Better glow/illum options, normal attenuation fade on glow material with alphas
More "skies"/environments
UV Channel and tiling settings.
For example my main use for marmo is while texturing, would be useful to quickly be able to tile a texture if you are working on a tiling texture.
Oh yeah and do something about the licenses. Give at the very least a floating style which would let you open it either at home or at work as long as only 1 copy is open at a time... like everybody else says, tedious DRM only punishes the people who purchased your software, pirates just copy paste an exe and forget
been using marmoset a lot lately, really cool to see you guys reaching out tothe community for suggestions.
I agree to many of the features mentioned here. they've mostly been mentioned. Just throwing my support and 2cents behind...
______
- Dustin's suggestion to have independent control over the reflection map would be very welcome. using a texture mask to help control this would be great.
- sync'd tangents
- I'd like materials to stay in the scene when you update the .mesh or .obj. Also i've noticed that if you have a .mesh file loaded. and you choose to replace this with a .obj the scene camera will often change. Ive got a work around where i duplicate the camera and switch back to this after the import. only seems to happen when going from a .mesh to a .obj. Having materials try and auto apply if you update the obj would be great to as has been mentioned here.
- a more advanced SSS shader would be great if possible.
- a save all function. its annoying having to save mesh and material. .. save post processing and save scene. 3 steps to save my "scene" A true "save scene" which saves all these separate functions at once would be welcome. (keeping them their separate .mesh .post is still very cool though)
- a save lighting rig preset. i think this was mentioned earlier.
- agree with dustin a split view or even having per camera attributes so i could switch to another camera where dof was disabled to adjust lights. both even.
- glass shader
- had a strange bug recently where if i alt tabbed to say firefox then back to marmoset my spec map would be blown out and reading wrong. id have to remove the map and re apply for it to fix its self. (windows 7 x64, 310.xx driver, nvidia 460 )
I'm seeing a lot of people saying "sorry if this has been mentioned" or something to that effect. Don't let this stop you from posting! Think of this thread as more of a poll or a survey, the more people that post and share their opinions the better. The more users we get asking for similar features or improvement the more important we know those suggestions are. Even if you just want to post to quote some else's suggestion, that's great.
+1 to being able to pan UVs/textures to get some animated stuff going. I guess adding the framework for being able to keyframe stuff, and more things can be added on later.
+1 to saving everything within a scene. It's still awesome to be able to save out material and mesh files separately if needed, so both of these together would be ideal.
The ability to increase how much an emissive texture adds to the bloom effect, without having to increase bloom across the entire thing. I found when trying to fake colored lights on a character with emissive maps, I had to crank it all the way up in my material which basically made that part of the texture completely white, or I had to increase bloom across the board which made things I didn't want to bloom. Neither really got the effect I wanted which was bright, bloomy colored lighting.
Would also love to have a basic particle emitter or something, though that's getting into territory you guys might not want to head.
Something I suggested some time ago, is a way to import cameras from the 3d package via fbx for example, so you can keep the same camera angle as when/if you render the highpoly.
As stated in the toolbag .pdf, material order is important... yet there is no system in place to rearrange materials. You have to remove the materials and open them in the order that you need. So that is what I would like to see as a new feature.
-some sort of vfx support
-saving is really weird and clunky. Even if I save a scene, mesh and material I still get asked every time I quit if I want to quit without saving something I already saved.
its been said many times, but as this is more of a survey/poll, here I go.
Reassign materials upon mesh reload.
Independent reflection control
Independent emissive bloom control.
More texture channels for floor.
Tiling ocntrol of textures.
Don't know if the last one is just me, but when doing a turntable export in image sequence, i have to start a mesh rotation and then hit the turntable export button, thus never being able to get it to start on the frame that I actually want. Then I have to duplicate the sequence in After Effects and slide it around to get it to where I want to be. No big deal, but a little time saver on this would be great
Better Ambient Occlusion. Screenspace AO is amazing in next gen viewports, would love to see more detailed implementation in Marmoset, along with the ability to get it darker ( or even tinted)
Better SSS.
Better DX11 tessalation?
Save light rigs, along with being able to name lights.
Have it be possible to have lights affect only certain chunks of meshes?
I'll probably come back and edit this when I think of more!
Definitely appreciate getting feedback from users on polycount
For license issues I got a pirated .exe, easier for me to deal with, I've paid for a license, though if it was more portable I wouldn't bother with the pirated .exe.
Would love to see:
Multi-material support (B option, pref read in via fbx)
Saving the scene in one save, though with the option to export materials kept for replacing materials.
uv channel selection
scaling tileable textures/uv's
synced tangents
toggling map visibility, eye icon
Would be nice but not fussed:
Relative file paths
Improvement to shaders/material editor, a node editor like udk has would be great.
Ability to use individual channels of an image for materials.
Animation would be nice, updated plugins or a method that doesn't rely on waiting on plugins, bought marmoset early 2012, no updates on this since.
Which feature(s) of Marmoset Toolbag do you find the most useful?
I agree with Baj Singh: "One thing that can be a little annoying, opening a new mesh in Marmoset resets the .mat settings so they have to be reapplied (this is the case when you have a mesh already in the scene or when you auto-reload a mesh)."
I also agree with the comments regarding licensing and disassociating it from your computer fingerprint. Though in defense of the Marmoset guys, they've always been super fast about getting me a new serial to activate Toolbag.
Which new feature(s) would you like to see in Marmoset Toolbag?
User-defined baker-specific tangent space support
Multi-material support by material ID
The ability to name your lights.
A split view, so you can keep one view focused on your character while you move your light around in the other view.
What do you like least about Marmoset Toolbag?
That it doesn't handle normal maps with mirrored UVs very well.
Might it be possible to open it up so we can write/submit/share our own shaders?
This one would be what I'd like to see. If it could just read standard cgfx shaders that would be fine. Adding its own node-based shader creator is overkill.
I also see people asking for UV-panning and whatnot, and this would solve that problem as well.
I have to second a particle system eventually. The advantages to having a minis scene are great.
For example:
Let us say that I was presenting an arena net art test, and I wanted to have a fire effect, it would add a decent amount of atmosphere mixed with smoke coming out the back.
Animated texture things, I could make the window glow flicker and add a lot of awesomeness to the presentation. Sure, I could do that in a game engine but don't you want people to use marmoset to present?
Also that mixed with better tile-able support would be amazing.
Then I could use Marmoset for presentation.
Also, I don't know if this is in there since I can't use lights in my marmoset in Windows 8, but having area lights not just spot lights.
It sounds to me like you people need a full on functional render/game engine instead of a 'simple' model viewer. They might as well release their game engine for all stuff that is being asked, especially since half of what is being asked is a direct toolset from UDK.
I bought Marmoset back in 2010 (If I remember correctly), so I've been a long-time user... I guess...
I've generally found pretty much every-and-anything put into previous updates useful. It's difficult for me to think of specific features that I've found particularly useful.
The only things I'd like to see in future releases are:
- the previously mentioned cgfx shader thing
- the ability to select multiple mesh chunks and apply a material to them. You have no idea how much time this would save on a long-term basis. If I could shift-select multiple mesh chunks and apply a material to them, that would save at least two minutes.
- the previously mentioned particle thing
As for what I like the least about Marmoset Toolbag, there is one issue that comes to mind. When I have faces on two objects a little close together (not close enough to produce z-fighting in Maya or Max, mind), and add a light, it causes some pretty bad z-fighting (at least on this computer - never tried it on other setups...).
That's all I can think of. Thanks for making an awesome product!
It sounds to me like you people need a full on functional render/game engine instead of a 'simple' model viewer. They might as well release their game engine for all stuff that is being asked, especially since half of what is being asked is a direct toolset from UDK.
Then I might as well use UDK. The reason I like Marmoset is that it gives me those same features, but it's a thousand times easier to set up a scene and tweak to get a good result, as well as more direct control over things like Post Processing. As far as just being a simple model viewer/presentation tool, I'd say it's already there.
Then I might as well use UDK. The reason I like Marmoset is that it gives me those same features, but it's a thousand times easier to set up a scene and tweak to get a good result, as well as more direct control over things like Post Processing. As far as just being a simple model viewer/presentation tool, I'd say it's already there.
This is what I'm trying to say. I bought Marmoset, I do like it a lot, but I don't do a ton of props that in my opinion are worth looking at by themselves and would generally be accompanied by AT LEAST a few props.
I love Marmoset, but these things we ask for are not a total game engine, just making it something use-able for mini-scenes at the very least.
If that was the case I could use it for a lot more things.
We don't need a terrain system, a complex material editor, programming and all sorts of stuff, just a few more features to make it a better and more versatile program over all.
I read that version 2.0 would be a full game engine though, and that would be amazing. (where ever I read that.) And a game has already been made with the engine as well.
A usable workflow for creating scenes, not just viewing singular assets is certainly going to be a big priority for us. Sure you can set up some of it(like multiple meshes etc) in your 3d app and export a scene, but the current UI and scene handling isn't really set up well to deal with it. I think its clear here with a lot of the comments that scene management is a very important issue. So far I don't think many of the features and improvements requested have been all that unrealistic(not to say we'll be able to get everything in of course). So yeah, I definitely do not want to discourage anyone from making suggestions, even over the top suggestions.
However, one thing I want to clear up(and you're probably referring to one of my posts Alex): We currently have no intention of doing a full on game engine. Marmoset Toolbag is based on the rendering pipeline we used for the game Darkest of Days. Marmoset Toolbag is essentially the rendering engine with most of the game specific (ai, pathfinding, physics) stuff stripped out. Rendering wise it is very much a game engine, just without all the extra stuff that it takes to you know, actually make a game.
We are going to continue to improve and expand the feature set of what Marmoset is really good at, a straight forward, easy to use presentation tool. Call it a model viewer, call it a real-time render, whatever you like. That's our thing.
I guess I'll toss in what I'm looking for too, in order of priority.
It's also worth mentioning that even though it may be unrealistic, I'd really like marmoset to come as close as possible to the graphical prowess of Cryengine and Unreal.
Better handling of license stuff, I've had to request a new license key several times over the years
Ability to reload a mesh and still have the mat applied
-more sky/lighting presets
-glass shader with options
-a shader with reflections.
-reload a mesh and still have the mat applied
-outline by edge texture color
...Better Ambient Occlusion. Screenspace AO is amazing in next gen viewports, would love to see more detailed implementation in Marmoset, along with the ability to get it darker ( or even tinted)
Even though the AO slider goes from 0-1, you can type in numbers higher than 1 to darken the AO.
- Hoping there is a way to script it or atleast open the exe to taking in commandline arguments so one can export out of maya say and launch marmoset with an argument for the fbx file to import.
- +1 to the idea of fx files.
Complex refraction within complex refraction (e.g.water within a glass).
Alphas, transparency & Gloss packable into a single RBG texture rather than using the alpha of individual textures...currently it is a pain in the ass to save out iterations quickly from a single open PSD file.
Support for mikktspace (so bakes from xNormal will match) and other tangent spaces.
Relative remapping for save project files and materials so you can change the location and everything still works.
The ability to separate meshes into chunks within Marmoset.
Particle goodies for simple creation of things like fire etc.
Animation support for Blender, Max and Maya.
SSDO support.
Support for other image formats such as png.
Animated material for water (2 normals blended together and offset + flow map + foam + Complex Refraction).
The ability to displace the ground plane and have it have a full material and set of textures.
Which feature(s) of Marmoset Toolbag do you find the most useful?
Marmoset is super easy to pick up and use to previs work for clients, pushing stuff into properietary engines or even udk or cryengine takes longer and is less performancehungry
Which feature(s) or function(s) of Marmoset Toolbag do you think would benefit from a workflow overhaul?
- if i save a scene, i'd like to reload the model without losing my material presets, right now it resets everything if i save over an existing model or use the load function in a saved .mesh file
- let us rotate the skydome in every direction, not just horizontally, often enough i like the light andcolours but want the sun to shine from a different angle
Which new feature(s) would you like to see in Marmoset Toolbag?
- I think DX11 support would be great, better skinshading, a modern hairshader, etc.
- custom shader support
- better shadows
- better AO
- cubemaps with configurable blur
- material IDs
- a webplayer
What do you like least about Marmoset Toolbag?
the material reset thingy, and well that it tends to look aged, compared to current builds of the cryengine or udk
What do you mean by vfx? Particle system or something?
Yes, perhaps a particle system is beyond the scope of Toolbag but it would be great to have some simple vfx support. Maybe it could be added as an addon module for some extra money.
Yes, perhaps a particle system is beyond the scope of Toolbag but it would be great to have some simple vfx support. Maybe it could be added as an addon module for some extra money.
Well we actually had a full on particle editor at one point within Toolbag, so its not an outrageous idea. It was in the beta version I think but dropped for V1 because of technical issues, middle-ware/physics issues causing crashes on certain hardware.
We're doing some massive overhauls to a lot of our systems so I can't really tell you guys if a particle editor is feasible or likely, but I'll say its not outside the bounds of reality.
Complex refraction within complex refraction (e.g.water within a glass).
Alphas, transparency & Gloss packable into a single RBG texture rather than using the alpha of individual textures...currently it is a pain in the ass to save out iterations quickly from a single open PSD file.
Support for mikktspace (so bakes from xNormal will match) and other tangent spaces.
Relative remapping for save project files and materials so you can change the location and everything still works.
The ability to separate meshes into chunks within Marmoset.
Particle goodies for simple creation of things like fire etc.
Animation support for Blender, Max and Maya.
SSDO support.
Support for other image formats such as png.
Animated material for water (2 normals blended together and offset + flow map + foam + Complex Refraction).
The ability to displace the ground plane and have it have a full material and set of textures.
About 6 months back I tried to build an environment piece (http://vimeo.com/45562813) with the intent of using Marmoset to render it out (didn't know how well Marmoset would take it).
From that experiment, I realized marmoset wasn't really set up to work with environments as much as center pieces (characters and smaller grouped assets) though it is possible in a limited way.
In addition to what most have already said, if there is any room for improvement it would probably be for marmoset to include larger scale scenes as part of its primary design (for environment artist). This mixed with a means to have more camera control (tracks, splines) would be great for environment artist wanting to show scene based content.
I would love to see Marmoset have some form of water shader, and of course some particle effect capabilities (loading sprites or presets).
OSX Support (that's a huge one I think there is no alternative aside from Unity on OSX)
Adjusting the emissive color through color picker as well as lighting with the emissive texture
Gloss map support (EDIT: I keep forgetting it's in the alpha channel of the spec map atm)
Material IDs (Current chunks system makes sense but can be a chore sometimes. Seems like this would be easier)
Improved alpha support.
A GoZ style set up for Maya/Max (at least) would be amazing honestly. To click a button in either of those programs and have the selected mesh open up right in Marmoset would be great. Not sure how feasible that would be.
The program works great for characters and single assets. Seems like the camera clipping planes (or something similar) makes it difficult to bring in full scenes. I would love to see that change. At least being able to put together small scenes in the Toolbag would be nice.
Awesome that you're asking for user feedback and trying to implement it. I'm really looking forward to what you guys have cooking.
never used it, but please turn it into the 3d artist's diorama maker
add scripting (LuaJIT yay) so one can program animations of material/camera or whatnot. Assuming most things stay static.
ideally use json/lua for storing material settings for a file (judging by feedback on material presets are not stored, keep "non geometry" related stuff in separate human editable fileformats).
add all optic effects that make Joe happy
In the end it's like a photography studio for your own models
I think more material to actually learn this software. its simple yes but most people dont know how to even use it effectively.
Make a character tutorial showing how to do make skin look amazing, how to make different types of cloth materials, amazing eyes etc.
And please make a video tutorial to go along with, I think most people find it so much easier to learn by actually watching someone doing it.
completely disagree with this. what you're asking here is to teach people how to be better artists. that's not the job of the marmo guys at all... by all means teach people what each slider does, or teach them how to set up a material. but you can't in all fairness ask them to teach you how to make skin look amazing, as that's up to YOU the artist, to do.
it doesn't matter how good you are at setting up your materials, they could give you the ultimate preset. but if you're a shitty texture artist, the end result will still look shitty.
More and better documentation is certainly something we're planing on doing. We just released a tutorial going over the material system with tips on how to represent different material types withing Marmoset Toolbag, which you can find here: http://www.marmoset.co/toolbag/learn/materials
We've also got a tutorial covering displacement mapping that we will be releasing shortly as well.
While I would say we will probably not focus too heavily on art-centric tutorials (because there are many good resources for that sort of stuff, Eat3d and the like) there is a lot we can do to fill out the more technical and tool-specific side. A tutorial going over the skin shader is a great example and is certainly on my short list of topics to cover. There are a lot of technical aspects to things like hair, eye reflections, etc that would benefit from clear documentation as well, so I can't say I agree with the sentiment that there wouldn't be a need for those sort of tutorials. You're not likely to see me posting any tutorials on how to sculpt characters in zbrush or anything like that, but the technical side, I see a lot of potential for documentation there.
Video tutorials are something we may consider in the future, we've been sticking to text tutorials because we lack a real extensive "help" style documentation system at the moment, so its good to have things documented in text.
by all means, document what the tools do, as you did in your other tutorial. i think it's something that should have been done a long time ago!
but there's a difference between explaining the tools so that artists can use them, and giving recipes for "amazing skin", which just builds unrealistic expectations. because that recipe will only really be useful for the textures that you created for that specific tutorial... an end user who isn't as good at texturing will end up using all the same settings and then going "why does mine look so bad?! YOU LIED TO ME!"
by all means, document what the tools do, as you did in your other tutorial. i think it's something that should have been done a long time ago!
but there's a difference between explaining the tools so that artists can use them, and giving recipes for "amazing skin", which just builds unrealistic expectations. because that recipe will only really be useful for the textures that you created for that specific tutorial... an end user who isn't as good at texturing will end up using all the same settings and then going "why does mine look so bad?! YOU LIED TO ME!"
it leads to more harm than good.
No I agree, thats what I meant about not making "art-content" specific tutorials, IE: how to create the actual art content, which I think is what you mean. I just wanted to make a differentiation there between what we can document on the tutorial/technical side, and what we can't really give you from a tools perspective(the actual art content).
I think its important to split those two things up and understand the differences, which is similar to what you were saying I think. Really I just wanted to let everyone know what you can realistically expect from us.
Which feature(s) of Marmoset Toolbag do you find the most useful?
It's overall ease of use, in terms of the UI
Which feature(s) or function(s) of Marmoset Toolbag do you think would benefit from a workflow overhaul?
I'd like to see cubemap reflections become independent of the sky lighting system, so I can turn down sky lighting and use my own lights, but cubemap reflections would still be visible. I'd also like to be able to import my own cubemap DDS files.
As much as I love the shader presets, the biggest downfall I see to using Marmoset is the inability to access a node-based material editor to make custom shaders. Not everyone would use it, but it would put more power in the hands of those who are interested in doing such things.
I agree with Baj Singh: "One thing that can be a little annoying, opening a new mesh in Marmoset resets the .mat settings so they have to be reapplied (this is the case when you have a mesh already in the scene or when you auto-reload a mesh)."
I also agree with the comments regarding licensing and disassociating it from your computer fingerprint. Though in defense of the Marmoset guys, they've always been super fast about getting me a new serial to activate Toolbag.
Which new feature(s) would you like to see in Marmoset Toolbag?
User-defined baker-specific tangent space support
Multi-material support by material ID
Skin shader improvements
Flow map support
The ability to name your lights.
Higher quality shadow maps and more accurate self-shadowing
A split view, so you can keep one view focused on your character while you move your light around in the other view.
What do you like least about Marmoset Toolbag?
That it doesn't handle normal maps with mirrored UVs very well. This is almost certainly due to the lack of a synced TS.
Replies
This!
More "skies"/environments
UV Channel and tiling settings.
For example my main use for marmo is while texturing, would be useful to quickly be able to tile a texture if you are working on a tiling texture.
Oh yeah and do something about the licenses. Give at the very least a floating style which would let you open it either at home or at work as long as only 1 copy is open at a time... like everybody else says, tedious DRM only punishes the people who purchased your software, pirates just copy paste an exe and forget
Yea this is a really good idea. I like using the built in floor better than importing a plane but its features are limited.
I agree to many of the features mentioned here. they've mostly been mentioned. Just throwing my support and 2cents behind...
______
- Dustin's suggestion to have independent control over the reflection map would be very welcome. using a texture mask to help control this would be great.
- sync'd tangents
- I'd like materials to stay in the scene when you update the .mesh or .obj. Also i've noticed that if you have a .mesh file loaded. and you choose to replace this with a .obj the scene camera will often change. Ive got a work around where i duplicate the camera and switch back to this after the import. only seems to happen when going from a .mesh to a .obj. Having materials try and auto apply if you update the obj would be great to as has been mentioned here.
- a more advanced SSS shader would be great if possible.
- a save all function. its annoying having to save mesh and material. .. save post processing and save scene. 3 steps to save my "scene" A true "save scene" which saves all these separate functions at once would be welcome. (keeping them their separate .mesh .post is still very cool though)
- a save lighting rig preset. i think this was mentioned earlier.
- agree with dustin a split view or even having per camera attributes so i could switch to another camera where dof was disabled to adjust lights. both even.
- glass shader
- had a strange bug recently where if i alt tabbed to say firefox then back to marmoset my spec map would be blown out and reading wrong. id have to remove the map and re apply for it to fix its self. (windows 7 x64, 310.xx driver, nvidia 460 )
I'm seeing a lot of people saying "sorry if this has been mentioned" or something to that effect. Don't let this stop you from posting! Think of this thread as more of a poll or a survey, the more people that post and share their opinions the better. The more users we get asking for similar features or improvement the more important we know those suggestions are. Even if you just want to post to quote some else's suggestion, that's great.
+1 to saving everything within a scene. It's still awesome to be able to save out material and mesh files separately if needed, so both of these together would be ideal.
The ability to increase how much an emissive texture adds to the bloom effect, without having to increase bloom across the entire thing. I found when trying to fake colored lights on a character with emissive maps, I had to crank it all the way up in my material which basically made that part of the texture completely white, or I had to increase bloom across the board which made things I didn't want to bloom. Neither really got the effect I wanted which was bright, bloomy colored lighting.
Would also love to have a basic particle emitter or something, though that's getting into territory you guys might not want to head.
Also on par with most that is being said too.
-saving is really weird and clunky. Even if I save a scene, mesh and material I still get asked every time I quit if I want to quit without saving something I already saved.
Reassign materials upon mesh reload.
Independent reflection control
Independent emissive bloom control.
More texture channels for floor.
Tiling ocntrol of textures.
Don't know if the last one is just me, but when doing a turntable export in image sequence, i have to start a mesh rotation and then hit the turntable export button, thus never being able to get it to start on the frame that I actually want. Then I have to duplicate the sequence in After Effects and slide it around to get it to where I want to be. No big deal, but a little time saver on this would be great
Better Ambient Occlusion. Screenspace AO is amazing in next gen viewports, would love to see more detailed implementation in Marmoset, along with the ability to get it darker ( or even tinted)
Better SSS.
Better DX11 tessalation?
Save light rigs, along with being able to name lights.
Have it be possible to have lights affect only certain chunks of meshes?
I'll probably come back and edit this when I think of more!
For license issues I got a pirated .exe, easier for me to deal with, I've paid for a license, though if it was more portable I wouldn't bother with the pirated .exe.
Would love to see:
Multi-material support (B option, pref read in via fbx)
Saving the scene in one save, though with the option to export materials kept for replacing materials.
uv channel selection
scaling tileable textures/uv's
synced tangents
toggling map visibility, eye icon
Would be nice but not fussed:
Relative file paths
Improvement to shaders/material editor, a node editor like udk has would be great.
Ability to use individual channels of an image for materials.
Animation would be nice, updated plugins or a method that doesn't rely on waiting on plugins, bought marmoset early 2012, no updates on this since.
This +1
What do you mean by vfx? Particle system or something?
This one would be what I'd like to see. If it could just read standard cgfx shaders that would be fine. Adding its own node-based shader creator is overkill.
I also see people asking for UV-panning and whatnot, and this would solve that problem as well.
For example:
Let us say that I was presenting an arena net art test, and I wanted to have a fire effect, it would add a decent amount of atmosphere mixed with smoke coming out the back.
Animated texture things, I could make the window glow flicker and add a lot of awesomeness to the presentation. Sure, I could do that in a game engine but don't you want people to use marmoset to present?
Also that mixed with better tile-able support would be amazing.
Then I could use Marmoset for presentation.
Also, I don't know if this is in there since I can't use lights in my marmoset in Windows 8, but having area lights not just spot lights.
I've generally found pretty much every-and-anything put into previous updates useful. It's difficult for me to think of specific features that I've found particularly useful.
The only things I'd like to see in future releases are:
- the previously mentioned cgfx shader thing
- the ability to select multiple mesh chunks and apply a material to them. You have no idea how much time this would save on a long-term basis. If I could shift-select multiple mesh chunks and apply a material to them, that would save at least two minutes.
- the previously mentioned particle thing
As for what I like the least about Marmoset Toolbag, there is one issue that comes to mind. When I have faces on two objects a little close together (not close enough to produce z-fighting in Maya or Max, mind), and add a light, it causes some pretty bad z-fighting (at least on this computer - never tried it on other setups...).
That's all I can think of. Thanks for making an awesome product!
Then I might as well use UDK. The reason I like Marmoset is that it gives me those same features, but it's a thousand times easier to set up a scene and tweak to get a good result, as well as more direct control over things like Post Processing. As far as just being a simple model viewer/presentation tool, I'd say it's already there.
This is what I'm trying to say. I bought Marmoset, I do like it a lot, but I don't do a ton of props that in my opinion are worth looking at by themselves and would generally be accompanied by AT LEAST a few props.
I love Marmoset, but these things we ask for are not a total game engine, just making it something use-able for mini-scenes at the very least.
If that was the case I could use it for a lot more things.
We don't need a terrain system, a complex material editor, programming and all sorts of stuff, just a few more features to make it a better and more versatile program over all.
I read that version 2.0 would be a full game engine though, and that would be amazing. (where ever I read that.) And a game has already been made with the engine as well.
However, one thing I want to clear up(and you're probably referring to one of my posts Alex): We currently have no intention of doing a full on game engine. Marmoset Toolbag is based on the rendering pipeline we used for the game Darkest of Days. Marmoset Toolbag is essentially the rendering engine with most of the game specific (ai, pathfinding, physics) stuff stripped out. Rendering wise it is very much a game engine, just without all the extra stuff that it takes to you know, actually make a game.
We are going to continue to improve and expand the feature set of what Marmoset is really good at, a straight forward, easy to use presentation tool. Call it a model viewer, call it a real-time render, whatever you like. That's our thing.
It's also worth mentioning that even though it may be unrealistic, I'd really like marmoset to come as close as possible to the graphical prowess of Cryengine and Unreal.
-glass shader with options
-a shader with reflections.
-reload a mesh and still have the mat applied
-outline by edge texture color
-re sizable/color marmoset logo upon rendering
Even though the AO slider goes from 0-1, you can type in numbers higher than 1 to darken the AO.
- +1 to the idea of fx files.
Which feature(s) of Marmoset Toolbag do you find the most useful?
Marmoset is super easy to pick up and use to previs work for clients, pushing stuff into properietary engines or even udk or cryengine takes longer and is less performancehungry
Which feature(s) or function(s) of Marmoset Toolbag do you think would benefit from a workflow overhaul?
- if i save a scene, i'd like to reload the model without losing my material presets, right now it resets everything if i save over an existing model or use the load function in a saved .mesh file
- let us rotate the skydome in every direction, not just horizontally, often enough i like the light andcolours but want the sun to shine from a different angle
Which new feature(s) would you like to see in Marmoset Toolbag?
- I think DX11 support would be great, better skinshading, a modern hairshader, etc.
- custom shader support
- better shadows
- better AO
- cubemaps with configurable blur
- material IDs
- a webplayer
What do you like least about Marmoset Toolbag?
the material reset thingy, and well that it tends to look aged, compared to current builds of the cryengine or udk
Yes, perhaps a particle system is beyond the scope of Toolbag but it would be great to have some simple vfx support. Maybe it could be added as an addon module for some extra money.
Well we actually had a full on particle editor at one point within Toolbag, so its not an outrageous idea. It was in the beta version I think but dropped for V1 because of technical issues, middle-ware/physics issues causing crashes on certain hardware.
We're doing some massive overhauls to a lot of our systems so I can't really tell you guys if a particle editor is feasible or likely, but I'll say its not outside the bounds of reality.
This x2
From that experiment, I realized marmoset wasn't really set up to work with environments as much as center pieces (characters and smaller grouped assets) though it is possible in a limited way.
In addition to what most have already said, if there is any room for improvement it would probably be for marmoset to include larger scale scenes as part of its primary design (for environment artist). This mixed with a means to have more camera control (tracks, splines) would be great for environment artist wanting to show scene based content.
I would love to see Marmoset have some form of water shader, and of course some particle effect capabilities (loading sprites or presets).
OSX Support (that's a huge one I think there is no alternative aside from Unity on OSX)
Adjusting the emissive color through color picker as well as lighting with the emissive texture
Gloss map support (EDIT: I keep forgetting it's in the alpha channel of the spec map atm)
Material IDs (Current chunks system makes sense but can be a chore sometimes. Seems like this would be easier)
Improved alpha support.
A GoZ style set up for Maya/Max (at least) would be amazing honestly. To click a button in either of those programs and have the selected mesh open up right in Marmoset would be great. Not sure how feasible that would be.
The program works great for characters and single assets. Seems like the camera clipping planes (or something similar) makes it difficult to bring in full scenes. I would love to see that change. At least being able to put together small scenes in the Toolbag would be nice.
Awesome that you're asking for user feedback and trying to implement it. I'm really looking forward to what you guys have cooking.
Would be great with xnormal synced normals!
Support for a LUT function would be awesome as well!
Agreed, my hardware figerprint has changes a couple of times without even connecting a usb device and i usually reinstall Windows every 4 months.
Set what UV channel a texture/material uses.
Import and place more than one mesh.
Tiling, panning (static).
And a set of basic built-in meshes (sphere, cube, plane, cylinder, etc)
What tangent space does Marmoset use? Or does it use its own?
add scripting (LuaJIT yay) so one can program animations of material/camera or whatnot. Assuming most things stay static.
ideally use json/lua for storing material settings for a file (judging by feedback on material presets are not stored, keep "non geometry" related stuff in separate human editable fileformats).
add all optic effects that make Joe happy
In the end it's like a photography studio for your own models
A man after my own heart.
Currently Marmoset Toolbag uses its own proprietary tangent space.
poor environment artists get no love.
ability to create and animate a camera for fly throughs of a scene.
scene fog that can be tinted by distance with a gradient.
bounced lighting.
ability for objects to emit light using the emmisive channel.
translucency shader similar to the one frost bite2 uses
an easier way to switch licenses between systems.can we just have a login and user name like steam?
a better way to have more control over shadows, tint, softness, fall off.
Make a character tutorial showing how to do make skin look amazing, how to make different types of cloth materials, amazing eyes etc.
And please make a video tutorial to go along with, I think most people find it so much easier to learn by actually watching someone doing it.
In other words, they should hire this guy:
http://www.youtube.com/user/maya3dcg/videos?view=0
This as well, since we are running a pole. I also love the cubemap options that Neox mentioned, and everything he said really.
completely disagree with this. what you're asking here is to teach people how to be better artists. that's not the job of the marmo guys at all... by all means teach people what each slider does, or teach them how to set up a material. but you can't in all fairness ask them to teach you how to make skin look amazing, as that's up to YOU the artist, to do.
it doesn't matter how good you are at setting up your materials, they could give you the ultimate preset. but if you're a shitty texture artist, the end result will still look shitty.
We've also got a tutorial covering displacement mapping that we will be releasing shortly as well.
While I would say we will probably not focus too heavily on art-centric tutorials (because there are many good resources for that sort of stuff, Eat3d and the like) there is a lot we can do to fill out the more technical and tool-specific side. A tutorial going over the skin shader is a great example and is certainly on my short list of topics to cover. There are a lot of technical aspects to things like hair, eye reflections, etc that would benefit from clear documentation as well, so I can't say I agree with the sentiment that there wouldn't be a need for those sort of tutorials. You're not likely to see me posting any tutorials on how to sculpt characters in zbrush or anything like that, but the technical side, I see a lot of potential for documentation there.
Video tutorials are something we may consider in the future, we've been sticking to text tutorials because we lack a real extensive "help" style documentation system at the moment, so its good to have things documented in text.
by all means, document what the tools do, as you did in your other tutorial. i think it's something that should have been done a long time ago!
but there's a difference between explaining the tools so that artists can use them, and giving recipes for "amazing skin", which just builds unrealistic expectations. because that recipe will only really be useful for the textures that you created for that specific tutorial... an end user who isn't as good at texturing will end up using all the same settings and then going "why does mine look so bad?! YOU LIED TO ME!"
it leads to more harm than good.
No I agree, thats what I meant about not making "art-content" specific tutorials, IE: how to create the actual art content, which I think is what you mean. I just wanted to make a differentiation there between what we can document on the tutorial/technical side, and what we can't really give you from a tools perspective(the actual art content).
I think its important to split those two things up and understand the differences, which is similar to what you were saying I think. Really I just wanted to let everyone know what you can realistically expect from us.
+1