Most people prefer number 1.
You can easily find pins/size compatible LCDs so it will be easy to put another one instead.
@bronxsound : you should definitively give a look at the leaflabs board
@mystran: I would have used floats without hesitation if I could.
A sin wave has values between 0 and 1 all multiplication do not need any adjustment.
When you get your integer values (16 bits) and multiply them together you always have to do some bitshiftings so that the result is still 16 bits.
Moreover the order how you do your arithmetic and bit shifting really impacts number of significant bits in the final result.
With float you don't have to think about that.
Even when reading from the wave table it does not work with integer. If you do that the frequency of your waveform won't be exactly the one you asked for...
And in FM the ratio of the frequencies modulation/carrier is very important and must be as accurate as possible.
I used fixed point arithmetics for that, with as many bits as possible in the decimal part.
I think my "integer FM engine" is finaly interesting and is part of the sound of this sound module
Xavier
