Ok, the hearing thing being addressed in comment, I'll focused on the main topic here.
In an audience (unless you're in a really specific conference on the specific subject), you will usually have many kind of persons. Many of them aren't expert in the specific details of your work. So, you should be able to give details on a lot of things, without getting in the very specific and technical stuff.
When someone ask a specific question that would require a lot of technical aspects to be mentioned, you always have the possibility to give a short answer which explains the gist of the answer and/or explain this is too technical for this particular talk, and offer to meet with the person afterwards for a more detailed talk.
In the same idea, when you are put in front of the question you don't have an idea about, you can either ask for a few seconds, think about it and give your input on it, and/or just admit this is a perspective you haven't thought about yet, possibly try to give it a shot. It is OK to not know everything, it is also OK to not know all the methods that exists, if you did your state of the art correctly and they are referring to something that is not within the scope of your work. Just be honest about it, and propose to discuss it after the talk, just don't close the door or try to bluff it.
Obviously you should have done your homework and know about the reference papers in your field, why they apply or don't apply to your work. Also, what are the obvious questions that will be asked about your presentation (this can be prepared in rehearsal session, and even driven by how you present things).