I am currently building a pair of rear horns for some 3.5" speakers i have based on Marting King's mathcad worksheets. My design has a length of 120", throat area of 2.25 in^2 and a mouth area of ~144 in^2. the predicted response is fairly smooth to 50hz. however, i've been reading some other information online and a few sources state that the required mouth area is related to wavelength of the expected cutoff frequency. so is the mathcad model wrong or is the required area just a suggestion? I didn't see anything about this in the papers on horn design that King created.