and you don't want the sentence too short or too long
that's the Target Word field, by the way 😄
harsh-harlequinOP•15mo ago
thanks for your help 😄
i found the issue
since ive been using migaku for almost 2 years now (means its been a while since i installed it) the migaku cards i used were outdated
the older one i used didnt have a target word section
no wonder i was looking for such a section without success xd
it seems to be working now.
i have one more issue though:
i made a copy of the Migaku Japanese card by going to note menu -> Add -> Clone: Migaku Japanese.
now when i use this clone in the Dictionary and send data from netflix to it, it doesnt fill any of the fields.
i only made a copy as described above, i didnt change anything in the options or migaku options
foreign-sapphire•15mo ago
you need to do the mapping in the dict
so sentence -> sentence, target word -> target word, ...
harsh-harlequinOP•15mo ago
there is no "target word" option in the dict to choose from the drop down menu though
foreign-sapphire•15mo ago

harsh-harlequinOP•15mo ago
ah, i see it works now. thanks for your patience 😄
foreign-sapphire•15mo ago
the initial setup is a bit of a pain 😉
harsh-harlequinOP•15mo ago
it always is when something has so many features xd
ok so there is one (hopefully) last thing.
i made a clone of the migaku cards and moved the section order around. but that didnt change what is depicted on the front and back on the flash card. how do i do that?
per default the sentence is on the front, but i want the unknown word to be what is tested
foreign-sapphire•15mo ago
scroll to the bottom of the note and click on "vocabulary"
foreign-sapphire•15mo ago

foreign-sapphire•15mo ago
on the back of the card
that turns it into a vocabulary card
sentence version:
foreign-sapphire•15mo ago

foreign-sapphire•15mo ago
vocab version:
foreign-sapphire•15mo ago

foreign-sapphire•15mo ago
you can also set it at the card creator level
foreign-sapphire•15mo ago

foreign-sapphire•15mo ago
if you select vocabulary, the card will be created directly as a vocab card
harsh-harlequinOP•15mo ago
nice, thats good to know.
now what if i want to test the english word?
foreign-sapphire•15mo ago
that's not really recommended
it could be done probably, but the issue is that the note isn't storing the English word in isolation
also Anki is really only there to support your immersion... the real learning comes from repeated exposure through immersion
anki preloads 'some knowledge' of the vocabulary
but it shouldn't be the focus
harsh-harlequinOP•15mo ago
okay that makes sense. thanks for your help 👍
i think im finally ready now xd
foreign-sapphire•15mo ago
my pleasure 🙂
also... I'm using the stack in a slightly different way 😄
more in line with "preload some then immerse"
harsh-harlequinOP•15mo ago
is that not the standard way?
it sounds sensible enough to me
foreign-sapphire•15mo ago
the standard way is watch and mine
I mine, I rep then I watch raw 😄
harsh-harlequinOP•15mo ago
thats what i do too actually xd
foreign-sapphire•15mo ago
I'm using the migaku stack to mine and morphman to sort the mess
study plan is amazing 😄
harsh-harlequinOP•15mo ago
morphman?
foreign-sapphire•15mo ago
it reorders the cards in your decks to study them in as close to i+1 as possible
and with the study plan feature, you can make it sort them in order of appearence in your immersion media
you set a target comprehension rate, feed it the subs or the chapters
it spits out a card order, then reorders the cards
harsh-harlequinOP•15mo ago
that sounds very powerful
mind telling me how to set that up?
xd
foreign-sapphire•15mo ago
it has one major issue
it doesn't work with the 23.xx.y versions of Anki
there's a fork/rework of it for 23.xx+ but it doesn't have study plan yet
so yeah... basically now I just use Anki for targeted study to increase show comprehensions
I just finished a cycle to get 95% on a specific show season... I'm about to start a new ~2 months cycle to reach 95% across 18 series
then I'll do another cycle to bring them to 98%