I just read this article
http://www.futurity.org/top-stories/to-learn-a-skill-practice-while-you-sleep/
Basically they did an experiment where two musical tunes were taught to a person and when the person went to sleep that person was exposed to just the first tune. The EEG showed the brain replayed the second tune as well. It could be great to review kanji mnemonics by playing the keywords while we sleep. However I'm not sure I could sleep while the kewords are spoken.

Calling this a great way to review anything is quite far fetched. It seems to be the equivalent of sitting on your couch with some electric device strapped to your ass. There might be a benefit, but so slight it's negligible.

It sounds cool! However, you say "two musical tunes were taught to a person" ....so do they know it already before they do the test, so they aren't really adding to their knowledge while asleep?

templerschaf,
I never said "anything" but I did say "great". And the reason I did say "great" was it makes what was previously thought impossible possible. The time we spend sleeping can be used productively towards our goals now. I'm not saying we should rely only on this method. Rather I feel it adds to what we already do.

Catherine,
You are correct. It is good for reviewing but not studying new materials.

I remember reading about how sleep is crucial to learning anything and part of the reason is the brain seems to process and do housecleaning on newly acquired material while we sleep. So, we've all been learning in our sleep the whole time, it turns out!

I agree, though, that the prospects of tweaking this time somehow are pretty interesting.

In the meantime, consider the following: the head first series of books from O'Reilly, which claim to be presented according to principles designed around the way the brain learns, suggest you should study shortly before going to sleep to ensure that the target skill will be worked on during the night.

Maybe schedule a bit of skritter time every evening before bedtime.

I used "anything" as shorthand for "Melody, Kanji, etc." like "anything of use".
And I remain skeptical of the idea of "learning or reinforcing memory in your sleep with headphones" at least to a useful degree, which btw. has not been shown in this study.
From what I have read the Effect was minimal and it could be attributed to something else, for example the non blinded environment it was conducted in (at least I couldn't find any mention of that).
I don't really want to argue about this, just don't get your hopes up, the future for this idea remains slim to none.
Learning is hard and if you plan on doing something extra, do something (skritter) before you go to bed, this is probably much more effective.

I am also quite skeptical about the idea that this would be useful for language learning.

From their results, there was a small but statistically insignificant improvement in recall when the tune was played during sleep.

Even if this effect is really there for learning melodies, assuming that the effect will also be present for much more complex information with many relationships such as a language is a big assumption.

When I go to sleep..well.. I want to sleep.
It would not be useful for a person who sleep like a log like me anyway.

It is interresting in a sifi kind of way. One day they might delvelop a hat that we can put on and then wake up with a new language. There go the fun of learning and reviewing.