Just going back with the previous question,
If you had something like

But if t > 0,
That means you would have to make the RHS - 4y, right since you have modded the argument of the logarithm?
What I don't get is why:
If it's
You don't change it to -loge(2)? Is it something to do with transformations?
No - in your first case, if t<4, then your equation becomes
=4y)
.
In your second case,
)
.
You keep using the word argument, and I think that's getting you confused with complex numbers. So, I'm going to call it an input.
If in your complex number you have a negative length, you make it positive and put the negative into the argument.
If in your logarithm you have a negative input, you cannot evaluate the logarithm - it's as simple as that.
If the input for the logarithm has a modulus sign, you simply make the input positive - that way, you can evaluate the logarithm. You do not take a negative out the front like you would a complex number, that is the equivalent of taking the reciprocal of your input - which seems pretty silly, doesn't it?