Getting into the gameindustry with blender?
polycounter lvl 5
Hello,
I started to learn 3d art with blender. Unfortunately I read, that the industry standart programm is 3ds max. So I switched for that, made many tutorials on that, used it for a while and got so frustrated with 3ds max that I swithced back to blender, which was a relief!
After being using blender for the last half year I am questioning myself, is this taking me anywhere?!
Sure, I love working with blender and 3ds max is a pain in the ass for me. But to be able to get into the industry I have to bite the bullet. Is this so?
Or is there any hope to be successful using blender? I mean in the end what counts is to bring the models as obj, fbx, 3ds to the game engine.
Are there any positive success stories to motivate me going on with blender, or do I have to face reality that it is taking me nowhere?!
Greetings Andr
I started to learn 3d art with blender. Unfortunately I read, that the industry standart programm is 3ds max. So I switched for that, made many tutorials on that, used it for a while and got so frustrated with 3ds max that I swithced back to blender, which was a relief!
After being using blender for the last half year I am questioning myself, is this taking me anywhere?!
Sure, I love working with blender and 3ds max is a pain in the ass for me. But to be able to get into the industry I have to bite the bullet. Is this so?
Or is there any hope to be successful using blender? I mean in the end what counts is to bring the models as obj, fbx, 3ds to the game engine.
Are there any positive success stories to motivate me going on with blender, or do I have to face reality that it is taking me nowhere?!
Greetings Andr
Replies
Knowing Blender won't hurt, but you need to know at least one of the standard 3d packages used.
In a production environment, artists end up working on each other's work all the time. What happens if you are sick or on vacation and someone needs to open up your source file to work on it?
Basically this. For more reasons than its even worth listing.
Even if you import your meshes to their app of choice to use the tools they write and to do final exporting there still could be a bunch of issues that carry over from the first app. Issues you probably won't be able to trouble shoot so then someone higher up in the pipeline will have to figure it out. It could be as simple as triangle stripping or as maddening as explicit normals.
Considering there is a large pool of candidates that already know the software and don't come with crazy pipeline breaking requirements its easier to just hire someone who doesn't have all that baggage.
Maybe if it was a contract and they said "deliver .fbx .obj and .tga's by xx date, we don't care how" But for in house work its going to be better to know the software they are working with.
The only way I see it working out for a blender only artist is if they are so super amazing that the studio couldn't possibly pass up the talent and is willing to invest the time in training them. But that is a long shot at best, they would have to beat out everyone else in raw artistic talent and even then a studio is more likely to go with someone else.
Tools can be / are usually more integrated than that - and there's usually more to than just popping in an obj and calling it a day.
Having said that I do use things like Digital Raster's Max Switcher to change some of it's navigations and selection schemes to behave more like Maya.
learn max or maya use them at the job and have fun at night with blender !!
edit: apart whats said above about pipeline integration
I'm bored of the Max/Maya shit. They're great tools but It's time to say very loud that we, as artists, have the right to choose our OS, our 3d softwares, and our workflow we want to work on and introduce other 3d software as a standard in the industry.
I know many artists that live with Blender (including myself) so it's definitly doable but you'll certainly have to fight more than if you'd use Max or Maya.
Honestly though, I don't think you'll get a chance to work on Crysis 3 or such title using Blender... You could learn the whole 3d thing with Blender and then just "transpose" your workflow into 3dsmax or maya, if you have to, once you'll have a job but, sadly, don't expect more
I'm sure most of the people above learned 3d with 3dsmax or maya and are fine with it, and are also fine running windows and paying thousand of dollars in softwares and it's great, honestly. But they don't care about us, a minority of people that are not fine with the "standard". So I'd say, no offense, don't listen to them. Max and Maya are surely the main standard in the industry, but using Blender or any other tools, still will allow you to develop your skills and talent and make you a great artist. In the end, Blender, 3dsmax, Maya are just tools, so use whatever you like.
Plus, 3dsmax and Maya are not that hard to master, the whole 3d thing is. The interface is different, some tools are better in one software, some others are better in another one, shortcuts are different but that's all... It's all about personnal preference. In the end, if you want to join a very big company, I don't think you'll be able to avoid 3dsmax or Maya but you will be able to learn and practice with blender
Anyway that's only my humble opinion.
Haters gonna hate.
Good luck
if linux or blender were truly better alternatives, don't you think people would use them?
One other upside to learning them is if you are needing.advice at work your coworkers may not be familiar with blender whereas in Maya they could tell you to go to x, y, and z to solve your problem.
That's a point that falls apart when your being compared to an equally talented artist who already knows Max/Maya. They will get the job.
Max/Maya is free for students so there is no longer an entry barrier. You can be an idealist if you like but it wont put money in your pocket and food on your plate.
That's not a fair assumption. Being "better" doesn't mean people will use it over other software. The reason MS and Apple are the only two names people at large are aware of is because they've put in a lot of money to get in with PC retailers (e.g. Dell, Compaq, HP, etc). MS went through a lot of effort to build customer dependence on their software. I don't see retail box PCs coming with an option for Windows or Linux. It's just Windows (Dell has a few selections that come with Linux if you really look for it). Windows comes with 90% (a very rough estimate) of retail PCs as part of the package and people don't like to learn a new piece of software when they already know one. Linux can't compete with that, no matter how good it is. They worked hard to get people addicted to their goods so that they won't go anywhere else, no matter how good the alternative may be.
The same goes for Max and Maya. They've cozied up to developers and given them special attention in order to get them to make their software standard in their studio. Who wants to scrap an entire pipeline and retrain an entire department for something "better"?
It's all matter of who works the hardest to become the "standard". Once you have the majority dependent on your software, they'll have a hard time changing to something else. Quality is only partly involved.
I say the above being a staunch supporter and promoter of Blender btw, having assisted a number of studios only to find they switch for the aforementioned reasons others have said above.. *shrugs*
I am going to learn 3ds max. I know very well your site katsbits (one of my favourite blender sites by the way!) and if you and all the others say so, then I have to go trough this. I just find modelling much more faster using blender. Maybe with time I get there with 3ds max.
By the way: is there a keyboard-bindings scheme downloadable that simulates the blenders hotkeys in 3ds max? That would make the entry easier and I needn't learn every shortcut from scratch.
Cheers Andr
I do have one caveat, and that is that the Maya viewport navigation system is pretty standard in most apps and max is the oddball so in that case using a program like switcher can help unify the viewport nav. However when I switch back and forth its rarely the viewport nav that trips me up and its some other bind that screws me over. I've jumped back and forth enough times that it just takes a few min to fall into the right groove.
Edit poly has sub-object modes (vert, edge, boarder, poly, element) You want to be in one of those sub-object modes when you copy the geometry, probably in poly mode so you duplicate the verts and the polys, duplicating just the verts will only give you verts no polys.
You can dupe the geo a few ways:
Hold shift and drag it to a new location, release and select "clone to element".
Hit the box next to the detach button, and select "as clone" and "element". This will make it a separate copy but keep it in the object. (editable poly just has a detach button)
That's kind of a fundemental differance between max and blender. In max you mostly work with edges, loops, rings and polys and work less with verts. To create new geometry someone might select an open edge, hold shift and drag that edge out to create new verts, edges and polys. That method is sometimes refered to as "strip modeling" where you drag out strips from various spots then fill in the gaps by either welding, bridging or using step build (more on that later).
To create just verts in max you switch to vert mode (1 on the keyboard) and click create, or shift drag an existing vert to create a copy. To create faces between the verts switch to poly mode, click create then click each vert to draw a face between them. This is a slightly longer process and a bit annoying but most people don't work that way in max except in specific cases. There are some advancements in the graphite modeling tools called "step build" that function more like the way you work in blender so you should check that out.
Hope that helps! In a few weeks there will be less head-on-keyboard moments, it gets better
Having said that, you will need to get familiar with max or maya for pipeline purposes and in many cases you'll learn both of them. In the end it's really not a big deal. The problems you are running into are possibly because you are new to 3d in general. Once you get your workflow figured out, you'll be able to pick any package and get tr hang of it much faster.
Is max the best tool for animation? No. Is maya the best tool for modeling? I can think of a half dozen apps I would rather use. But does this mean they are not the "best" option for a studio, certainly not. You guys can feel free to call me when Blender, C4D, lightwave or any other lesser app comes out with a professional baking workflow, quality realtime shader support, a solid API that allows for easy integration into your inhouse tools, or support for the major 3d engines that we use in the games industry. Please let me know when these tools offer ALL of this, because Max and Maya do, and blender sure as shit does not.
I'm a Modo user, and I will be the first to tell you that Modo, Silo, etc are awesome modeling tools. They are however not full studio solutions, even mentioning modo in this sort of discussion is foolish. Modo is not an app that will ever replace Max or Maya, its complimentary, like Zbrush, topogon, 3d coat, crazybump or whatever. I do all of my high/low/uving in it but I know full well that it isn't going to get me from creation to ingame all in one app.
This is what it comes down to, in the games industry we need professional, flexible tools that can serve the needs of the entire production pipeline and the entire studio. Blender isn't there.
Trust me, I know full well that Max and Maya are not perfect and certainly have their flaws, in fact many in the industry would LOVE to switch if there was something that was honestly better. There isn't some big conspiracy here, if there was an app that was better for us than Max/Maya, we would be using it.
Stick with blender, learn how to do good art, it will carry you much further than knowledge of a specific program.
You can learn Max and Maya later on, but knowing blender will give you knowledge in a program that you can apply to any worksituation. You wont ever know what your future studio will use, it migth be Max, it might be maya or it might be silo and other combinations of plugins, tools and Third party apps.
Earthquake, you are being almost unrealistically unfair towards blender, it is not fully industry supported but it has been fully capable for quite some time already. And it took ages for something like Max to catch up with the uv tools of blender, and when it did, it took lcsm.
90% chance it will be max or maya
80% chance knowledge of one of the two will be preferred over blender, even if you know max and its a maya studio
2% chance it will be silo(just a simple modeling app).
Its really not that big of a mystery.
Max and Maya are two options, not one.
Lol what, suddenly knowing more than one app makes you a worse artist?
Look its cool that you're a blender guy and Its not surprising that you would get a bit butt-hurt over this sort of topic but i'm simply being a realist. I hate Max and Maya equally but still have the common sense to recognize how widespread their use is, how valuable they are to most studios and the benefits of knowing at-least one of them when trying to get a job.
Whether an artist is technically skilled enough for the job is neither here nor there, and sticking with one program or learning a second should have a very very minimal effect on artistic growth. If anything learning a second app should expose you to different workflows and enhance your artistic development, working in a vacuum is rarely productive.
A studio with mixed versions of the same software is a headache enough, a "use whatever you want" studio would be a nightmare.
For straight up modeling work, there really shouldn't be a difference between a freelancer and an onsite artist. Both should be delivering you a clean, editable file in the studio's app of choice at the end of the day.
If you have lax standards and people are just sending OBJs out of random apps, sure, that is a mess. But with clearly defined file delivery standards there shouldn't be any issues, offsite or onsite.
For things like rigging, animation, engine specific tasks like lighting and level design, etc it really isn't possible though.
For every job I've ever had I've used modo(lightwave too if we wanna go back that far), and then the studio's "Mother app" for final delivery and any app specific needs like baking etc. Both onsite and offsite.
The typical argument against this is "oh what if someone else needs to edit it?!?" Well, as long as the file standards are met there are no issues. Having everyone share the same app doesn't by any means guarantee another artist's files are going to be workable. Some people do some really messy shit, more important than the specific app you use is having some set workflows.
No, what I meant was, if you're this artist that worked with maya all your life you'll still have a 50% chance to end up with a studio that is Max only, which is not unlike the situation where you're a blender-user.
But if we look at the numbers, quite a vast majority of people who use max and maya will be amateurs and will most likely never work in the industry, them picking max & maya over something like blender will affect that in very little ways.
You missed my point there, I do recognize their wide spread use, whereever you end up chances are you will end up having to use max or maya but maya artists get hired all the time at studios that use max, and vice versa. And believe it or not, Blender artists can get hired just as much, as long as they are kick-ass artists.
The fault on your side was degrading blender with "facts" that it isn't good enough, we all already know that max and maya has a more widespread use, there's a ton of good posts to read in the big blender thread on polycount.
What thread started wanted to know was if he could get a job even if blender was his main choice of application, and yes, he could, and when he did he could end up learning whatever tools (big plural) the company is using, but in the end the choice of application to use primarily should be up to the artist himself.
And in that scenario, any versatile tool such as blender, max or maya goes.
No, I know that quite well, I use a variety of tools, there are things I like in max and maya, and then there are more things I like in blender, all have problems though, none is superior, two of them are more widely adapted in the industry though, and one of them I can use wherever I work.
But its not the same situation, its a similar situation, but you'd be sorely mistaken if you believe HR at most studios values blender experience the same as maya experience, even if its a max studio.
If you're a Maya guy applying for a Max job:
A. There are likely some ex-Maya guys at the studio already that can help you transition to max.
B. You show that you know one of the two industry standard apps, this is seen as more professional than someone who only knows blender, who will often be viewed as less professional. You may not like this, but from a HR perspective this is the reality of the situation.
Now this isn't to say you CANT get a job if you use blender, but learning one of the main two apps would help you out tremendously. Hell, your application could get passed over entirely by someone in HR simply because you do not display the required max/maya experience. You'll say "oh that can happen if you don't have the correct max/maya experience too" but studios very commonly look for one or the other.
Advising someone specifically stay with blender and NOT learn max or maya is quite simply terrible advice if they want to get hired in the games industry.
Knowing max or maya doesn't automatically get you a job or make you a skilled artist, i'm not sure where any of that is coming from. The % of Max/Maya/Blender artists that are pros or noobs is irrelevant as well. The only situation that is worth discussing is a group of equally tallented artists, but with varying software experiences, the blender artist will always come out on the bottom in this situation. Getting a job in the industry is ultra competitive, sticking with blender only and hoping for the best is nothing short of negligent.
If anyone is seriously considering sticking with blender for their career, they should know where it falls short in comparison to other apps. I know you blender guys have a real chip on your shoulder and would love to believe that blender can do everything that the big boys can do, but its simply not the case.
And knowing both blender + max or maya is even better. This is my main point which you seem to have missed.
At a company with a large volume of applicants the HR person might not be looking over resumes by hand. It will probably be done by automated keyword matching. By not having either "Max" or "Maya" on your resume it might be detrimental to getting past the first phase of applicant culling.
No, I wouldn't say that, my reaction was to you saying that he should not learn blender at all, and go with max or maya, which is just a standard misinformed reply, even when the threadstarter enjoyed blender so much more even after trying out the other applications.
I'm not saying he shouldn't learn the other applications, he can take a few weeks to learn the basics in those, without having to become fluent in modelling in it, and then write it down in his cv, problem solved.
And unless the studio is highly nazi about what you can and cannot install on your work-computer you can download blender anywhere and import your modelling and uv work through the mother-app of the company and continue from there.
So yes, blender can be your main choice of app and you can do very well with it by just pushing your work through whatever program the company uses, and that will not require deep intimate knowledge of the tools.
Here's once again where you come of as uninformed, I'm not one of the guys on blenderartists.org, I've been here for some time, I value knowing a wide variety of tools, and I've seen blender gone from truly shit to awesome, the same with max and maya which until recently required third party plugins to actually be most useful.
Being like you are now is just as bad as being like the guys on blenderartists.org that believe blender is the be all end all, blender is no better and no worse than max or maya, all of them are the most fully featured tools we have right now.
Let's look at something you said earlier:
Blender has full baking support of any type of rendering output, including normals, it has realtime shader support through material nodes or pure handwritten shaders and featured this with realtime shadows before max had it fully out of the box, it has a fully fledged python API that interacts with every part of blender, even more so with the big rewrite in 2.5 which means that every UI element is written in python and fully customizable.
If you wish to make comments like that, please read up on facts first.
No I agree fully with that, I wouldn't say that anyone needs deep intimate knowledge in all of them combined, but he should be fluent in ONE of them when it comes to modelling.
agree with most of what you said but have been noticing some pipeline breaking stuff that stops blender from easily working with other packages, such as it;s different smoothing methods and not being able to import vertex normals from other apps, forcing you to redo smoothing, no FBX import ATM, none of the exporters preserve instances, also baking doesn't support cages.
if they would address these things in blender and get bmesh in and stable, i think it will be fit into other peoples work-flows a lot better, since even know i dont like it;s modeling tools, it does a lot of things really well, i really love the animations tools of blender more so than maya, and the simulation tools.
True, I agree with those things especially the lack of a cage. Blender should be capable though, it's the import/export scripts that has to be up to date and working.
In max I want the split-edges smoothing from maya and blender, but they keep insisting on using the smoothingroups, in blender I want the cutting tools and uv-preserving functionality from max which I will be in b-mesh if we get it anytime soon, and in max I want the unified UV and mesh editor from blender that uses the exact same tools and functionality in both mesh and uv editing mode, they really shouldn't be seperated.
It's there and open for people to fix, and when looking at unity they have seamless importation functionality of blend files, where everything gets imported intact, and then any issues on the importation side of maya and max has to be taken into consideration as moving stuff between any program has always been problematic.
Blender will always use edge-split modifier for as a "smoothingroup" equalent, it's similar to what maya uses, and none of them will move over to smoothingroups, but if fbx which is a max format saves smoothinggroup information, then the python export script should in theory be able to save edge-split down as that.
Thats all well and good, except for the fact that I never suggested he give up blender entirely, i've been saying this entire time he should learn another app in addition to blender. You, on the other hand, DID suggest he shouldn't bother learning another app. He already knows blender, do you think i'm expecting him to un-learn it or something?
This is simply BAD ADVICE in an ultra competitive job market, especially when the majority of artists you'll be competing with know max, maya, or often both. There are very few "entry level" positions out there these days where teams are willing to take on raw tallent.
Being the "blender guy" is bad for a variety of reasons already mentioned. It makes you less "hirable", not un-hirable, but less. I'm not even saying this is how it should be, as blender guys do probably have an unfairly rough time, but from being involved in the hiring process at multiple studios I can tell you this is simply the reality of the situation. Again, your application could get rejected by some low level HR drone simply because you don't have Max or Maya exp, why risk that?
Even learning one of the big two to a rudimentary level just so he can put "basic maya knowledge" on his resume would be a big plus.
Blender doesn't have professional baking tools, no cage, screwed up hard edge/SG workflow(I read that you need to manually break verts to get hard edges?) completely useless for baking. It can't export custom normals. Where are the 3ps, Xoliul, Bryce etc quality shaders for blender? No way to get synced normals into or out of blender. Blender's baking workflow is not usable for production. Baking is the backbone of modern asset production.
In addition to that, you'll never be able to use blender to bake onsite at a Max or Maya studio, if the workflow is quality and locked down that is, you'll need to know how to use the correct tools to bake to get synced normals into game. That means Max, Maya or even Xnormal, 99.9999% that does not mean blender. If all he does is learn how to import, properly set up smoothing and bake in Max and/or Maya that would be a HUGE plus.
Zbrush can "bake" stuff too, its also unusable for production for a variety of reasons.
Sure it has an API, as all 3d apps do, is it any good? Have any major studios put it through the gauntlet in a production environment? I'm not being snarky here; i honestly do not know.
Your argument is that it has the "potential" to be usable, there is a pretty big difference between that and actually being usable.
All you end up doing in any modelling program is move verts.
They all do the same shit if youre a modeller. I challenge you to find me one model that can NOT be made by another program.
Learn them both. Learn them all. Get good at learning shit because you'll be required to learn new stuff your entire career.
When I started back in 2000, we used max 2.5. I learned Alias Wavefront Power Animator 9.0 (which later became Maya) and it had very little 'polygon modelling' tools or support.
People who used the first few itirations of Maya were NURBS modellers. There was talk of voxel modelling programs and 'blob modelling' programs. The future looked like Sub-D modelling techniques (again it does 12 years later)
Back then there was a debate between photoshop vs painter.
It wouldn't surprise me if Mari replaces Photoshop in 5 years, and I'd be forced to use Mudbox instead of Zbrush.
In the end the program really shouldn't matter. They're theoretically all the same, and shouldn't take you longer than a week or 2 to learn.
Its almost like saying you can't draw with HB pencils, and you need H2 or a mechanical pencil to draw. You should be able to draw with any pencil. It might take a little bit of getting used to, but its certainly doable.
It's enough with a month of dabbling in max and maya to learn the basics enough so that you can work with it without having to go deep into any of the modelling tools, this is what is important when working with the tools of a company, you need to know your way around the program well enough, but you don't have to make your actual mesh in it.
It has professional baking tools, the process is nearly the same as in max and just as crude, it was the cage that was missing, did maya have a cage yet?. hard edge workflow is exactly the same as in maya:
It's a modifier, you mark edges as sharp and the modifier keeps them apart without actually giving you two vertices when you move it around. A split edge is actually seperate vertices and two disconnected triangles.
The normal data in blender is exposed, it's the exportation scripts that has to do the work, and then the target program that has to import them.
These third party shaders you speak of were externally written, blender, max and maya were around even before that, autodesk did not make them, and they could just as easily be written for blender.
But sure, if you want to attribute the lack of these in blender as one of blenders failings, go ahead.
Even though I did bake a few things in blender during a max and maya dominant industry job studio working on console games, I did prefer max, DESPITE the mismatching tangents resulting in less then stellar bakes if you didn't put down work into splitting up the mesh to ensure a nicely smoothed mesh.
And as mentioned, the thing missing is the cage, other things are there, you'll get proper bakes out of it if you wish, I just wasn't too fond of fixing up the smoothinggroups twice due to formats not working together between the programs.
So modelled highpoly and lowpoly, uv'd and set up in blender, then moved over to max via .obj exportation, and then using a third party written obj importer in max because the one that shipped with max by default was prone to running out of memory.
Baking In any application is essentially the same, I learned it in max in a day and then used that same knowledge when doing it in blender, but for already mentioned reasons ended up doing it in max.
Everything is exposed, python is a widely used programming language, all plugins in blender are written in it, all exporters, all importers, did you have anything specific in mind when you asked "is it any good?" or are you just expecting it to not be up there with "the big boys" ?
No, my argument is that I've actually used it in a real scenario, I actually have real experience surrounding that "potential" event you speak of, I know the downsides and upsides, and from my experience as using blender primarily is that as long as you learn the basics of maya and max outside of that you'll be fine.
The main thing is being able to fit into the company's pipeline, and basically there are no/extremely few companies that have a Blender-based pipeline. Most are Max or Maya, so knowing one of those programs you've got about a 50% chance of being able to fit right in without needing to adjust to their main program. If all you know is Blender, there is about a 0% chance of fitting in without adjustment.
Our company is Maya based and I can use it for all that is required, but use Blender for most of my modeling and UV mapping as I prefer modeling in it and find the UV mapping vastly superior to Maya's. As pointed out in one or two of the posts I read there are indeed limitations (eg. lack of preserving vertex normals) so sometimes it's easier to just do some of the work in Maya instead of faffing around jumping back and forth. Also keeping organised work files helps make sure if someone else needs to get at your work it's not a pain in the ass for them.
I'm a big Blender fan and really looking forward to Bmesh integration so the modeling tools can pick up, but my job would be so much harder if I didn't know much Maya. And the reality of the job (character artist) is that things can be akward enough having to bugfix and reexport things to the game, do fixes to geometry, skinning etc. that trying to do that outside of Maya would just make things more difficult.
Short version:
1. Be able to do the job with what the company uses
2. Then incorporate Blender (or Modo...whatever) to improve your personal workflow (if the company lets you use other software).
Learning Max and/or Maya helps with step 1.
I think this should be emphasized. Most larger corporations in any field filter applications heavily to get a manageable amount of resumes to the departments hiring head. This is why it is advisable to look at the employer's job listing and make sure to "integrate" it into your resume. If you get filtered out by HR your portfolio won't even reach the department.
Well a "cage" can mean a few different things and is a fairly vague term, it doesn't necessarily need to have a "cage", however it does NEED to have the ability to use an averaged projection mesh when you have hard edges on your lowpoly. From my understanding it does not have this functionality, only the basic ray cast straight from your lowpoly mesh normals, which will cause gaps in your bakes if you have any hard edges(most meshes) and is not suitable for production. Like using the default XN ray distance settings as apposed to a proper, averaged cage; not a good idea.
Maya has the ability to use averaged normals, or your lowpoly's mesh normals for projection direction. For all intents and purposes this is the same as having a "cage" in max. The biggest difference is that manually editing the "cage"(envelope) in Maya only adjusts ray distance, not ray direction. In Maya its a simple toggle, its actually extremely convenient.
Its really not important which part of the workflow is broken, you can't get custom mesh normals out of blender and that is a big deal, requiring weird work-arounds like manually detaching faces. Again, not a solution that I would want to use in production.
Lack of comparable tools is certainly a downside to using blender, whether its built in or 3rd party does it really matter? Worrying about where the tools come from or how much time people have had to make them is just sentimentalism. You've got the tools or you don't, end of story. Fews studios have a tech/tools team and the time to rewrite various tools and features that are available for max/maya.
Again, "could just as easily be written for blender" we're back to going on potential alone.
Yeah max's crappy tangent basis is awful to work with, but there are third party solutions(and autodesk working on a fix - even if its still a bit messed up). Is there any way to get the same quality bakes as default Maya or Max with qualified normals/3ps QM? Any way to get that out of blender? I'm fairly sure the answer is no to both.
I think you understand a lot less technically about baking than you think you do. You might want to do a little research on the topic.
But what if I wanted to adjust the cage to prevent waves, would it be fair for me to call mayas baking tools unprofessional when it lacks an adjustable cage?
Lack of comparable tools is a downside of blender, especially when most of max nowdays is built up from things the community wrote, and things like popular plugin roadkill that had its origin in lscm, which was something that came from blender, and how long did it take for max to get a proper node-based material editor?
It's a bit of a catch 22, people want a ton of tools for blender that won't exist until they write them themselves, much like how it has gone for maya and max users.
No I guess you're right, Blender won't really be a major player until one of the major players makes it a major player, but as you said and as it is: when a studio decides to pick up an application to implement into their pipeline they'll wont pick the one with only a tiny 5% userbase in the game-art community, that's a given, no matter if it would work out fine.
Not to mention, most studios are already in place, pipelines are already built, exporters and tools are already there.
But this should not be a testament to the quality or professionalism of blender, it's just about how common an application is in its use in the industry.
It took quite some time, countless of games have all been baked this way, all in the name of professionalism, it took some time for people to voice up about that.
But no, had blender had the same kind of vocal support when it came to that thing for autodesk, things would probably move along a bit faster though, fare towards blender?, not sure, apparently max was a perfectly fine option for years even when maya produced better bakes, but blender is not, even in the times of constant change?
There has been shoutouts at polycount from the blender development crowd, they want to know the things that can be improved, and a ton of it has improved now up to 2.6, if the thing that you have mentioned is what's missing from blender then it's a short trip to being a good enough application SANS the community and developer-written tools.
No I understand it, it's an average normal over a vertex that has been split into two, and since blender doesnt use a cage(yet) it will just use the two split normals and not the average one over that split edge.
And look at it this way: I'm not saying that every studio should suddenly switch their pipeline to blender, that won't happen, the only studios that will are going to be indie studios with a big area being in unity.
But.
Recognize that a bunch of people will be or are using blender as their main modelling and uv tool, and that it's perfectly viable and not even wrong in any way to have worked with blenders for years before learning a bit of max and maya so that you can pop it in your cv. I use blender all the time because I'll never reach the speed of modelling and unwrapping in max or maya as I do in blender, and I'll never sculpt in anything but mudbox or zbrush because both blender, max and maya are lacking in that area, and I'll keep baking in max because it is where I'm most familiar with baking in.
Telling someone that they should not in any case have blender as their primary application even if they enjoy using it is not only shortsighted, it's hurtful to the area where blender cannot expand on its own merit, the amount of users.
The default answer should really be "Feel free to pick any favorite amongst blender, maya or max, but learn enough max and maya so that you can put it in your cv" not "learn max or maya INSTEAD, they're much better"
I've used both Max and Blender. Strictly from a sense of modeling and UV mapping, Blender is a better tool than Max. The interface is more intuitive and faster than Max. They've managed to simplify complicated things in other tools. It's worth using.
I don't get what the big issue is with SG vs. edge split. In the game engine, don't smoothing groups get split anyway because no vertex can have two normal values? Aren't custom normals defying how game engines render?
@Fuse
"Working in a major app open you to plenty of community resources like tutorials, scripts, plugins etc. More so than a niche application like blender."
It's not so niche as you think. There is a large number of tutorials for Blender. CGcookie has been my primary source for tutorials, but there are others (CGtuts for example). I also tend to gravitate to tutorials that don't cater to specific tools. Blender is certainly not resource limited.
my point and others poeple point with thigns like this, is even know hte game engine cant handle multiple vertex normals per vertex, maya and max do handle that and can export them via fbx.
say i did all my edge hardening in Maya and wanted to bring my mesh into blender for animation, well i would need to redo all that work with the normals because blender wont import or export that data.
like i said on a lot of blender topics here, yes blender is a fine tool to use on it;s own, but it does have some things that make it very difficult to use it along side other 3d packages, such as the vertex normal thing, and not supporting FBX import so having now good way to get animations in from maya, and the lack of ngon support also.
Not that I'm particularly amazing at it, but the moment I started scripting in Blender, I realized exactly how haphazard its entire UI and workflow is. The features are all there, but the means of accessing them is completely random and poorly designed. It's like these menus were built for use during modeling simply so the user would have access to the tools, without much thought put into their fluidity. I've personally made my own context-sensitive menus and tools to speed up modeling, and am sad to say it's probably preferable that you learn to script python and access the API if you're going to use Blender seriously, due to its lack of widespread use.
2.5's new UI system makes it possible to do a complete rewrite of the application interface for essentially any task, and I should think even moreso now that you can script new GL elements directly in the viewport. People have used this to create prototypes for radial menus, etc. See this random example of edge filleting:
My argument for Blender remains 'potential' based. No, it isn't exactly feasible for production use as it ships. Yes, it could be scripted and rewritten to fit a project perfectly if you've the resources and interest in doing such a thing. I personally think it would be nice to see that done more, since I can't really think of many examples other than the open movie projects which have done this (coding a unique branch for the project, the modular features of which are often eventually integrated into trunk).
The argument is not about how good blender is but about what will get you a ticket in to the ndustry aned listing max and maya/softmage even lightwave will carry a lot more weight.
TBH there is stil that element of blender being regarded as the happy amateur software
listing blender on the cv is like trying to get in to a good nightclub wearing trainers:)
Yeah, this is essentially how I use Modo in my workflow. High/low/uvs. Smoothing and bakes are done in Maya, Max or XN(generally client preference). I'm not even a big max or maya supporter, I just know how important having a basic knowledge of them when working in the industry. I would advice anyone thinking about getting a job to learn that before hand so your potential employer doesn't have to worry about training you up.
Sorry if I have bit a bit overbearing in my criticism with blender, I probably got a little carried away. However, I never suggested that someone should learn Max/Maya over blender, it was my intent to state how important it is to not be viewed soley as a guy who knows how to use blender and nothing else, as I think even you would agree that carries a certain stigma in the games industry.
Yeah, I agree.
PS: RE: "maya being unprofessional etc" No, I wouldn't call maya an unprofessional app for not allowing you to tweak ray direction in the cage, I would call the user trying to do that unprofessional. =P
But that's me, I'm pretty anal about that stuff, you can check out the big thread I wrote on the issues in tech talk to see how I feel about all that.
IMO Maya has the best baking tools out of the box of any app(aside from the totally broken AO stuff - in the versions I've used). There really isn't another app that bakes and displays normals as well as maya out of the box. It offers a couple other improvements over max too, like the ability to select multiple lowpoly meshes and give them unique "cage distances"(envelope % in maya). Maya has its own terrible issues too, but just baking a normal map and displaying it in the viewport, Maya gives the best results out of any app i've seen(without extra plugins).
PS PS: It certainly is a huge catch 22, Blender needs a large user base for people to have the desire to make tools to make it competitive, but Blender needs to have competitive tools to get the user base. IMO this more than any other factor contributes to the low adaption rate blender has.
Yeah, this is completely right. Blender is not used because it's not used. If it were, there'd be more tools for it, and with more relevant tools it'd be stronger in a professional environment. It can't be that hard to write an fbx import script; if no one has it's simply because Blender has not seen extensive use in an environment requiring that... or legal reasons I am unaware of.
The whole vertex normals and baking thing really irritates me too, but I cannot help but wonder if there is not a creative solution out there involving python; just because it doesn't exist, doesn't mean it isn't possible (of course you know this). It does however mean Blender is, at present, much more volatile in professional use... and therefore not really a convenient option for projects with tight time and budget constraints. In that sense, it's up to the hobbyists to build it up to the point of production viability. That is, passionate professionals with a "hobbyist" level interest in developing Blender into something that has a chance, driven by the desire for a clean alternative to apps which are commercially tried and true, if not without flaws.
Well there are multiple parts to this:
A. Custom mesh normals as in anything but a completely averaged mesh. Simply having edges set as hard qualifies as "custom" so if you're exporting straight from blender and into a game engine that is an issue.
B. Exporting not only normals but also tangents and bi-normals is a very important step to getting synced up results, or best results if a proper sync isnt possible, in your game engine.
C. Mesh normals can be more complicated than simply soft or hard. You could have normals that appear soft even along a broken geometry border, custom mesh normals let you do this sort of stuff. Most game engines can load custom mesh normals.
In addition to that there is the discussed issue of averaged cage normals vs explicit mesh normals used for ray direction when baking, which is a big deal.
Honestly whether blender can import or export mesh normals to an app like Max isn't a very big deal. I use a script for both Max and Maya to set hard edges based on uv islands, just 1 click. Its more getting that information out to a game engine.
@Vrav
I agree. If you have knowledge of that in addition to Max or Maya, you shouldn't be penalized for putting it on your resume/cv. In fact, the increased appearance of Blender showing up in resumes next to Max/Maya might cause employers to take pause and consider looking into turning Blender into a tool they can exploit for their studio if many people list it as one of their tools.
It does, I catched up quite a bunch after getting my first job, and learned how to use the modeling tools and such, even if I'm not a big fan at all in them, since there were always times when needed to know my way around when I needed mess around in someones source-files, but in short a vertex is a vertex, both in max and blender.
But when it came to really raw-pushing a new mesh or highpoly I went straight to blender, moving stuff between these programs still is a big issue with many platforms, and were quite a big issues with maya and max before autodesk gobbled maya up. I guess the lack of multiple-program users within the blender community has halted the progress on proper scripts.
Nah, I can see where you are coming from, and hopefully you've seen where I come from in this, and I strongly agree with that. I don't think anyone should be the one trick guy when it comes to apps, since as JacqueChoi said; things are always changing and you always have to learn new stuff.
It was mainly that someone would be suggested to avoid blender that pained me the most, as it is one of those generalizing believes that has been really hard to get rid of and prevents the blender community to growing into something more serious, especially in the very lonely game-art one.
There has even been times where I'm in a good mood, doing happy art, moving between programs like max, photoshop and max, but then suddenly when I have the blender window open I might've gotten an alien comment that made me feel like I was using poser.
Truly, it's a shame it took so long for autodesk to clean things up for max in this area, I struggled for ages hunting down badly shaded models and cutting up uv's and smoothing-groups due to how things just wouldn't bake easily.
It does, it's still hard though as it doesn't just have to excel in some area, it has to excel in all areas to even start pulling someone away from his main application, the only hope to get someone on-board is new people, and as long as they go to proper places to learn game-art they can become a good artist with every tool.
In reality I don't think blender will pull any users from people already cemented in autodesk tools, but I hope for a world where something stupid as other artists not taking blender serious doesn't exist.
It's really nothing I enjoy seeing at polycount, especially since polycount is quite big on "it's all about your portfolio"
I'm not sure anyone is suggesting that blender in addition to another industry standard app is going to look bad. The main thing to avoid is being the "blender guy" who only knows how to use blender.
I have personally reviewed more portfolios than I would like to remember, but I can't recall seeing a single really good artist's portfolio, and then reading that he used blender exclusively on his resume. Most blender artist's have below average work from my experience, so while it may be a bit unfair to assume all blender users are amateurs, it isn't a stereotype that is unfounded.
Lightwave, Softimage, Cinema4d and variety of other lessers apps that started around the same time as Max/Maya did have had plenty of money too, and they couldn't buy market share. At some point the tools that are industry standard are so for very valid reasons, not because of money-stuffing conspiracies.
Maya was industry standard long before Autodesk ate them up. Mudbox had a huge amount of traction before Autodesk ate them up, and mudbox hasn't exactly taken the world over from zbrush since selling out.
To a large extent blender being free hurts them. No, not for the reasons you think, but because they are a non-profit and can offer no professional support to their license holders. The big boys can, and most studios aren't likely to take the risk on completely unsupported software with multimillion dollar projects. Not that Autodesk support is really that great, but this is an important factor to the studio heads writing the checks.