Transitioning studio from 3DS Max to Maya

maya
max

#1

My studio is in the process of considering a switch from 3DS Max to Maya for our rigging and animation departments across various project teams. I know that many of you have made similar transitions in the past and probably others have done the math and decided against such a transition.

For us, the reasons for the switch center around Maya’s apparent ascent to become the game industry standard for rigging and animation applications (If any one has real data on this, please share it). This presents the following disadvantages for us as a 3DS Max house.

  • It’s difficult for us to find experienced (and even relatively inexperienced) candidates that have a good level of familiarity with 3DS Max (Part of this I think is due to the geography as we’re located in relatively small and somewhat isolated pocket of the industry.).

  • As a result of the above point, we often end up hiring candidates that have strong Maya backgrounds, but little experience with 3DS Max. This of course increases our onboarding costs significantly.

  • Similarly, many of the outsourcers that we work with are not well equipped to efficiently interact with a studio that principally relies on 3DS Max for its rigging and animation pipeline.

  • Because 3DS Max’s python implementation lags behind Maya’s, our existing tools and current tools development is based on MAXScript and DotNet. Neither of which is terrible, but a move to a Python development would offer some compelling advantage to us.

  • We miss out on a lot of really cool, non-Autodesk tools that are either Maya exclusive or just better supported on Maya (Red9’s tools, Epic’s A.R.T tool set and Fabric Engine come to mind).

  • There’s also a perception at the studio that the skills and knowledge sets of our TAs and animators are falling behind industry trends, leaving our team members out of sync with the current and future needs of the broader industry. Admittedly this last point is a bit self-serving. However ultimately, I do believe that the transition to Maya represents a “rising tide lifts all boats” scenario in that these sorts investments in personnel generally provide a favorable a ROI for companies.

Regarding cons, the switch would obviously be very expensive short-term, both in terms of direct costs and in terms of indirect cost such as reduced productivity during the transition (and the time afterwards while the new tool set and pipeline mature). While there’s a lot that needs to improve with our current 3DS Max-based pipeline, it is very functional and has the benefit of (already being built and) having been thoroughly production tested.

If the Studio decides to commit to a Maya transition, the switch would likely be executed in one of three basic ways (listed in the order of increasing anticipated cost):

  • By Project - some games are fully transitioned to Maya and other projects remain exclusively 3DS Max-based (my personal favorite)

  • By Project Hybrid - on a per project basis, projects are either fully transitioned to Maya, remain exclusively 3DS Max-based, or are partially transition to Maya to one extent or another

  • Studio Wide - all projects are fully transition to Maya

So with all of that said, I’d very much appreciate any thoughts that you gals and guys have on the notion of such a switch. Specifically, I’d love to know whether you believe Maya to be worth the transition cost, and if so, does it make sense to pay additional costs to port old rigs and animations (or whole projects for that matter) to the new pipeline? Advice or input of any kind would be greatly appreciated.


#2

I was working in a game dev studio in the UK when the Senior Tech Artist of the studio decided to switch to Maya (many years ago).
The experience was horrible, the result was horrible as well.
Not only the transition to a new tool went bad, we lost many good artist that left the studio because of that decision.

During the transition we barely produce a 10% of what we used to.
Then many of the Maya tools weren’t supported by the engine, so more work for programmers.
Some Maya tools that used to look amazing when we started using we realize they were not as good as we thought initially (even some were total crap).
Bake is still awful in Maya, modeling is very inconsistent and slow, it used Y-Up (which for us broke many rigs in the engine), it handles waaay less polygons than Max (right now viewport performance and modify mesh in Max must be about 10 times faster than in Maya), vertex color in Maya is very basic (I could do a lot more in Max), the history is a problem because it’s a very destructive way to work (stack in Max is its biggest advantage here) and you have to delete history if you don’t want to crash Maya very often, it has hundreds of bugs…

When the transition was over, my estimate is that we were about 30% slower (or less efficient) with Maya than with Max, and that situation I’ve live it in another studio in the US some years later. Again my estimate was that we were about 30% slower. I blame it on the Maya workflows, is not made for speed (the equivalent of one action in Max is several actions in Maya).

I usually compare it to Apple. Everyone think it’s great, there is this perception that they’re the best. But the reality is that it’s a minority and it has a lot of problems.

Since a few years ago I avoid every studio that work with Maya. I don’t think it’ll be a problem in the long term since it looks like lots of Maya users are transitioning to Houdini, in a few years we may see Maya down the same path of Softimage and Mudbox.


#3

I’d approach that kind of decision VERY carefully. While Maya definitely has its merits and is a welcome addition to any pipeline, Max is still a pretty big powerhouse in gaming department. Some inside numbers show that it is, by FAR, the most used DCC to export stuff into Unreal. Going point to point:

“It’s difficult for us to find experienced (and even relatively inexperienced) candidates that have a good level of familiarity with 3DS Max (Part of this I think is due to the geography as we’re located in relatively small and somewhat isolated pocket of the industry.).” - True, and the best ones are already employed in large studios and pipelines. Just recently a new founded studio had trouble finding Max artists in NY area. The competition to grab Max artists is getting bigger and I blame the lack of them in two fold - the education material favors Maya big time and the marketing also tends to lean towards Maya too. It also does not help that the Viz market is getting hotter and hotter (and generally pays more for entry level artists than gaming industry does).

“As a result of the above point, we often end up hiring candidates that have strong Maya backgrounds, but little experience with 3DS Max. This of course increases our onboarding costs significantly.” I’m sure I’m not the only one here that finds that teaching a Maya artist how to use Max is easier than the other way around. And (as Xerges said) a Max artist is prone to spill his talent on other areas, helping productivity and delivering faster when deadlocks are found. My hint? Train them to use Max and see what comes as result.

“Similarly, many of the outsourcers that we work with are not well equipped to efficiently interact with a studio that principally relies on 3DS Max for its rigging and animation pipeline.” We are! Get in touch with us if you need help! We do everything from modeling, to rigging to animation to final frames. We are small studio in Brazil but ready to tackle and help any Max pipeline deliver!

“We miss out on a lot of really cool, non-Autodesk tools that are either Maya exclusive or just better supported on Maya (Red9’s tools, Epic’s A.R.T tool set and Fabric Engine come to mind).” That I can’t help you with, but there are plenty of awesome tools made in max and for max from several guys and companies all around. Regarding Epic’s A.R.T. isn’t this kinda of a “biped” version for Maya? And Fabric Engine it’s already on its final Beta for Max http://fabricengine.com/fabric-3ds-max-final-beta-come-give-spin/

“There’s also a perception at the studio that the skills and knowledge sets of our TAs and animators are falling behind industry trends, leaving our team members out of sync with the current and future needs of the broader industry. Admittedly this last point is a bit self-serving. However ultimately, I do believe that the transition to Maya represents a “rising tide lifts all boats” scenario in that these sorts investments in personnel generally provide a favorable a ROI for companies.” This perception is just wrong and prone to change.

Anyway, make this move after some serious thought! Losing productivity equals dying in a competitive field as games.


#4

Sounds like you had a bad experience that’s coloured your view of Maya. I Personally disagree that studio’s are 30% slower using Maya than Max. Some Artists might be slower in Maya, but that’s not the software’s fault. You could reverse the situation and the same issue would apply.

It sounds like either your artists were taking longer to adapt than people assumed. It might have been an issue with the on boarding or perhaps people were resistant to the change (im sure we’ve all worked with artists that HATE either Max or Maya for some reason and would hate to swap), or the time frame they were expected to get up-to speed was too short. Again personally these are all solvable problems. Its not Software specific. Id worry if after a year any artist was still 30% slower in either package after swapping, they might have other problems that some tools, training or workflow changes could fix.

Yes there are some things that Max does better, but there is also a lot that Maya does better. The last thing we all need to do is turn it into a Max Vs Maya Cage match. The software’s just a tool at the end of the day. Its what produced with it that is important in the end.

I’ve personally been an animator and TD at places who used Max and Maya, and for myself i prefer animating in Maya by a long shot, and doing any pipeline /rigging/setup work i think Maya is much more flexible for a few reasons.

I find Maya is a lot more open ended than Max, there’s hardly anything that’s broken in Maya you cant fix depending on your scripting ability with Python and C++. Scripting in Python (PyMel and Pyside for included) is a lot more flexible than Max-Script, it also has immensely more resource available to help with issues you might run into. There’s libraries available for just about anything you might need. In the long run its also not a dead end when it comes to learning a programming language. Max-Script has one applicable use. Learning Python may have a wider use within your studio. In the long run you aren’t painting yourself into a corner with your tools and pipelines, Python is applicable for Maya, Houdini, Motion Builder etc.

A lot of it is going to come down to how flexible your artists are, how willing they are to embrace change, and how much time your allowing to totally re-tool and re train the studio.

/2cents


#5

That sounds like a pretty terrible experience. We would of course aim for a smoother transition. Many of our modelers already use Maya (we have a Modo user or 2 as well in addition to our 3DS Max modelers). That said, our modelers would hopefully be largely unaffected by the switch to Maya as we’re really only looking at transitioning our animation and rigging teams.

For the animation teams though (and riggers) this transition would represent a massive change in work flow. Even in a best case scenario, the lost productivity early on is going to be pretty significant.

My hope would be that we could develop the Maya pipeline and tool set in relative isolation from our live projects, only bringing Maya into a production capacity when we hit a point of viability (and only then on new projects).


#6

Good points all around, davius. Honestly, I’m not convinced that Epic’s A.R.T. is really that compelling an argument for Maya (haven’t really looked at it closely), and I know that the Fabric guys are working on getting Fabric’s 3DS Max support caught up to their Maya efforts. Also, you’re right 3DS Max has cool exclusive tools of its own as well.


#7

Do you have any resources that you can share on this? I’ve not had a lot of success in locating information on relative market share for 3DS Max vs. Maya in games (or in any other industry for that matter).


#8

I’m certainly with on you the “subject to change” comment. :grinning:


#9

Unfortunately no, but this is not so surprisingly if you actually think about it. But I didn’t hear this from 3rd parties or read on some dubious webpage - it’s reliable but take it as you will.

@sscott [quote=“sscott, post:4, topic:8646”]
The last thing we all need to do is turn it into a Max Vs Maya Cage match.
[/quote]

And then goes on and piss poor Max interoperability? Did you ever know that Max supports Phyton? I give you that Max Python implementation lacks in comparison to Maya’s, but just how long have Maya had that option? Max has it for 3 versions? And I can assure you that Max’s Python implementation is being worked out version after version. The question I like to make is - what did you try to make in Max using Phyton? Or what tool did you try to use in it coming from somewhere else? I can direct you to people involved in Max dev so you can chime in what are the roadblocks you’re facing.


#10

About 8 years ago I was talking with someone from Autodesk that assured me that Max was selling 10 licenses for each license of Maya (he is not working at AD anymore and I’ll not disclose his name, you have to take my word on it, sorry).

here is the annual report of AD:
http://www.annualreports.com/Company/autodesk-inc
(2015) Maya is falling 9% while Max rise 5%

Then a few years ago I found a post of somebody who was using this annual reports to extrapolate the number of licenses of Max vs Maya, and it was about 8 licenses of Max for each license of Maya.

some figures I’ve found
11054 companies using Max against 3113 using Maya
https://discovery.hgdata.com/product/autodesk-3ds-max
https://discovery.hgdata.com/product/autodesk-maya

Ubisoft Montreal for instance, huge game developer, Max based, several thousands of licenses of 3ds Max.

My advice is being away from Maya as much as you can, there is nothing you can do in Maya that you can’t do in Max.
And make your modelers work in Max and Zbrush, you’ll be much more productive.

some more figures:
https://idatalabs.com/tech/products/autodesk-3ds-max
https://idatalabs.com/tech/products/autodesk-maya


#11

That’s some great info., Xerges (see here for annual reports straight from the horse’s mouth). Clearly I didn’t search hard or effectively enough. :grinning:


#12

Last figures I read was 110k active licenses of Max and 14k licenses of Maya,
but I can’t confirm or provide the source
but looking at those reports and companies using software data it sounds quite accurate


#13

I have used Max and Maya as both a tech-artist/tools dev, and as an artist/user. I have been through the transition between both software packages twice, and my current studio is going through another (my third) transition right now.

I can say that I prefer Maya as a tools developer but prefer Max as a modeler. It’s not a huge preference, it’s a mild one.

The biggest change is understanding Maya’s DAG. Reading:

This is a great resource for mapping functionality between workflows between Max and Maya:
http://paulneale.com/tutorials/MaxMaya/maxMaya.htm

The NEX plugin steals the best from Max and puts it in Maya:
http://draster.com/

But the best advice I can give isn’t a plugin or a website, it’s adjusting your expectations. The Maya workflow is fundamentally ‘different’ from 3D Studio Max. At first, you’re going to want to learn to work the max-way in Maya, and in some cases (like with NEX) it’s fine. But you’ll never truly be comfortable until you learn the intended workflow promoted by Maya power users. Go on youtube and look up some videos on how to rig and animate the Maya way.


#14

@PaK pardon my question but in what Maya version are you? NEX was integrated in Maya 2014 EXT1 (or was it EXT2? Can’t remember)! And that page from Paul Neale (a world renowned 3dsMax Artist/TD/Teacher - he taught me some rigging in Max #hint #hint) was meant to show that Max ISN’T second to Maya when it comes to animation/tools (TBH, there are quite a few tools available in Max that are NOT in Maya like Expression Controllers, LimbSolver, Visual Maxscript…). Granted, those are core tools, and that list was made quite some time ago, prior the launch of great new additions like MCG and DCM.

My 2Cents (since everybody is giving) - talk to your bosses/decision makers. Tell the true facts, show the links posted here, show how Max is widely used in ANY industry and make a stand to make the pipe rely on the tool you’re already very much proficienty at. Maya can live along Max (if needed be), and if your challenge is finding good Max artists, that’s because they are (most) already taken. You can outsource (like most studios are doing anyway - did you know that a big chunk of Horizon Zero Dawn was done by my neighbors here?) or you can train your artists to be proficient with Max.

All in all, you’ll be able to deliver faster, stay competitive, and manage your projects with confidence in an environment you are already familiar with. Don’t fall that easily for the mermaid song. :slight_smile:


#15

Epics A.R.T isn’t particularly great and Fabric is supported in Max.

I am at an all Max studio and we are intending to make a move to Maya during the course of the next project. This is being done using Fabric. We are creating our rigging pipeline with Fabric in an agnostic manor and then allowing the animators to choose from Max or Maya. This is the high level plan and I am sure there will be difficulty along the way. However taking this agnostic approach is the lowest risk when compared to a hard switch which could have a massive negative impact.


#16

This is an interesting thread and something that we’re seeing a lot more of recently, studios moving over to Maya (often from XSi) and looking for help in that transition. Doing it alone is a tough call unless you have an experienced Maya TD at the helm and time to develop and stabilize your pipeline. The thing with Maya is that out of the box you’re still left with only 60% of what a production actually needs. There are rigs out there, there’s help out there but there’s very few integrated solutions. Yes our Red9 StudioPack is a nice addition for animators but it’s only an animation aid / api, you still need a rig, exporter setups, maybe project management integration, and so it goes on…

Yes I would say this as we’re in the business of helping studios just like you get up and running with a pipeline out of the box that has been in full production at many AAA studios for many years. This is exactly why we did the ProPack and the Red9 Puppet Rig.

We’ve seen it a number of times with clients who have gone down the road of writing a rigging solution themselves, got 6 months into, the guys who writes it leaves and it becomes redundant because it was never actually finished or production tested. There are plenty of rigging solutions out there for Maya, ours is just one of them of course.

It all comes down to timing, how much down time have you got for this transition and where do you need to be in 6 months time. Maya is a good bet but does need a lot of TD support to really get the most out if it.

thanks

Mark


#17

I wasn’t making a partisan inference with my link, I was trying to provide helpful info for someone who wants to do something in Maya and knows the equivalent Max functionality, a look-up table of sorts.

I think you misunderstood my post davius. I’m not saying I think one package is better than the other.


#18

We are transitioning from Max to Maya/Mobu. I’m a 3dsMax veteran and just couldn’t ignore some facts. Here are some of the reasons we are moving - whether these are valid or not is debatable:

  • Referencing in Maya, although not perfect provides a less waterfall pipeline which is invaluable. 3dsMax’s X-Ref, even after the advancements provided in 2017, is not production ready from our preliminary tests.
  • Maya can run standalone. 3dsMax sort of does, but drags all of the UI with it and is cumbersome (seemingly 2018 should fix this).
  • One game engine in particular provides Maya Tools out of the box.
  • In our experience, the number of animation/tech applicants familiar in Maya is superior to those in 3dsMax (the film industry tends to gravitate to Maya).
  • The majority of our animators preferred and were originally trained with Maya/Mobu.

An interesting exercise is to establish short to long term advantages and costs. Ultimately, I do believe that if the majority of the team prefers a platform, despite it’s failings, you should probably cater to this, if only for retention.

All that being said, we are still debating whether or not to do our animation entirely in Mobu. Our environment team will not budge from 3dsMax since it is incredibly risky and they are familiar with it from a past production.

At the very least, you’ll know that you’re not alone. :slight_smile:


#19

Thanks for hoping into the discussion and sharing your perspective, Mark. Red9 is certainly one of the solutions that we’ll be looking into. We have the luxury of being able to set our own transition timeline to a degree (and of making the more central call of transitioning or not). I’m actually not the principal driver for the proposed transition to Maya, but as a soon to be DCC Tools and Pipeline guy, I’ll be pretty involved in a transition should one occur.


#20

Thanks, CLR. That genuinely is a comforting thing to know. :smile: