`math domain error` during quality computation

[fvirdia]

This is more a suggestion than an issue. Sometimes, line 87 of tools/quality.py may result in a ValueError: math domain error when floating point precision is too low (eg, if using "d" rather than "dd" for a large enough basis), due to trying to take a log of 0, I believe.

Such exception could be caught and relabled, to help more easily identify the problem.