I generally use Cohen's article (Cohen, J (1992) A Power Primer, Psychological Bulletin, 112 (1), 155-159.) to answer this type of question but there are a number of online calculators as well. On page 158 of the Cohen article is a table that you use but you will need to know your research design (eg t-test, anova, regression). So let's assume it is a t-test, you go down the column on the left hand side, to mean difference, across the top to alpha 0.05, medium effect size, and the sample size is n = 64. That is 64 per treatment condition. Since this is a t-test, that means you will need 64 x 2 or n=128 in total in your study.
