How much is Mari used?
I am starting to feel like I'm being taught and learning to use a texturing program that is not used much at all. How much is The Foundry Mari used out there? Is texturing with Mari a lot different than texturing in Photoshop? Because with Mari, I don't have to think about seams much, a mindset that could be a problem when I have to switch to texture with Photoshop.
Replies
Yeah because traditional art died out as well when digital art emerged. :P
There still pretty much situational tools, not gonna replace Photoshop or Max/Maya
I don't think it's a weak comparison at all. Art is Art, no matter how it's made. Would it be better if I compared a paintbrush and paper to zbrush?
Maybe it's not has popular with younger people, but traditional art is still around, go into any college and I can guarantee that there'll be more traditional art courses than there are art related CG courses. Been into any art galleries lately, yup nearly all traditional art.
I didn't take it out of context, you said anyone that resists these new tools will be extinct. So someone that sticks with Photoshop and Max in five years will be extinct because they didn't learn some specific situational program that does one thing good? Max and Photoshop cater to a much broader feature set than than these new tools.
I know it's good to learn new software and look forward in to future tech, but these programs will still be in a niece are of CG.
Learning ZBrush is all fine and dandy, but again if your traditional art knowledge fails, learning ZBrush doesn't teach you anatomy and form, but traditional sculpting does.
I can bet any employer would choose someone with awesome traditional art skills and medicorce program knowledge, than someone with awesome knowledge of Zbrush, but poor sculpting skills.
Traditional art will always be priority over program knowledge, anyone computer literate can learn max or maya and get the basics in a day, try learning something multi-tiered like art is gonna be a lot longer and more beneficial at the end of the day.
This obviously holds true to varying amounts but is a story we hear often.
http://vimeo.com/65408722
ROD just released their mari pipeline
http://blog.selfshadow.com/publications/s2013-shading-course/rad/s2013_pbs_rad_slides.pdf
The shader support that MeshModeler linked is pretty cool. What separates Mari from its competitors though?
I actually believe that it was created at WETA
Well, that answers my question
There's also this, for anyone else wondering...
http://vimeo.com/61793424#at=150
Desperad0, to answer your question, yes, texturing in Photoshop alone is very different from something like Mari. However, modern workflows often use either both, or would stick to a 3D paint program like Mari. The advantage to using both is that you can paint directly onto your model, where seams become less of a concern, but you still have access to all of Photoshop's post-processing features. What's important is that you have a solid understanding of UV mapping and how it impacts your texture, regardless of the program.
I haven't used Mari, but I have used Bodypaint, 3DCoat, Mudbox and ZBrush. From what I see, your skills with Mari should carry over to those applications very easily.
That being said, I'm sure many people can find use for it working with game assets. It's after all a pixel based painting program with deep PSD integration(and adjustment layer if memory serves me well), procedural layers such as fractal noises and more. I really think it's worth for any texture artist to at least give the free trial a drive.
While I agree anyone learning about the industry should be learning about Mari right now, I disagree with the notion that you can learn Mari and skip Photoshop, which the OP is suggesting. We still use Photoshop at RAD - but we also use Zbrush polypaints and Mari - they're all useful tools for manipulating color data and generating color maps.
I think Mari's acceptance in the industry is about where Zbrush's was 5-6 years ago.
That's why I ask how much it is used because it seems to me only big studios can afford Mari. If you're freelance or indie or small... will you still consider Mari?
Also, substance designer. Everything I have seen looks great for vfx and environmental artist. But other than tweaking settings on skin tones, I cant see it being much more useful than photoshop adjustment layers for character work.
He was auguring that anyone that didn't learn the listed software would be extinct which is why I compared traditional to digital art saying that traditional was still around despite digital art emerging.
I already said above that learning new software was good, but saying that someone creating art is extinct because the're not using the new tool becomes available, is wrong.
It's about learning the fundamentals of art, once you get that down using a paintbrush, charcoal or digital shouldn't really matter. Trying to learn art in combination with learning software is just making things more complex for you.
Absolute horse shit...
this is probably the most irrelevant argument there is! art is art, there is no classic vs digital. it's all art.
nowadays, max/maya are more situational than the other tools available. the only place max has in my workflow now is retopo and uvmapping, and both of those things i could do elsewhere. it's become possible to do 99% of your work in zbrush, if you can get good results out of their texturing tools. if you can't, then use mari. the actual NEED for programs like max/maya/photoshop is becoming smaller and smaller with every new iteration of these other programs.
to say that software like marvelous designer or substance designer are "situational" is just downright ignorant. marvelous designer is a powerful cloth simulator... it's not just a character artist tool, it can and has been used by environment artists too... and it speeds up workflows immeasurably. substance designer and Ddo speed up texturing workflows to no end... and the day those softwares become integrated into software like Mari (you know it'll happen sooner or later), you can practically say goodbye to photoshop for texturing all together at that point.
now, i realise that CURRENTLY studios who focus on mobile gaming probably don't have much need for those programs right now, but the mobile platform is becoming more and more powerful tech wise, it won't be too long before we start to see "AAA like" games on mobile, where these softwares will cut dev time by a lot.
every time i see someone say "oh no, i can live with my classic workflow and art skills", all i can think is "absolutely, but for how long?"
You still can't animate, composite, track and render in these programs. Like I said Max/Maya have a much broader toolset.
Mari and SD are focused on a set area of the pipeline.
If Max and Maya are more situational than Mari and SD, then try getting a fully animated character or environment in a game engine just using Mari and SD.
STUDIOS WONT BE ABLE TO SURVIVE WITHOUT ADOPTING TOOLS LIKE THESE.
not all studios but alot for sure, traditional methods can achieve as good or better results but time costs too much and projects are scaling exponentially
its smart to use tools like this but they are still at the point where you can get ahead by having knowledge you dont need it.....
yet.
ark - you jumped on his statement way too abruptly "Anyone who resists new tools like Mari, Substance Designer, Marvelous Designer, etc is going to quickly find themselves extinct" NOT "He was auguring that anyone that didn't learn the listed software would be extinct" important word being LIKE
also they are situational now... but i think they wont be pretty soon... material definition is so important and things are becoming industry standard very quickly that were not dreamed of at the beginning of at the beginning of the current gen
also who do you know that still models by typing in co-ordinates?
Maybe I did jump too abruptly, I didn't mean to be pointing a finger, but if that's what it came off has then I apologise, but using words like "Extinct" was maybe the wrong choice of word on his part?
Just because some new tool is released doesn't make it's predecessor extinct. Sure it's viable and good to learn the new software, but all pipelines are based around a strong all-rounder CG package.
People actually modelled typing in co-ordinates? That sounds awesome, but Michelangelo painted with a paintbrush and the same brushes are used today.
What I'm trying to say is we keep hold of ancient tools that will forever give certain properties that can't be superseded. Our current set of industry packages can and will be superseded. Don't bury your head in the sand ;-)
Max and Maya or some other core 3D package will always be the hub.
yup, our prop and character texture artists use it regularly, for masks, clr maps, and flow maps. and all sorts of other maps.
If you can show me tools that let me make a AAA fancy pants character in the same timeframe as lowpoly stuff - about 2-3 days, then I'll hop on it. It's the weeks working on one asset that turns me off of AAA the most.
there's nothing that good, YET.
but that's what i was saying, how long will it be before the tools allow that kind of turn around? especially now that you can take a 3d scan, cleanup, zremesher, cleanup, uvmap, done.
the two cleanup passes don't even take all that long in most cases, as 3d scanning tech gets better.
the point i was trying to make, is that all of these tech advances either speed up, or remove portions of time that artists spend doing tasks that are detrimental to their workflow. not embracing them is like saying you love doing everything the long way.
also this.
you never know when the nest you're living in will fall out from underneath you and you're forced to find alternate working opportunities. and you will lose out on a job over someone else with equal artistic merit if they're up to date on tech.
artists generally love their craft... that means they enjoy the process, so will always mourn the loss of need for techniques painstakingly learnt and mastered. You will cling onto them as long as they can for good reason, if you spend alot of effort into something you want to get the most out of it.
but its as much about what makes your work pleasurable, most artists enjoy the fact that things are painstakingly hand painted etc etc, as it makes them more of an artist and less of a technician.
but all that ignores what is/will be generally commercially viable. videogames are heading for a crisis IMO where the escalating production costs outweigh the difference they make to the consumer and any tool like these becomes extremely important to bridging that gap
We're getting there but not quite 2-3 day turn arounds yet. But I do think we're getting pretty close to that point. The latest Zbrush Z-remesher lets you create your lowpoly in a matter of minutes after you have your high rez done. It's amazing!
For texturing in next gen I think it's going to be a lot of material library assignments in the engine and then painting the wear and tear on it quickly with masks and such. We'll see it move towards almost being like prop fabrication on movie sets where you figure out creative ways to ding and dent your assets to give them a story.
The more mari gets integrated with games I think it will really start to take off. For any kind of higher detailed stuff with baked normal maps and spec it works great. But so far for low poly diffuse only I still use 3D coat.
The sins art dump link in my sig has some examples of Mari for games. That being said, the game models had to be used to make the cinematic. Which was the reason why I jumped on mari. If it was more of a stylized lowpoly project probably wouldn't have used it. Mari makes creating crazy detail very fast and efficient. Which makes perfect sense considering what it was developed for.
On a side note, I loved painting the sky-boxes for sins in mari. absolutely a blast. I never had to worry about seams and I could just paint on a decent rez sphere and see all of the focal points in context.
Could you let me know what GPU / Driver version you have? Thanks.
We're actively looking for feedback from the games community about how we can improve Mari so all of this is incredibly useful.
what is more important is a powerful GPU for Mari. what GPU do they have ?
from my experience it wasnt as slow as some are saying.
also, 8GB ram is actually quite low considering windows7 on idle eats away 3-4gb by itself leaving you only ~4gb for other apps and rest is virtual memory and i doubt your schools has even set up proper page file disks for virtual memory.
It's essential to have the latest stable drivers for your GPU installed. We tend to find (and get fixed) a large number of driver bugs.
8GB (at least under Linux) should be fine. Large parts of Avatar were done on 8GB-12GB machines.
If I'm not mistaken, Mari needs a good chunk of GPU memory when dealing with large/many textures.
I've only used Mari a little bit but found the UI to be terribly unintuitive and cluttered. Had problems doing even basic things like flood fill or importing PSDs, and there's no decent training materials I've run across focused on game artists either for the current version.
I can't seem to access info for GPU on these computers.
It doesn't really support non-square textures to my knowledge, which is a pity, but you still can get around that by using squares - higher resolution and then squashing down and doing some sharpening.
It doesn't really makes sense for lowpoly art, because you have to have expensive GPU too run Mari well. Few of our artist have 32gb ram/i7/SSD rigs, because they use PCs for Nuke as well. So there is really not a problem, but otherwise investing in machines like that only for low-poly it doesn't balance out.
Our artist love it, me as tech artist haven't got chance to go into scripting and shader-writing, because there was no need too, but i would strongly recommend looking into Mari for VFX/Movie/Beauty production.
Mari uses a virtual texturing system (like ID tech 5) and has a fixed Memory footprint for normal textures. 1gb of Ram should be able to cope with many 10s of Gbs of textures.
Mari does use extra GPU ram for live tiling and projection textures, you can however cache these to manage memory effectively.
I have to agree with this, 30 days seems pretty standard for most software but I think even that is a little on the short side to become familiar with a new app, and for it to prove its worth in a production context.
I'll certainly let people know about your request for a longer standard eval.
If you drop me an email at greasley@thefoundry.co.uk I can sort you out with a longer eval.
that's probably one of the biggest reasons for uptake of Autodesk products vs anything else at this point. it's like smoking... we all know it's bad for you, but they get to you early enough that you struggle to break the habit!
I'd like to know what might be making this run slowly.
Thanks
How can I do that? The archive is 600+MB, and even DropBox doesn't accept upload that big.
I think it's just our classroom computers' built, because everyone in the class who textures with Mari came to understand that we'd better save-a-lot, and not have other programs running at the same time. Only mp3 or movie.
I can setup an FTP account at the foundry for you. That will take 600+MB without any problems.
Also a log file from a sessions will help us work out what GPUs and driver versions you're using.
The log can be found in C:/Users/<your account>/Mari/Logs/MariLog.txt
The log would probably be more use than the archive at the moment and you should be able to attach that to an email.
If you can send the log to greasley@thefoundry.co.uk that would be great.
Thanks