You should always have standards that span the area where your samples are. So if you except samples to be maximum 10 fold downregulated, you don't need to do 0.01 dilution in your standard curve.
Usualy more wider dilutions are used because they span all possible concentrations, but if diferent dilutions have different amplification efficiency (due to inhibition for example) it may skew the slope of the standard curve. In that case for concentrations around 0.5 - 0.25 the standard curve with 0.5 and 0.25 dilutions could be more accurate. It really depends on the range of your samples.
thanks again !