I am not really happy with the new features.These should have been in Maya like two years ago considering how much it cost.Thats why I switched to Blender cause its integrated u can model,sculpt,texture paint or make a video game(game engine)in Blender.
On a side note,these features like grease pencil,vertex modeling,fluid retopology have existed in Blender since.But I am guessing a lot of studios will have to stick with Autodesk because they have obviously spent money and time familarising their staff with the apps and built plugins and inhouse tools around it.Switching to another app is gonna cause more money and time.
I switched to Blender cause its integrated u can model,sculpt,texture paint or make a video game(game engine)in Blender.
Has any commercially released game ever used the Blender game engine?
And I don't think Blenders paint feature will ever be taken seriously until it has proper layer support (ie. Photoshop). Though I guess the guy who made B-Mesh is working on something like that, but the last build he did was a few months ago and crashes Blender 100% of the time.
Has any commercially released game ever used the Blender game engine?
And I don't think Blenders paint feature will ever be taken seriously until it has proper layer support (ie. Photoshop). Though I guess the guy who made B-Mesh is working on something like that, but the last build he did was a few months ago and crashes Blender 100% of the time.
Not that I know of.But I have watched the demos made with it and they pretty impressive.Commercial games seem to go for udk most of the time.
Well,I have used the texture paint feature and got good results.There is an addon that comes with Blender called Texture layer manager.It enables u to create layers easily with mix,add e.t.c.
I think it's pretty clear the focus is on Maya now, most of the XSI dev team was switched over to concentrate on Maya as well. Funny they picked the one with the oldest core and legacy code.
Utterly ridiculous what there charging for subs, especially in EU for the little features you get. When Camera Sequencer is classed has a major update and looks like something that could be done in ICE relatively easily, then you see Modo putting out a release that has more features/fixes than Autodesk has done for all 3 of its major 3D apps over the last couple of years, you know innovation went out the window a long time ago at Autodesk.
Wouldn't be too surprised to see Max going the route theyve chosen for XSI and classing it has some support package, since they want everyone on the suites. Develop one package with the other two thrown in as support, charge you close to the cost of all three, yet developing only one!
I think it's pretty clear the focus is on Maya now, most of the XSI dev team was switched over to concentrate on Maya as well. Funny they picked the one with the oldest core and legacy code.
Utterly ridiculous what there charging for subs, especially in EU for the little features you get. When Camera Sequencer is classed has a major update and looks like something that could be done in ICE relatively easily, then you see Modo putting out a release that has more features/fixes than Autodesk has done for all 3 of its major 3D apps over the last couple of years, you know innovation went out the window a long time ago at Autodesk.
Wouldn't be too surprised to see Max going the route theyve chosen for XSI and classing it has some support package, since they want everyone on the suites. Develop one package with the other two thrown in as support, charge you close to the cost of all three, yet developing only one!
exactly, maya is like a dinosaur in between these programs...
yes, but it is also true that modo is still missing a lot of stuff and it is easier to add new high profile features. BUT, the way these features are added puts autodesk to shame.
exactly, maya is like a dinosaur in between these programs...
though, of course lets not forget, Max launched in '96, Maya appeared in '98, Softimage (nee XSI) in 2000, Modo 2004, Blender '95, Houdini '96, and Cinema4D '96 -although it was available on the Amiga before that, same as Lightwave. So you could say, most of the 3D software today has been around abit.
There is an addon that comes with Blender called Texture layer manager.It enables u to create layers easily with mix,add e.t.c.
I've tried it, can't even figure out how to delete layers, and its got problems with seams which Blenders paint normally doesn't. The UI also makes it completely unwieldy for more then maybe 10 layers maximum. It needs a ton of work.
To delete layers,go to the textures section in the properties window(where u set up ur materials,constriants e.t.c) and click the minus sign.The textures layer manager addon works with the textures section.
though, of course lets not forget, Max launched in '96, Maya appeared in '98, Softimage (nee XSI) in 2000, Modo 2004, Blender '95, Houdini '96, and Cinema4D '96 -although it was available on the Amiga before that, same as Lightwave. So you could say, most of the 3D software today has been around abit.
You are right, I was ultimately thinking about previous alias products that originated the ui of maya...
So it's safe to say that they are all dinosaurs except for modo? We don't how much of lightwave was "stolen" into modo too.
I do wish they had implemented more features in Max, but I'd be a fool to neglect everything else they've done, massive viewport updates, Populate(true that it's not applicable to every field of work, but still it's an amazing tool), massive PFlow upgrades(you'd have to pay an extra grand to have the same features in PFlow, now it's native and not a subscribers only feature, think about that), a completely new and awsome menu system, and other smaller but important features like vector maps, etc...
I think right now they are focused more on upgrading the core and behind the scene stuff rather than adding more features, but it's just common sense to me that when they are done doing that they'll start adding all the features that they know is missing in there...
Smaller late changes with larger early prices. Autodesk in a nutshell. I can't believe Autodesk expects anyone to pay so much money for products that have improved so slowly, while competition is exploding so rapidly. More often. And at half the price.
@Macrow, they will do this as long as it works out for them. Their market share is insane. So many clients build custom pipeline around Maya or Max, that it's hard to imagine that they would change their core technology. Buying the cheaper ones extra, is always an option, but the core tool, don't think so. And AD knows that.
Yeah, it's momentum in a lot of cases. If every artist was to start fresh, from square one, and evaluate modeling tools - I don't think nearly as many would choose Max.
In terms of modeling tools, Max, Modo, and XSI are the kings. But the only program with a system comparable to Maxs modifier stack, and the flexibility it brings, is Blender.
XSI's operator stack is similar to Max's modifier in alot of ways and you can easy emulate it and control the order in which tools/actions get evaluated.
XSI's operator stack is similar to Max's modifier in alot of ways and you can easy emulate it and control the order in which tools/actions get evaluated.
I am unfamiliar with Max, is it not like the Maya's input nodes that are shown in the channel box stacked as history?
What do you think the would choose then? I thought Max's modelling tools were excellent?
Is it just Modo that beats it?
IMO, yes, but that's not really what I meant. I don't mean they would all choose Modo but rather that if Max didn't have the momentum and mind share that it does, artists would take a closer look at other apps before settling down.
As it is now, everyone around them uses Max and every job application lists Max as a required skill so ... they choose Max. Not because it's necessarily the best tool for the job.
I sort of have a thirst for learning new things so I try a lot of programs just for the hell of it, which is how I found Modo. You never know what's out there until you look.
I am unfamiliar with Max, is it not like the Maya's input nodes that are shown in the channel box stacked as history?
Modo rules for modeling on my side nowadays...
Maya uses a Dependency Graph were it passes attributes along it nodes to accomplish something, whereas Max and XSI have a stack system were the attribute/nodes can be reordered and selective history kept, wheras in Maya the nodes are dependant on their previous nodes, hence the name Dependency Graph.
I am unfamiliar with Max, is it not like the Maya's input nodes that are shown in the channel box stacked as history?
Modo rules for modeling on my side nowadays...
Maya has a history stack that can't be reordered. You are pretty much forced to work very linearly. Going back down the stack and making changes can be a dicey operation with horrific consequences farther up the stack. I think most people collapse the history and just muscle through with poly wrangling when they could be working a little less linearly and not spend so much time brute forcing their way to the final result.
Max has a modifier list, that can be reordered. The stack in max isn't dependant on previous operations being in a specific order. It can operate on an object level so it can be independent from changes made to the geometry. Meaning you can add a bend, go back down the stack and make changes to the It's modifiers become more like photoshop adjustment layers rather than a recorded event history list. You can work non-linearly with quite a few different modifiers ordering them however you need.
You can leverage the stacks functionality when creating modular assets, like bending walls, pipes or ducts. Go back down the stack and toggle a few modifiers on or off, reorder the stack, insert new modifiers anywhere in the stack and from the same base shape you get the various pieces all without fear, that changes made below will cause catastrophic results above. I can go back down the stack and make changes to the base straight piece, adding or removing geometry as I need, then I can toggle the modifiers and extract the various pieces that I need. At any point I can toggle on or off various modifiers to see how each piece will look.
If a modifier up in the stack freaks out it's because I left a sub-object selection lower in the stack and the modifier is working on that selection instead of the whole object. Which is a simple fix, go back down the stack, exit sub-object mode properly (like I should have) and everything is fine. In Maya, those kinds of operations have me very very nervous and I wouldn't risk it, I would stick to working in a straight line and would probably iterate less and be much less likely to try various things.
That's right, I used to disable history most of the time. Therefore I did not mind modo not having history as much. Although sometimes it is annoying if you create a base shape and to resize it to some precise dimensions, you have to recreate it... I got used to it by now.
Thanks for the thorough explanation of the difference guys.
That is close but as far as I know Maya doesn't handle changes to the geometry? Especially after the fact? Can you go back before the inputs were applied and make drastic changes? I thought that would cause serious problems and the only real way to work around it was to delete and redo the modifiers?
For example, in max I can insert the edit poly modifier before or after the bend modifiers and make drastic changes to the geometry or add additional object, you can't do that in Maya can you? Maybe you can...
For example, in max I can insert the edit poly modifier before or after the bend modifiers and make drastic changes to the geometry or add additional object, you can't do that in Maya can you? Maybe you can...
yes you can in most cases. if you just modify topology to the current mesh and keep the bend/lattice and lot of other deformers without deleting history then you can keep making changes to topology and still have the deformers affect the changed topology.
however, it may not be possible in all cases. it depends on type of merging/editing you are doing.
awesome thing about Maya is it is node based, so as long as you dont merge multiple nodes, you can keep editing that particular node. you can also remove certain poly tweaks from hypergraph.
it may not be possible in all cases, depends on type of merging/editing you are doing.
max has a linear modifier stack which re-evaluates when you change state. ie it is a DAG (Directed Acyclic Graph). the reason why things work out when you move modifies is that it re-evaluates the whole stack from the root upwards in a linear fashion. the problem with this is that is forces a large amount of memory copying. so when your stack starts to grow so does your memory. large stacks on dense geometry becomes unstable.
maya on the other hand uses what is known as a DG (Dependency Graph). maya also has a DAG built on top of the DG. the DG or the DAG can be used to make nodes. nodes pass messages through the DG or the DAG. nodes in the DG have a non-linear evaluation. nodes in the DG update only when they are called to do so by another node that needs some information to change its state. the reason why you do not have a nice little linear modifier stack window in maya is because nodes are not linear and live in a graph. you can view the maya DG in the 'hypergraph'. the DG solves the problem of excessive memory copying. you can have hundreds of thousands of nodes that operate on geometry with out copying that geometry in memory.
it would take a lot of print to explain how to work with the DG to be efective and get what you want. it takes more work to learn then a simple DAG. its a bit non-intuitive from the outside. but it works magic when you understand it.
undo is linear in both maya and max there is no branching in the undo in either package. the differences are related to edits of the DG vs DAG.
max has a linear modifier stack which re-evaluates when you change state. ie it is a DAG (Directed Acyclic Graph). the reason why things work out when you move modifies is that it re-evaluates the whole stack from the root upwards in a linear fashion. the problem with this is that is forces a large amount of memory copying. so when your stack starts to grow so does your memory. large stacks on dense geometry becomes unstable.
maya on the other hand uses what is known as a DG (Dependency Graph). maya also has a DAG built on top of the DG. the DG or the DAG can be used to make nodes. nodes pass messages through the DG or the DAG. nodes in the DG have a non-linear evaluation. nodes in the DG update only when they are called to do so by another node that needs some information to change its state. the reason why you do not have a nice little linear modifier stack window in maya is because nodes are not linear and live in a graph. you can view the maya DG in the 'hypergraph'. the DG solves the problem of excessive memory copying. you can have hundreds of thousands of nodes that operate on geometry with out copying that geometry in memory.
it would take a lot of print to explain how to work with the DG to be effective and get what you want. it takes more work to learn then a simple DAG. its a bit non-intuitive from the outside. but it works magic when you understand it.
undo is linear in both maya and max there is no branching in the undo in either package. the differences are related to edits of the DG vs DAG.
That makes a lot of sense, thanks for explaining it. I guess part of my problem is that in max, its very easy to actually work with and interact with those pieces but in Maya its hidden and tangled. What good is functionality if people can't find it and leverage it?
It also seems like Maya could potentially be a bit more buggy with its non linear path and could potentially be harder to track and reproduce bugs with such a diffused network of nodes, max might be a bit more stable or at least easier to track and reproduce because it's going in a straight line.
It seems like it might take max a little longer to plow through a long stack but maya might actually become more unstable the more nodes you toss in there? Especially if it isn't evaluating them all? It might have dormant nodes that seem to be working until they get passed some info and update? They probably have a way of dealing with that, but I have spent more time crashing in Maya than I ever had in max, tho I chalk that up to my inexperience.
It seems like Maya's problem is UI design related while max is just screwed in the performance department, ha. The Maya team could actually straighten out the UI if they wanted but for whatever reason they aren't?
most maya user arnt using the UI at all..
i use 90% shortcuts and marking menues...
For the most part I use shortcuts in max also, I have a pretty elaborate setup to use and reuse buttons for different tasks just because its faster. But there are some operations that it just makes sense to do through the UI. It seems like digging into the hypergraph would be annoying and awkward, maybe its not, I'll have to dig into and investigate it further.
i'm not suggesting one is better. what i was pointing out was some of the core design differences and why things feel and work different in the two packages. having a grasp of that helps to get a feel for whats going on under the hood.
maya at its core was designed to solve two problems with the DG and its design.
1 scale:
how do you grow very complex scenes with lots of components all working together in a way that is efficient.
2 pipeline:
how do you integrate different components like animation simulation etc into a coherent whole that can be managed in a sane way.
it does those things very well and provides a solid platform.
as for modelling i'm not going to get into that. but you can make it happen if you put in the time and develop your work flow. it will be different then your max work flow but just as fast and efficient. that's my experience anyways.
It also seems like Maya could potentially be a bit more buggy with its non linear path
Sorry, is that just something you made up or how do you justify the relation between being "non linear" and "unstable"?
Just because something is "non linear" it is more likely to be "unstable"?
I would argue the two are not related at all.
If anything, Maya's 'goal' is that one node does not need to know anything about the next node to be able to interact with each other. To me it seems like this would benefit stability.
Having used both, I would say that in Maya's API it is less likely (though certainly not impossible!) to breaks things by misunderstanding the correct usage of data/nodes/features then in the 3dsMax SDK.
Users/artist may end up messing with nodes and connect things that do not make sense and therefor give them results that are hard to understand and hard to visually debug what is going on.
I can totally agree to that. 3dsMax modifier stack is much more obvious and simpler to use for an artist.
That is important, but has nothing to do with program stability.
Sorry, is that just something you made up or how do you justify the relation between being "non linear" and "unstable"?
Just because something is "non linear" it is more likely to be "unstable"?
I would argue the two are not related at all.
If anything, Maya's 'goal' is that one node does not need to know anything about the next node to be able to interact with each other. To me it seems like this would benefit stability.
Having used both, I would say that in Maya's API it is less likely (though certainly not impossible!) to breaks things by misunderstanding the correct usage of data/nodes/features then in the 3dsMax SDK.
Users/artist may end up messing with nodes and connect things that do not make sense and therefor give them results that are hard to understand and hard to visually debug what is going on.
I can totally agree to that. 3dsMax modifier stack is much more obvious and simpler to use for an artist.
That is important, but has nothing to do with program stability.
I would defer to your knowledge, you would know far more about both systems than I would. If you think its more stable and easier to work with (on the backend) I believe you.
I was more or less thinking out loud and hoping someone could help fill in the gaps. I thought I had some info on the differences between the two but that wasn't quite right. Which is why I started asking questions. I'm not trying to promote one over the other just trying to understand where they differ and how, which one is better at what and why.
It seems to me that max is a little more stable than maya, I have very little proof of that other than anecdotes and personal experience. Max has been buggy in the past and they put a focus on improving its stability, just like it seems the Maya team has, but max got more stable and it seems like Maya is just as buggy as it has always been, maybe a little better in some areas and worse in others. Like some of the interactive tools are scary to use and can cause instant crashes without warning, at least for me. Maybe I was doing something wrong, but I thought I was using the tools as designed and whatever I did that it didn't like, wasn't communicated back to me so it will keep on happening. It's that kind of instability that lead me to question if the node based approach might make it harder to repro and fix bugs? Maybe there are some tools to help with that kind of debugging? I don't know.
Maybe after they finish innovating they will focus on stability again. Hopefully the non-linear flow doesn't make providing stability a big challenge?
Maybe this isn't right but I'm looking at it as if I was following a treasure map with a single defined path vs staring at a street map of the entire earth with no defined path and trying to figure out how it is getting from point A to B. It seems like it might be confusing not just for the user but for the programer as well? Maybe you guys have ways to deal with it? I'm not sure, if you could shed some light on it that would help me understand and the next time the topic comes up I'll have a more complete picture to work with.
I ve used max for 4 years, maya 2 years. Overall, it seems to me Max is more stable than Maya.
All I know is maya is very prone to crashing when you model in smooth preview mode and using tools like bevel does all kinds of crazy things such as invisble faces, mesh moving back to origin while beveling. Not only that, combining and seperate mesh creates a lot of useless nodes. In the end I still prefer to use maya because of the amount of polys it can handle and the scene management- being able to edit things globally ( editing in component mode with several mesh selected, Uv editor etc.
I use to be one of those max users that swear by Max and hated Maya with a passion.
a node is not= to a modifier. its fundamental. everything in maya is a node. from an animation key to a particle emitter to a edge collapse. there are lots of nodes that are not designed to be edited in the gui. if you try to think of nodes like modifiers your going get a headache and get into trouble.
know what shape and transform nodes are and what there doing and why. which nodes you edit and which you do not. play around with connection editor so you can see whats really there and connect inputs and outputs. for more complicated setup this is where the power is.
know what inputs and outputs are and how to navigate them. ie, attribute editor arrow keys, hyper-shade etc.
use the channel box stack. its as close to linear as your going to get.
delete non deformer history is your homeboy.
blind data is not your homeboy.
etc.
shaderfx: I can totally agree to that. 3dsMax modifier stack is much more obvious and simpler to use for an artist.
for me its not that way. i think its just that most people never really get an introduction to how maya works. so its all a big mystery and seems very opaque and complicated. as for speed and efficiency that tends to be tied more to getting use to the tools to the point where you don't have to think about it anymore and you just "DO" stuff. you can do that in both packages and modify your work flow to get faster and faster.
Replies
On a side note,these features like grease pencil,vertex modeling,fluid retopology have existed in Blender since.But I am guessing a lot of studios will have to stick with Autodesk because they have obviously spent money and time familarising their staff with the apps and built plugins and inhouse tools around it.Switching to another app is gonna cause more money and time.
Has any commercially released game ever used the Blender game engine?
And I don't think Blenders paint feature will ever be taken seriously until it has proper layer support (ie. Photoshop). Though I guess the guy who made B-Mesh is working on something like that, but the last build he did was a few months ago and crashes Blender 100% of the time.
Not that I know of.But I have watched the demos made with it and they pretty impressive.Commercial games seem to go for udk most of the time.
Well,I have used the texture paint feature and got good results.There is an addon that comes with Blender called Texture layer manager.It enables u to create layers easily with mix,add e.t.c.
Utterly ridiculous what there charging for subs, especially in EU for the little features you get. When Camera Sequencer is classed has a major update and looks like something that could be done in ICE relatively easily, then you see Modo putting out a release that has more features/fixes than Autodesk has done for all 3 of its major 3D apps over the last couple of years, you know innovation went out the window a long time ago at Autodesk.
Wouldn't be too surprised to see Max going the route theyve chosen for XSI and classing it has some support package, since they want everyone on the suites. Develop one package with the other two thrown in as support, charge you close to the cost of all three, yet developing only one!
exactly, maya is like a dinosaur in between these programs...
yes, but it is also true that modo is still missing a lot of stuff and it is easier to add new high profile features. BUT, the way these features are added puts autodesk to shame.
THIS!^^
From this
http://www.polycount.com/forum/showthread.php?t=68173
though, of course lets not forget, Max launched in '96, Maya appeared in '98, Softimage (nee XSI) in 2000, Modo 2004, Blender '95, Houdini '96, and Cinema4D '96 -although it was available on the Amiga before that, same as Lightwave. So you could say, most of the 3D software today has been around abit.
I've tried it, can't even figure out how to delete layers, and its got problems with seams which Blenders paint normally doesn't. The UI also makes it completely unwieldy for more then maybe 10 layers maximum. It needs a ton of work.
You are right, I was ultimately thinking about previous alias products that originated the ui of maya...
So it's safe to say that they are all dinosaurs except for modo?
I think right now they are focused more on upgrading the core and behind the scene stuff rather than adding more features, but it's just common sense to me that when they are done doing that they'll start adding all the features that they know is missing in there...
No, if i recall its all in python/pyscript so its unbelievably slow.
What do you think the would choose then? I thought Max's modelling tools were excellent?
Is it just Modo that beats it?
I am unfamiliar with Max, is it not like the Maya's input nodes that are shown in the channel box stacked as history?
Modo rules for modeling on my side nowadays...
As it is now, everyone around them uses Max and every job application lists Max as a required skill so ... they choose Max. Not because it's necessarily the best tool for the job.
I sort of have a thirst for learning new things so I try a lot of programs just for the hell of it, which is how I found Modo. You never know what's out there until you look.
Maya uses a Dependency Graph were it passes attributes along it nodes to accomplish something, whereas Max and XSI have a stack system were the attribute/nodes can be reordered and selective history kept, wheras in Maya the nodes are dependant on their previous nodes, hence the name Dependency Graph.
Modo dosen't have history at all.
Maya has a history stack that can't be reordered. You are pretty much forced to work very linearly. Going back down the stack and making changes can be a dicey operation with horrific consequences farther up the stack. I think most people collapse the history and just muscle through with poly wrangling when they could be working a little less linearly and not spend so much time brute forcing their way to the final result.
Max has a modifier list, that can be reordered. The stack in max isn't dependant on previous operations being in a specific order. It can operate on an object level so it can be independent from changes made to the geometry. Meaning you can add a bend, go back down the stack and make changes to the It's modifiers become more like photoshop adjustment layers rather than a recorded event history list. You can work non-linearly with quite a few different modifiers ordering them however you need.
You can leverage the stacks functionality when creating modular assets, like bending walls, pipes or ducts. Go back down the stack and toggle a few modifiers on or off, reorder the stack, insert new modifiers anywhere in the stack and from the same base shape you get the various pieces all without fear, that changes made below will cause catastrophic results above. I can go back down the stack and make changes to the base straight piece, adding or removing geometry as I need, then I can toggle the modifiers and extract the various pieces that I need. At any point I can toggle on or off various modifiers to see how each piece will look.
If a modifier up in the stack freaks out it's because I left a sub-object selection lower in the stack and the modifier is working on that selection instead of the whole object. Which is a simple fix, go back down the stack, exit sub-object mode properly (like I should have) and everything is fine. In Maya, those kinds of operations have me very very nervous and I wouldn't risk it, I would stick to working in a straight line and would probably iterate less and be much less likely to try various things.
Thanks for the thorough explanation of the difference guys.
that is not completely accurate as far as i know.
you can in fact reorder many of the deformers from inputs panel. you can MMB click-drag individual inputs and change order.
For example, in max I can insert the edit poly modifier before or after the bend modifiers and make drastic changes to the geometry or add additional object, you can't do that in Maya can you? Maybe you can...
yes you can in most cases. if you just modify topology to the current mesh and keep the bend/lattice and lot of other deformers without deleting history then you can keep making changes to topology and still have the deformers affect the changed topology.
however, it may not be possible in all cases. it depends on type of merging/editing you are doing.
awesome thing about Maya is it is node based, so as long as you dont merge multiple nodes, you can keep editing that particular node. you can also remove certain poly tweaks from hypergraph.
it may not be possible in all cases, depends on type of merging/editing you are doing.
actually its the other way around.
max has a linear modifier stack which re-evaluates when you change state. ie it is a DAG (Directed Acyclic Graph). the reason why things work out when you move modifies is that it re-evaluates the whole stack from the root upwards in a linear fashion. the problem with this is that is forces a large amount of memory copying. so when your stack starts to grow so does your memory. large stacks on dense geometry becomes unstable.
maya on the other hand uses what is known as a DG (Dependency Graph). maya also has a DAG built on top of the DG. the DG or the DAG can be used to make nodes. nodes pass messages through the DG or the DAG. nodes in the DG have a non-linear evaluation. nodes in the DG update only when they are called to do so by another node that needs some information to change its state. the reason why you do not have a nice little linear modifier stack window in maya is because nodes are not linear and live in a graph. you can view the maya DG in the 'hypergraph'. the DG solves the problem of excessive memory copying. you can have hundreds of thousands of nodes that operate on geometry with out copying that geometry in memory.
it would take a lot of print to explain how to work with the DG to be efective and get what you want. it takes more work to learn then a simple DAG. its a bit non-intuitive from the outside. but it works magic when you understand it.
undo is linear in both maya and max there is no branching in the undo in either package. the differences are related to edits of the DG vs DAG.
max 2014 is no longer bloated and has many common bugs fixed, caddies can be disabled and you can save files all the way back to max 8
april fools
It also seems like Maya could potentially be a bit more buggy with its non linear path and could potentially be harder to track and reproduce bugs with such a diffused network of nodes, max might be a bit more stable or at least easier to track and reproduce because it's going in a straight line.
It seems like it might take max a little longer to plow through a long stack but maya might actually become more unstable the more nodes you toss in there? Especially if it isn't evaluating them all? It might have dormant nodes that seem to be working until they get passed some info and update? They probably have a way of dealing with that, but I have spent more time crashing in Maya than I ever had in max, tho I chalk that up to my inexperience.
It seems like Maya's problem is UI design related while max is just screwed in the performance department, ha. The Maya team could actually straighten out the UI if they wanted but for whatever reason they aren't?
most maya user arnt using the UI at all..
i use 90% shortcuts and marking menues...
I guess the audience feel the same about the intro, as we do about the new features.
i'm not suggesting one is better. what i was pointing out was some of the core design differences and why things feel and work different in the two packages. having a grasp of that helps to get a feel for whats going on under the hood.
maya at its core was designed to solve two problems with the DG and its design.
1 scale:
how do you grow very complex scenes with lots of components all working together in a way that is efficient.
2 pipeline:
how do you integrate different components like animation simulation etc into a coherent whole that can be managed in a sane way.
it does those things very well and provides a solid platform.
as for modelling i'm not going to get into that. but you can make it happen if you put in the time and develop your work flow. it will be different then your max work flow but just as fast and efficient. that's my experience anyways.
I was previously a Max fan and am a 100% Maya nerd now. haha
Sorry, is that just something you made up or how do you justify the relation between being "non linear" and "unstable"?
Just because something is "non linear" it is more likely to be "unstable"?
I would argue the two are not related at all.
If anything, Maya's 'goal' is that one node does not need to know anything about the next node to be able to interact with each other. To me it seems like this would benefit stability.
Having used both, I would say that in Maya's API it is less likely (though certainly not impossible!) to breaks things by misunderstanding the correct usage of data/nodes/features then in the 3dsMax SDK.
Users/artist may end up messing with nodes and connect things that do not make sense and therefor give them results that are hard to understand and hard to visually debug what is going on.
I can totally agree to that. 3dsMax modifier stack is much more obvious and simpler to use for an artist.
That is important, but has nothing to do with program stability.
I was more or less thinking out loud and hoping someone could help fill in the gaps. I thought I had some info on the differences between the two but that wasn't quite right. Which is why I started asking questions. I'm not trying to promote one over the other just trying to understand where they differ and how, which one is better at what and why.
It seems to me that max is a little more stable than maya, I have very little proof of that other than anecdotes and personal experience. Max has been buggy in the past and they put a focus on improving its stability, just like it seems the Maya team has, but max got more stable and it seems like Maya is just as buggy as it has always been, maybe a little better in some areas and worse in others. Like some of the interactive tools are scary to use and can cause instant crashes without warning, at least for me. Maybe I was doing something wrong, but I thought I was using the tools as designed and whatever I did that it didn't like, wasn't communicated back to me so it will keep on happening. It's that kind of instability that lead me to question if the node based approach might make it harder to repro and fix bugs? Maybe there are some tools to help with that kind of debugging? I don't know.
Maybe after they finish innovating they will focus on stability again. Hopefully the non-linear flow doesn't make providing stability a big challenge?
Maybe this isn't right but I'm looking at it as if I was following a treasure map with a single defined path vs staring at a street map of the entire earth with no defined path and trying to figure out how it is getting from point A to B. It seems like it might be confusing not just for the user but for the programer as well? Maybe you guys have ways to deal with it? I'm not sure, if you could shed some light on it that would help me understand and the next time the topic comes up I'll have a more complete picture to work with.
All I know is maya is very prone to crashing when you model in smooth preview mode and using tools like bevel does all kinds of crazy things such as invisble faces, mesh moving back to origin while beveling. Not only that, combining and seperate mesh creates a lot of useless nodes. In the end I still prefer to use maya because of the amount of polys it can handle and the scene management- being able to edit things globally ( editing in component mode with several mesh selected, Uv editor etc.
I use to be one of those max users that swear by Max and hated Maya with a passion.
a node is not= to a modifier. its fundamental. everything in maya is a node. from an animation key to a particle emitter to a edge collapse. there are lots of nodes that are not designed to be edited in the gui. if you try to think of nodes like modifiers your going get a headache and get into trouble.
know what shape and transform nodes are and what there doing and why. which nodes you edit and which you do not. play around with connection editor so you can see whats really there and connect inputs and outputs. for more complicated setup this is where the power is.
know what inputs and outputs are and how to navigate them. ie, attribute editor arrow keys, hyper-shade etc.
use the channel box stack. its as close to linear as your going to get.
delete non deformer history is your homeboy.
blind data is not your homeboy.
etc.
for me its not that way. i think its just that most people never really get an introduction to how maya works. so its all a big mystery and seems very opaque and complicated. as for speed and efficiency that tends to be tied more to getting use to the tools to the point where you don't have to think about it anymore and you just "DO" stuff. you can do that in both packages and modify your work flow to get faster and faster.