I find that part very hard - for all of the 25 years I’m freelancing… 
What I try is to get as much information beforehand and an as complete picture as possible.
Especially with new clients.
And it depends a lot who you are working with:
An agency may have a budget planned already and sometimes I just ask how long they want me on the project (daily rate). In some cases I would have guessed much lower but the agency knew their client and knew that it would take several iterations and how they planed to go about the job. In many such cases you could do something more quickly, but if the budget is there, you may be able to go into more detail or improve the design over a longer time.
Some agencies are cool like that, others get a dozen estimates and take the cheapest. I usually can live without those
A small company with less experience on the other hand may not know anything about the work process itself (I see that rather often) and may not calculate “Design” and “Ideas” as real things or “work”. So sometimes it is about education of the customer (“If you think of Apple products, do you think the design quality makes a difference?”).
I once had a request many years ago about what “10 minutes of animation” would cost (they did not know the content yet, but they liked the then hot early Pixar movies…)
If you did similar things in the past, you may be able to take that as a measure stick, but I personally love to do new things I never did before, so that usually does not work too well and one client may be extremely different from the next…
So I sometimes take my personal learning experience out of the calculation, since somebody else may be much faster.
In other cases when there simply is nobody else to do it and finding out how to do something is actually part of the job, it’s the other way around.
So I adapt my rates to the customer at hand (a big car company pays more than a small startup), to the length of the job (short/fast jobs are more expensive than longer ones), if it’s tricky or simple, if it includes programming or not…
And then there is also a part gut feeling. If the first contact is rather complicated or arrogant or bean-counterish or otherwise not supportive, then the work will probably be the same which translates into higher rates.
If on the other hand I know somebody is very clear and straightforward, fun to work with and pays quickly then I may reduce my rate.
I may also take on a job that I’m personally interested in that I otherwise would not consider financially.
3D rendering is no longer as much of a deal as it used to be. With a renderer like Thea Studio and good CPUs & GPUs, things that used to render for hours or days now render in just some minutes and realistic materials are much easier to do than in the early days of 3D, so it depends on what you do, how large the renderings are and if you are doing animation (still much more effort than stills if it’s more than a turntable).
I sometimes deliver just renderings, then that is the actual thing I sell and the customer does not get the 3D data (which is a value in itself and not automatically included - you don’t get the recipe if you buy a bun at a bakery), at other times I use renderings more for communication and the main thing is the 3D file for manufacturing, then the rendering is not extra and I don’t spend much time on it. And sometimes it’s both, high quality renderings for pitches and customer communication plus the final 3D data for manufacturing…
Really depends on the job and customer at hand.
If you just start out, you probably will make some mistakes, some project will take a loooot longer than expected, some customers may turn out much uglier than you thought and some things will be much harder to do than they originally looked, but that’s how we learn…
A famous banker was asked how he got so successful: “Good decisions”.
But how do you make good decisions? “Experience”.
And how do you get experience? “Bad decisions”.
Cheers,
Tom