Talk:Royal guards/Gallery/@comment-24255866-20140903195314/@comment-1276314-20140903201855

The magic user must be able to imagine the area around which to put one's magic like your hands, and keep the size of the affected area constant.

In a helmet's case they must imagine the area around which to put grip on and then move it in certain direction. If you happened to imagine your whole head along with the helmet inside the grip, you'd get weird results, but helmets are durable and never change shape so you could form the grip around the top of the fake hair. Grabbing not a whole target wouldn't be optimal for living beings; imagine doing that to a Ursa Minor unlike Twilight Sparkle did in Boast Busters. Another proof of this what I explained above is that TS just placed grip on Rainbow Dash's tail in Lesson Zero and she was hanging by it for few seconds.

Since the in-grip area isn't limited by anything else but one's abilities, the form and size of the in-grip area can be different, so a good enough user could form the grip around just the helmet for maximum grip.

It's not really that hard. And even a magic user could just use one's hooves to remove the helmet, just like Shining Armor did in A Canterlot Wedding Part 1, who is a unicorn. That's easier mostly.