possible naive RCAS negative output fix

[KirillAlekseeenko]

Additional multiplier in this fix is needed to account for the difference between e (central sample) and minRingSample if (e << minRingSample) here: // 0 == (w*(b+d+f+h)+e)/(4*w+1) -> w = -e/(b+d+f+h)

due to min max usage instead of individual taps we have: e/(b+d+f+h) > e/(4maxRing) which is safe But e/(4maxRing) < minRing / (4*maxRing) in (e << minRingSample) case So we risk getting such w that result output would be < 0