Point taken...
I did the same test but setting the lock to indipendent.
Now the two graph are identical.
In fact , setting the Q to 0.1 i quite like setting a very large band, so almost all the spectrum is attenuated.
Hope helps.