Personally speaking, if one's symptoms do not highly resemble an accoustic neuroma or lesions in the auditory nerve (a retro-cochlear disorder), it is very wise to avoid an MRI, as the noise levels prove it to be a double-edged sword.
I had an MRI performed on a 3T machine last week for my neck (which holds a stronger magnet and is hence a louder machine). Keep in mind that studies have measured the peak decibel level to be 130 decibels (equivalent to a jet engine at close range) in 3T machines; nevertheless, they have become the standard machine to use.
I ended up losing some of my low frequency hearing (no high frequency damage surprisingly), despite wearing specially designed earplugs the techs stated were adequate in protecting me from the noise of the machine. I did not wear the headphones because they only amplified the low frequency noises and were not designed for hearing protection.
The truth of the matter is that it is very difficult to shield oneself from high decibel low frequency noises, and even more so if that noise is accompanied by vibration (as it is in an MRI machine). I don't exactly understand the physics of why, but imagine a neighbor playing heavy bass music at a high volume or a boombox car driving close to you. It is very difficult to block the noise out, and earplugs do not attenuate the low frequencies if they are excessively loud as much as they do the high frequencies.
So the takeaway from my experience is that it is a good idea to be wary of an MRI order- up for ear problems, that MRI technicians are not properly educated in protecting patient's ears, and that even with maximum protection (earplugs and earmuffs) you are not necessarily out of harm's way.