In answering the question, be brief, and usually it makes sense to restate the question first. If you don't have an answer, ask to speak to the questioner later.
In general, your colleagues resent you if you take too long, seem to be hunting for the point at issue, or are off base. If you do this often, you will get a reputation you probably do not want. Never say that you are ignorant or out of the field or make excuses--ask the question! I've rarely seen questions that are "stupid," in the sense that everyone knows this--usually half the audience wished they had asked that question.
You will do better as a presenter if in the first thing you say is the main point of the talk. Whatever you would put in the conclusion, usually works better up front. With audiences that are likely to interrupt frequently, it is fine to say the main point up front, and if the questions are too frequent, toask for ten minutes to present before you take further questions.
One kind of question is often very helpful, an epitome. Something like, "Let me know if I have understood your talk. Namely, you are saying that [flying dogs use aerodynamic principles understood by Bernoulli]." Obviously fill in the appropriate summary. This should not go on for five minutes, but be maybe one sentence more.
By the way, these skills are not easy to master, and many never master them. They're simple to say, but hard to adhere to, or even just keep in mind. Most of us teach and we expect an audience, albeit these days in cyberspace. No one has to listen to your talk, or stay, or even be patient, at scholarly presentations. Walking out, doing email, etc is always possible. To hold people's attention, or at least have their attention often enough, is an achievement.
PS: There is an essay by Paul Meehl, in his Psychodiagnosis, about why he does not attend case conferences. Meehl is the source of the MMPI, and more generally of asking "clinical or statistical judgment."