#include <iostream>
#include <vector>
#include <cmath>
#include <algorithm>

int n;
long long F = 2000000, mini = 1000000000000000000;
std::vector<long long> fs, ps, ts;

void eval(const std::vector<long long> &es){
    long long t = 0;
    for (int i = 0; i < n-1; ++i){
        long long w = F;
        for (long long e : es) w = std::min(w, (ps[i]-(e+t)%F+F)%F);
        t += w+llabs(fs[i+1]-fs[i])+ts[i];
    }
    mini = std::min(mini, t);
}

int main(){
    std::cin >> n;
    ++n;
    fs = std::vector<long long>(n);
    for (int i = 1; i < n; ++i) std::cin >> fs[i];
    ts = std::vector<long long>(n-1);
    for (int i = 0; i < n-1; ++i) std::cin >> ts[i];
    ps = std::vector<long long>(n-1);
    for (int i = 0; i < n-1; ++i) ps[i] = fs[i] < fs[i+1] ? fs[i] : F-fs[i];
    std::vector<long long> qs(n-1);
    for (int i = 0; i < n-1; ++i) qs[i] = (ps[i]+llabs(fs[i+1]-fs[i])+ts[i])%F;
    std::vector<long long> es(4);
    eval(es);
    for (int i1 = 0; i1 < n-2; ++i1){
        es[1] = (es[0]+ps[i1+1]-qs[i1]+F)%F;
        for (int b2 = 0; b2 < 2; ++b2) for (int i2 = 0; i2 < n-2; ++i2){
            es[2] = (es[b2]+ps[i2+1]-qs[i2]+F)%F;
            for (int b3 = 0; b3 < 3; ++b3) if (b3 != 1) for (int i3 = 0; i3 < n-2; ++i3){
                es[3] = (es[b3]+ps[i3+1]-qs[i3]+F)%F;
                eval(es);
            }
        }
    }
    std::cout << mini << std::endl;
    return 0;
}