I don't use the reading field on the front, because I think it lacks
context and there are too much homographs/homophones.
It might help you to understand more what you listen, but I think it
needs more context to make sense, otherwise it become a time sink as
you described.
What I'm doing is creating three cards for each note.
All of them show the kanji on the front side, and one of them ask for
its reading, one of its meaning, and one both.
I still show both answers on each cards to help me remember the other
one (reading or meaning).
I found it good enough to force me remember both, because when I had
just one card showing all the information, I usually just memorized
one of them.
I'm also not interested in using the kanji on the back side, because
I'm not interested into writing them by hand, and when I type them the
suggestions are enough to pick the right one.
If you want to be able to write them by hand or have a stronger grasp
on them, I suggest using a different program/app to force you write
them and check all the details.
Obenkyo (
https://play.google.com/store/apps/details?id=com.Obenkyo)
has a test to write the kanji based on the reading/meaning (you can
choose one of them or both) and it has a weighted score to determine
which ones should be tested (as anki it tries to show you the ones you
fail the most more often).
It also allows you to introduce just a few new ones to be tested and
has some bigger categories using both JLPT or school levels.
When I card has two very different reading for the same kanji, or
different kanji for the same reading and meaning, I usually just
create a card for each and add the other reading/kanji below the rest
of the card (I just put it in the meaning field, no extra field
needed).
If you want to keep using the reading on the front side, I recommend
you to use the anki desktop to flag all duplicate cards based on the
reading field, export and merge them using a spreadsheet.
Then I would create a new template for homographs which would show the
reading and one of the kanji/meaning and ask for the other.
It is not ideal, but I think it might be enough as the spacing
algorithm will show you the failed cards more often.
I tried creating some cloze cards for other subjects, but I found even
the Anki desktop interface a little cumbersome when handling more than
one cloze per note.
I ended up using Emacs and created some keybindings to add and
manipulate the {{c1:: and }}.
If you want a simpler solution, I suggest using a simple text editor
and instead of {{c1:: just type XX (or something else that will not
appear on your text, it might even be some Japanese characters that
never appear together, such as '。ー') and after you finished marking
all your cards, you can search and replace them for the correct one.
It works well when creating several cards, but it still too much work
for a single card.
> To view this discussion on the web visit
https://groups.google.com/d/msgid/anki-android/26cc127a-32ce-47e4-a64b-b5e972f7cbb7%40googlegroups.com.