Massively n00bish question.
Hey guys,
I was wondering since I have just taken my hi-poly model into mudbox, smoothed some edges etc. Brought it back into 3ds Max and I have noticed that my mesh is a bit of a mess now. Not like massively messy, but un-even quads that are strechted more than others in places.
Is that still ok to use when baking it down for normals? or would I need to re-top the whole thing. Or would that only be done for animation purposes?
I was wondering since I have just taken my hi-poly model into mudbox, smoothed some edges etc. Brought it back into 3ds Max and I have noticed that my mesh is a bit of a mess now. Not like massively messy, but un-even quads that are strechted more than others in places.
Is that still ok to use when baking it down for normals? or would I need to re-top the whole thing. Or would that only be done for animation purposes?
Replies
And while (I believe) you're correct in saying that tutorials have basically ruined a generation, I'd say it's more to do with bad tutorials. Everything becomes a giant game of Chinese whispers in the quest to generate as much content for an ad-revenue based youtube channel.
It starts with perhaps a guy, who does a little write up for someone somewhere, probably on a forum, he states something along the lines of "so yeah, that's why I tend to start with a darker red for my base textures". This then gets rehashed to the point of someone doing a video where they tell you that you absolutely must start with a dark red. Because that is all they know, and that is all their viewers will ever know.
Sorry for dragging this more off-topic, I think this is actually a very interesting and possibly-serious talking point these days. I'd been thinking about it a lot for at-least year now.
Edit
It's probably worth noting that this has been happening since before my generation too. I'm only 4 years into the games industry, so still new by a lot of standards, but I've been learning this shit for a good 14+ years now (I think). I still struggle with the idea that I'll be as "smart" as those I learned from, so I'm sure this "diluting" process has been going on a lot longer than we would first assume.
Still though, it is worth noting that year after year the standards of game-art always reaches a new mind-blowing stage.
I feel I shouldnt even approach employment before I get all these rules and correct practises right.
So yeah thanks alot for the replies and no need for the apologies of going off topic. If anything it would have been another rule I would have run into xD.
Speaking from someone you guys seem to be aiming at, I do find that it can be quite confusing on what is expected. I have approached quite a few high up people and I am talking "technical/lead/directors" of places such as Naughty Dog, Crytek and Epic and they all kinda give the same but different views on what to include in portfolios etc. So its all abit all over the place. Unless thats just me.
Oh finally, if anyone wants to add anything to this thread to link people or whatever then feel free
These. For example, advising someone about how a certain sculpting program might behave when using stretched polys during the actual sculpting process could be some beneficial advice for them to know beforehand. But if the only message that gets across leads to them believe that stretched polygons are bad in general, then something isn't being communicated right.
Yeah, I got that a lot. It doesn't help that it's a totally loaded question. There's a lot of "it depends". Oh, and not to forget there are a few art roles you could dive into, which skews things more.
One common answer is to "aim your portfolio at a particular studio"... meaning you'd follow their work and create a portfolio that compliments things they've done. This is immediately "contradicted" by the next bit of common advice though:
"Show variance", to demonstrate a range of artistic ability is always a plus. It's very good if an employer thinks they can get more bang for their buck with you. And of course... to contradict this again:
"Deal your best cards only"... something that I hear a lot of, though have no factual proof for is that we are always judged on our worst piece, not our best. I guess it doesn't really matter, the only thing that does matter is that you get the message and make sure you're only sending the best of your abilities to a potential employer. "trim the fat" as they say.
So there we are, it sounds like a mess... but I guess the only thing you should be worrying about, is the quality of your work. If you're branching out of your comfort zone, make sure you do a damn good job.
So for example you do find that stretched quads do give you a certain (this time) undesired effect and that you should have probably set up the base mesh a little differently. Next time you might actually want that effect (maybe it creates a nice dented used metal effect) so you "screw up" your base mesh and get that effect without much effort. Or you simply learn how to build your base mesh differently to avoid the problem. Or you bake and find that it isn't a problem at all and you don't need to waste time figuring out ways to get around it.
Either way don't be afraid to test things out.
Failure is a necessary component to success. It is only through failure that you will be able to identify success. The trick is to not repeat the same mistakes over and over again, expecting different results. Which is easy to do when you only have a rule to work with...
If you do this I guarantee success! Except when it doesn't work then you're stuck trying to repeat the same failed steps... weeee...
Like Per said, if someone throws out an absolute eyeball them suspiciously because they stink like a noob.
Example:
Noob: ALWAYS SET THE SKIN POSE ON EVERY RIG NO MATTER WHAT!
Guy who's messed up enough to know better: Setting the skin pose is helpful because... bla bla bla. Skin poses are created to get around XYZ problems. Certain rigs actually deal with the same problems differently and those solutions have unique features to them that may or may not help you with the problems you face.
With the noob rule you get a simple rule to follow but you will never know why and are cut off from actually expanding your knowledge to a useful level.
Even if you are a "follow the rules, never deviate" kind of guy, it will do you good to find out why that rule was structured that way. What you find is a lot of absolutes actually aren't.
You can carry around a big book of rules and try to memorize a bunch of specific and weird things and always try to deal in absolutes but honestly when you strip away the absolutes and find out why things function like they do then things get a lot easier to handle. You find out what drives a lot of crazy absolute rules are just a handful of technical misunderstandings of how things work.
Once you figure out what drives the rules you can exploit them and do some really creative things.
So how do you get there? You ask questions and find answers. So instead of:
"should I retopo this because someone said long quads are bad"
You should ask:
"I read somewhere long quads are bad, why is that and how will it effect this model?"
Notice: You could run some tests and find out for yourself but asking isn't a bad thing because you might get more info then what you'll find on your own.
People are afraid of appearing stupid so they often keep their mouth shut and don't ask questions and miss out on an opportunity to learn. If you are serious about learning you will risk the tiny hit to your ego and ask. The more questions you ask the better you get at asking them and the better you get at answering them too.
TLDR
Don't be afraid to try things and fail. Failure helps you find what works.
Don't be afraid to ask questions. How else will you ever learn anything...
Well as has been said before "rules" are mainly guidelines, they can help at certain times and have nothing to say at others.
And you should just test it, it's OK asking for advice but you will always learn more when doing it yourself, maybe you find out that baking the normal map goes fine but the mesh is not suitable for animation,
so then you will think about that next time and maybe retop it before animating it.
Also as long as you work, completing projects not worry to much about "rules" that you have not tested or discovered yourself, you will after a while be able to pick up what tips that is true and what's BS.
Remember that every artist has a way of working that fits themselves, so what might work for one person might not work for you or for me, so do not try and force a way of working on yourself because it works for someone else,
it will only make things worse.
Personally I stopped worrying about all these supposed "rules" I had heard about (mainly from elitist idiots from /3/) when I realised that 3D isn't maths, it's art. So if it looks good and does it's job, then that's all that matters.
Also don't worry about hi-jacking the thread, I mean its all the same stuff that you are putting down and people who are affraid of asking the "stupid" questions can check this out and learn something. I always used to think that if you wanted to know something you should ask, but from experince with lots of different people, some people just really find it a challenge to ask questions becaus of various reasons. So hopefully my un-ashamed ignorance to this topic helps others
So yeah now thats out the way;
I completely understand the whole "dont come running on here asking anything and everything you get stuck on", because I always find the best way to learn is by failure and understanding the failure. However I am slightly pressed for time with trying to find work etc, I guess I thought I better ask before I do the bake thing incase it is a horrible mess and I might have wasted hours for a simple "you should have just checked that box" sort of answer.
One lesson I have learnt, more specifically from doing sub-d modeling; if that is the correct term, is that it is not what you sort of expect. For example, to get a rounded edge I would use chamfer, simple but obviously makes a nasty mess with TEH QUADS. I check out this guys iphone tut on another website that involves video tutorials (not sure if I am allowed to mention it... ) but yeah it was just making the turbosmooth do the work of the edges and the edge loops to control that. Which was a strange thing because you had to visualise where abouts the smoothing would happen, unlike chamfer which is right infront of you. I would guess chamfer would be more used in low-poly modeling, but still for me it was a whole new experince.
Also with the sculpting I agree. I want to learn the workflow... learn my own workflow
Finally just another thank-you for taking your time out to reply and answer my question and adding extra stuff in here. All of it is massively appreciated and hopefully helps others as well
Chamfer is ok when used in pretty specific ways, like on an entire loop or where there isn't a pole... ok yea chamfer just sucks, ha! But one smart polycounter came up with a great little "quad chamfer" script. You have to pay for it but honestly if I did more hard surface modeling I would buy it in a heart beat, but I animate so I work around chamfer when I need to =/
Aside from quad Chamfer you mention the turbosmooth method which can be good at times but you might want to check out the "Subdivision Surface" Rollout in Editable poly (not edit poly, but the collapsed object 'Editable Poly'). What this does is allow you to see the subdivided mesh while working on a low poly cage. This way you can place those edges and see their effect as you work. Swift loop and Connect are quick ways to reinforce edges.
Personally I can't work on a model like that all the time so I toggle on/off "Smooth Result" a lot. So often I created a keybind for it. This keybind doesn't exsist in max so you have to write a short macro script and execute it before you can bind a key to it. Kind of pain in the ass but you end up doing this a lot to speed up your workflow.
The macroscript looks something like this:
Macroscript SubDivSmoothToggle [COLOR=YellowGreen][COLOR=DarkOliveGreen]-[/COLOR][COLOR=DarkOliveGreen]-Name of the script[/COLOR][/COLOR] Category:" MyTools" [COLOR=DarkOliveGreen]--The Category that it shows up under inside Customize UI[/COLOR] toolTip:"SubDivision Toggle" [COLOR=DarkOliveGreen]--The text displayed in Customize UI as well as when you mouse over the button[/COLOR] buttontext:"SubDTog" [COLOR=DarkOliveGreen]--If you create a button it will say this[/COLOR] ([COLOR=DarkOliveGreen]--if subdivsion is on, turn it off, otherwise turn it on.[/COLOR] if $.subdivSmoothing == true then ( $.subdivSmoothing = false ) else ( $.subdivSmoothing = true ) )The top part is the header that tells max this is a macroscript and to add it to max.The bottom part in () is a simple if statement,: "if this is on, turn it off. Else turn it on"
Open the maxscript editor (Main Menu > MaxScript > MaxScript Editor)
Paste the text above and run it once (ctrl-e to evaluate)
This will look like it did nothing but go Main Menu > Customize > Customize UI and change the Category to " My Tools" your macro should be there and you can now assign it to a keyboard shortcut, toolbar, quad menu whatever...
I went into a little more detail about maxscripts below, but I tried to keep it as beginner friendly as I could.
http://www.polycount.com/2011/05/16/maxscript-for-newbies/#more-5510
I also have to wonder if the way games are designed now a days are factoring into it at all. See back when I would spend a lot of time gaming Atari and NES days the games where hard as fuck and repeatitive. Most of them never had an ending they just got harder and harder until you gave up. Being a master meant you put in the effort and hours to figure things out more than anyone else. You didn't have compasses and holographic NPC friends telling you to turn around and follow the glowing arrows to the way point. You got there by trying and dying over and over again.
Now if the player doesn't get there the first time the dev team sees it as a colossal failure and pines over kids throwing down their controller and playing something else. While I wouldn't trade the progress the industry has made for anything I wonder if this approach is helping to kill that 'die trying' attitude?
Ok back on topic...
really after a few days of doing subd work, you get a feel for it and know what things will look like subdivided and smoothed before doing so.
if you have problems finding that command it is in the main UI group under teh editable polygon object category in the keys editor.
(not at a computer to check, and it sounds the same)
You mean like this? :P
I must have been thinking of Edit Poly (modifier) + Turbosmooth, which doesn't show the cage by default. You have to toggle on and off show cage, then if you want to see just the low poly you have to toggle off turbosmooth, which oddly enough I have a script for toggling "show end result" too, so at some point I though it was necessary enough to script it heh.
Haha my apologies if you thought I meant that I made the whole shape in mudbox... no way I made it all in max first. The only reason why I took it into mudbox was that unless there is a better way that is less time consuming; making the slightly rounded body [BELOW]
I thought I would make the body slightly too thick, then I guess as you would sand/carve/hamer :poly121: away until I got the round look I was going for. That as well as I said before, I wanted some practice with mudbox since everyone seems to say (again maybe the she says he says thing) that you need to have some sort of experince within a sculpting package. The actual shape etc I made within max
From a personal point, I dont feel as if I do have the natural "art-farty" skills to make these crazy looking things in mudbox or zbrush straight way. Plus in a weird maybe sligtly anal way... I dig the sliders :poly136:
http://www.blendernewbies.com/tools/subdivisionmodeling/subd_PRIMER/page1.html#introduction
This is a great one, has a lot of information.
Ah a person of great taste. The model I am following is the one herman li from dragonforce used alot. I always liked the... forgot the name; wood finish on the top. However I always prefered by the smallest margine, the RG I think it was... cant for the life of me remember what it was completely, but it had a shaper edge + two black racing lines going down the middle. I always thought it would have looked awesome with EMGS thrown in.
Aha you are very right, I know there is no way I could say really "here look at mah mudbox skillz" because its not really anything spectactular + again it doesnt seem as able or atleast as easy to create shapes than that of zbrush... which annoyingly I do not own a full lisense to. So a student version of mudbox is the only sculpting package I have available and because of the lame internets I have, its not vailable to keep downloading the trial zbrush. Which is a shame because before I moved here, I bought myself some awesome books on zbrush.....
@Mark Dygert, yeah I do get the feeling mudbox is good for touchup work but thats about it. It seems pretty impressive in the way you can get a texture and boom its either a brush or a template depending on your own approach. Not sure if zBrush has that feature or not. However zbrush has things like dynamesh, zspheres and a whole load of cool stuff that would make making the guitar alot eaiser.
In all honesty, these sculpting packages are new to me so I am just trying to include them in my current workflow for the sake of it so I keep dabbing into them when I can
@Jack Ryan. Ah sweet, I'll give that a look now