Conjoint analysis is also a common approach. It requires less specific answers from customers about price points, which people are often unwilling to share.
If you can get enough solid survey responses, conjoint is ideal, but won't give you a lot in terms of a price elasticity curve. That's not necessarily a problem, but I'd recommend a combination of MaxDiff (baby conjoint) and Van Westendorp. Used in tandem you'll get some solid footing on packaging preferences, value props for positioning, and ultimately price elasticity.