# Is it a bug? when I calculate Mantle Bouguer anomaly by gravfft

I follow the example in gravfft to calculate Mantle Bouguer anomaly. The second step is to calculate the assuming constant crust with 600 of density contrast. But when I use 6 km crust and 3 km crust, the results are totally the same! That’s impossible. Who has the same problem?
Many thanks!
··
gmt gravfft bathy.grd=nf/1/-6000 -fg -D600 -Gmoho_g_6.grd
gmt gravfft bathy.grd=nf/1/-3000 -fg -D600 -Gmoho_g_3.grd
gmt grdmath moho_g_6.grd moho_g_3.grd SUB = moho_compare.grd
gmt grdinfo moho_compare.grd

moho_compare.grd: Title: Produced by grdmath
moho_compare.grd: Command: grdmath moho_g_6.grd moho_g_3.grd SUB = moho_compare.grd
moho_compare.grd: Remark: Parker expansion of order 3
moho_compare.grd: Pixel node registration used [Geographic grid]
moho_compare.grd: Grid file format: nf = GMT netCDF format (32-bit float), CF-1.7
moho_compare.grd: x_min: 45 x_max: 53 x_inc: 0.0166666666667 (1 min) name: longitude n_columns: 480
moho_compare.grd: y_min: -40 y_max: -36 y_inc: 0.0166666666667 (1 min) name: latitude n_rows: 240
moho_compare.grd: z_min: 0 z_max: 0 name: mGal
I’m afraid we somewhere broke backward compatibility here. It works if you use the modern syntax `=nf+s1+o-6000` and `=nf+s1+o-3000`. Or even simpler, just use the +o flag; `bathy.grd+o-3000` & `bathy.grd+0-6000` flags.