>.1 actually is 10,000 so if your off by one your actually off by 10,000
Log scale is x10 ea whole number, so 0.1 of that range would be 1 percent, no?
(I realize there's more to log scales, weighting diff ea. end of range, etcetera, but generalizing)