First of all, allow me to be clear about one very key point in learning Japanese...
LEARNING SOLELY FROM ANIME/MANGA/DORAMA/J-MUSIC=NO
What you learn from there is either incredibly archaic, or way overblown and will make you look strange when used in a casual conversation. Imagine if someone you knew spoke like they were in a cartoon or a soap opera every time they spoke. If entertainment media is all you learn from, that's what you're gonna sound like; not very inviting of conversation, if you ask me. Remember, lessons are the base, and media supports and reinforces what you're learning, never the other way 'round.
Moving on, though, like any other student of any other subject, you must be willing to learn from both the lessons offered and the stumbles you'll encounter; because if there's one thing you'll do a lot of it's stumbling. (for a while your Japanese may sound like this, and that's ok. There are no mistakes, only variation, as the Zen saying goes.)
For example, in Japanese, words calls for stress on certain syllables to differentiate them from one another(IE hashi[chopsticks] & hashi[bridge]). Also when you see words with an i or u at the end of them-specifically those ending in ou and ei- that's a sign to extend the end vowel. For example, the word sensei, as you're presently seeing it, is properly rendered into Romaji as such(though I'll render it in Hiragana once I get the basics down, more on that later), but when pronounced its sounded out like this(seh-n-sehh). Note that this also applies when the extended sounds are rendered like this(ou as oo[IE deshoo] or oh[IE Raoh from Fist of the North Star] and ei as ee [IE Sensee]).
With study and proper listening skills-gained through listening to other people's conversations and noting their manner of speaking(you certainly wouldn't do it for other reasons, right?)-you'll pick up on what you need to to carry out the conversation, which leads me to a key aspect of the language and how people interact with one another: reading the air(referring to the context of a conversation, not the air itself. This much I will cover in the next post).