3D Computer Animation to Physics?

[TdePayen]

I have a BFA in computer animation and currently work in that field, specifically game development. Since I was in high school, I've had an interest in Physics. Lately, I started brushing up on my math (the highest level I took in college was Precal) and decided that I wanted to start studying Physics (and I found this forum). Initially, I was only doing this for fun, but then I had the (perhaps somewhat crazy idea) that maybe I ought to try to go to graduate school and get a PhD in Physics. There is a state school near where I live that has a PhD program. I realize that I might have to take some undergrad classes in addition to studying on my own.

The reason why I'm posting here isn't to ask some general question about whether I can do this or not (I'm the kind that you really shouldn't stand in my way if I'm after something) or whether I'm too old (some of these threads have caused me to ask "Is this Logan's Run or what?").

Anyway, I digress. My questions are: what specific area(s) of Physics, if any (or maybe all?), would benefit by having knowledge of 3D modeling and animation software? (such as Autodesk Maya or 3ds Max) 3D software is useful for simulating things, and this sounds like something that might be useful in many areas of Physics. What kind of simulation software is currently used in Physics research? Thanks!

 

[ZombieFeynman]

I think it's unrealistic to expect that you "might have to take some undergrad classes in addition to studying on your own." It is very likely that in order to get into graduate school, let alone succeed, you will have to get a degree in physics.

I do computational physics and have never heard of or used the animation tools you mentioned. I (and many colleagues) primarily code in Mathematica or C++. We also use Python and it is not unheard of to use Fortran.

I strongly suggest you look at the syllabi for the typical BS in physics. It usually includes at least: a year of intro physics, a thermodynamics course, a modern physics course, a year of Intermediate Mechanics, a year of Electrodynamics, a year of Quantum mechanics, and a smattering of elective physics courses. It also usually includes several math classes (the calculus sequence, linear algebra, and differential equations). Students who do well in graduate programs usually have experience in undergraduate research.

I think it is plausible that you could pursue a PhD in physics, but I don't think it is as trivial as you may think. I recommend you convince yourself that this is not a whim and it is worth the time, effort and money.

 

[TdePayen]

I don't think it's trivial. I'm looking at two years of intense self-study with the idea of possibly getting conditional admission. (The two years is because I'm paying off my student loans from my previous degree, and then save money so I don't have to take out student loans again). The program I'm looking at says that if your degree is not in Physics, then they prefer that you have 12 credit hours of upper level course work in Physics, and if you don't have that they do conditional admission where you take certain undergraduate courses. I'd imagine that they want you to take the core classes for the undergraduate program, Classical Mechanics, E&M and QM. I was planning on studying like mad for the GRE and Physics GRE so that I would have very good scores. I also understand that I need all the math, especially Calculus. I was going to go through all of the math and physics at Khan Academy.

Why do you (or others) think it's necessary to get an undergraduate degree in it first, as opposed to just taking core classes? I mean, the prior would be ideal I understand, but the time involved might make it not very justifiable for me at my age, Logan's Run jokes aside. We'd be talking probably six years for the undergrad, because I couldn't just quit my job to go full-time, since I already have a Bachelor's I wouldn't qualify for any aid whatsoever. And then another six to ten years for the PhD. Not to mention, it's going to be two years before I can start anything. That's possibly eighteen years and that would put me in my fifties before I'd be finished. I would do it just for the learning aspect of it, because I think this whole thing sounds interesting, but I'm not sure that I can justify that length of time to get into something else. I thought it sounded from the program description like something I could do. Maybe it isn't. Worth it and justifiable are two different things.

Anyway, about the topic at hand - I'd never heard of Mathematica, but it sounds like it does 3D kinda things. I'll have to look into it, if just out of curiosity. I appreciate your advice.

Maybe I have no idea what I'm getting myself into... wouldn't be the first time

 

[TheKracken]

If you are serious about this, then prestudy all of your precalculus material and enroll in your local community college calculus course. From then if you still want to go with it, take more math and physics. After that you will have the first two years knocked out. Maybe then you can spend a year doing a few upper division courses and become admitted into a graduate program. This is out of my area of knoloedge though. Currently I am a community college student and there are pleanty of people here who already have degree's.

Good luck to you, if you are serious, don't focus on skipping everything, take the courses and get through them! Get to where you want to be.

 

[ZombieFeynman]

The amount of physics courses the typical grad school bound physics student takes is a lot more than 12 credit hours, even out of the upper division. Studying all of undergraduate physics and math by one's self is very very difficult. You're missing out on a lot of benefits like interactions with classmates and professors. More importantly, getting the degree provides you with repeated tests to prove your knowledge, to yourself and others. Self studying is easy when you're a senior grad student or post doc (I do it all the time) because you have the tools and the knowledge to make it work. But even years after starting my PhD, I find it immensely easier to learn in a class environment with a lecturer, assigned problems sets I must do, and numerous other students to bounce ideas off of.

I am not sure how applicable graphics design is for research, but my guess is not very.

I'm not trying to discourage you. Good luck.

 

[homeomorphic]

Seems like you ought to consider getting into physics simulation for games or movies or something. That's what your skills would probably transfer to, more than anything. I'm surprised you hadn't thought of that, rather than a physics PhD.

I don't think there's such a thing as being too old, but studying physics at the PhD level is a very serious undertaking, to make a very severe understatement. It's not a good idea for all but a tiny, tiny percentage of the population, regardless of age. One exception to that might be the sort of person who is independently wealthy or has an established career to fall back on, such that they have nothing to lose (other than having their life completely consumed by grad school for a while, that is).

 

[Vanadium 50]

And then another six to ten years for the PhD.

That's full-time. (Median is something like 6.5 years). If you do it part time and can spend 16 hours a week on this, it would be something like 20 years to finish. Which is why graduate schools, as a rule, don't admit part-timers: it ties up a slot for decades.

Also, a thesis that was fresh and interesting 20 years ago might not be relevant, interesting or even unsolved 20 years later. Finally, universities have caps on how long it takes. For example, where I graduated, classes to be applied to the PhD had to have been taken no longer than 10 (or maybe 12) years before the PhD.

 

[billthegeek]

I'm familar with those 3D modeling softwares you mentioned. But I have never seen them used in the context of physics. But I know there are engineering firms that use 3D modeling for prototype design.

Why not do a master program in scientific computing, Data Science or Mathematical Modelling? PhD is a huge commitment. Masters might be a better fit for you in terms of career benefits vs duration of study. There is also a strong demand for data scientists in the market.

You can read up the curriculum of master in scientific computing here:
http://math.nyu.edu/degree/ms/scicomputing.html
http://www.maths.ox.ac.uk/current-students/graduates/courses/mscmmsc

 

[TdePayen]

That's full-time. (Median is something like 6.5 years). If you do it part time and can spend 16 hours a week on this, it would be something like 20 years to finish. Which is why graduate schools, as a rule, don't admit part-timers: it ties up a slot for decades.

Also, a thesis that was fresh and interesting 20 years ago might not be relevant, interesting or even unsolved 20 years later. Finally, universities have caps on how long it takes. For example, where I graduated, classes to be applied to the PhD had to have been taken no longer than 10 (or maybe 12) years before the PhD.

I was hoping to be able to do the graduate school full-time, because they often employ people as TAs and such. I have teaching experience on the college level. I can see how getting an undergraduate degree might help my prospects of that. The school I'm looking at, the full-time PhD is supposedly 4.5 years, with a maximum of 8 to finish it. They do allow part-time people in their program however. Maybe it wouldn't take as long as I suspect even getting an undergraduate degree, because I might be able to get through the graduate program faster with better support (i.e. fellowship and stipends). You definitely give me something to think about.

Why not do a master program in scientific computing, Data Science or Mathematical Modelling? PhD is a huge commitment. Masters might be a better fit for you in terms of career benefits vs duration of study. There is also a strong demand for data scientists in the market.

I didn't even know about that, so thank you for the information. I'll look into that as well. I think that Maya would be interesting to try to use to model simulations. It has the ability to code in Python built into it. Perhaps it would be more geared toward teaching people rather than for research.

If you didn't know, the game development industry is a bit like wild-catting, to make a severe understatement. The only difference is, in wild-catting you have a better chance of making a fortune. I'm looking for something that is difficult to be out-sourced and is at least stable on average. But more than that, I'm trying to find something to do that I'm interested in.

 

[homeomorphic]

If you didn't know, the game development industry is a bit like wild-catting, to make a severe understatement. The only difference is, in wild-catting you have a better chance of making a fortune. I'm looking for something that is difficult to be out-sourced and is at least stable on average. But more than that, I'm trying to find something to do that I'm interested in.

Yeah, I know about that, since I've considered it myself. I think, for me, making indy games would be the best option, but I'd need to be independently wealthy or have an established career to fall back on in order to do it safely. If the academic climate in physics didn't look so forbidding, I might consider doing a 2nd PhD in physics and pursue that, but my hunch is that it's considerably worse than the video game industry in that regard. I would choose video games over anything in academia in a heartbeat--there's no way it could be anywhere near as hard, at least for me. Maybe it's different for someone with more of a knack for cranking out publications at lightning speed and teaching without having to spend hours and hours preparing for each class, but for me, I know that would undoubtedly be much harder than anything in game programming. Anyway, I'm more inclined to go for standard software developer type jobs.

 

[Rika]

Wow, is gamedev collapsing in US? In my country if you are good programmer/artist/designer you can land a job that pays reasonable money without many problems. It's not that easy like standard software developer job but it's not that hard either. And way easier than academia.

 

[homeomorphic]

Wow, is gamedev collapsing in US?

I don't know if it's collapsing, but it's very competitive. They say that usually, you have to work more hours and for lower pay than other programming jobs, plus it's harder to land a job. I highly doubt it's as competitive as tenure track positions in academia, though. On the plus side, I hear that qualifications aren't such a big deal and it's more about what you can do than what degree you have.

In my country if you are good programmer/artist/designer you can land a job that pays reasonable money without many problems. It's not that easy like standard software developer job but it's not that hard either. And way easier than academia.

Actually, neither is easy for me because my PhD is in math (and not particularly applicable math), so my programming, while not terrible, is not exactly top-notch, and I think people tend to be suspicious of random PhD career-changers. They can very easily be seen as overqualified or underqualified or both at the same time. I think I will eventually find success in the coming months, but if not, I will most likely be headed back to grad school for a masters in computer science.