Converting selection between sub-object levels is limited, cumbersome and unintuitive. The Select By... window the fault of this. The Select By... can only convert from selected segments/splines to knots, but not from selected knots to segments/splines. And the workflow feels excessive, like if you are in knot mode and want to select by segments, you switch to segment and make your selection, then switch back to knot, open the tool and click Segment to get your desired selection.    It should be possible to convert sub-selections in spline objects the same way you conveniently convert sub-selections in Editable Poly, by making the selection in the level you want to convert from and hold CTRL while clicking on the sub-level you wish to convert to.      There is no technical limitation why it shouldn't be possible to convert from selected knots to segments/splines, as you can think of it as converting selected vertices to connected edges. If the knot is selected, convert that selection to connected segments.   Addition: In Editable Poly there are shortcuts that convert sub-selections you can assign to hotkeys and menus. Splines should have similar context-sensitive shortcuts for quick selection conversion. Show More

A lot of 3ds max users who work with spline profiles expect Fillet operations to yield a fillet that corresponds to a segment of a circle with a radius of the specified value. A fillet value of 25cm should always produce a curve that fits a 25cm radius circle, as long as the original spline allows it that is.   The current way 3ds max does fillets is to affect the size of the fillet based on the angle. It means only 90° angles will give you the correct fillet radius. Obtuse angles gives you too large, and acute angles gives you too small.   The image below shows the current result on the left, and on the right how most users expect the fillet to behave: Most engineering drawings define fillets this way, so it makes very much sense to have 3ds max behave the same way. It's much more practical and intuitive.   What I suggest is that such new fillet behaviour becomes an optional setting, in case users want to use the classic mode.   The calculation required to correct the fillet is fairly simple, assuming we need to adjust the current input value: (fillet value) / (tan (knot angle/2) So "fillet value" is your desired radius, and the "knot angle" is the angle between the selected knot and its adjacent knots, or the positions a very small distance away from the knot along the spline (which would support non-linear knot types).    Until this is implemented, users who wish to convert fillets to always provide a radius segment can evaluate the following expression in the MAXScript Editor, it will automatically put the converted value into your clipboard for pasting as the new fillet value. Just change the desired radius and knot angle: setclipboardText (units.formatValue ((units.decodeValue "25cm") / (tan (45.0/2)))) This expression works for Chamfer values too! Show More

Compound nodes are great for compacting complex material trees into one single node.   We can go from this  To this But there are areas where compound nodes can be made more useful.   • One limitation is that assigning materials outputs to objects requires dragging a specific output from the Slate and release it on top of a single object you wish to apply the material to. Also, all shortcuts associated with assigning materials from regular material nodes are disabled for compound nodes.  This means you cannot assign materials from compound nodes to multiple objects at once and the usefulness of these nodes is significantly reduced. I just realized that with multiple selected objects you get an option if you want to assign to all the selected objects. That's a solution to assigning to multiple, but I still feel not having to drag and drop is very beneficial too. I feel this feature is of course better than nothing, but it's not so intuitive to figure out. Therefore having the ability to the suggested solution below would be beneficial:   A solution to this is to have the system check if there are actual material outputs in the selected compound node, and if so enable Assign To Selected in toolbar buttons, menu options and shortcuts. If the user runs the shortcut, because there can be multiple material outputs there should be a popup menu where you pick which material output to assign to your selected objects, such as this popup menu below:     Additionally, right-clicking on a material output should give you the option to Assign To Selected, because then it's obvious which output you are interacting with.   This can improve the usability of compound nodes greatly. Show More

It would be grate to have same bihavior for quick slice as fore spline creation with pressed Shift key. Show More

Ability to import multiple files, like STP, STL, OBJ or FBX at once. A checkbox that automatically renames the imported model in 3Ds Max to the name of the file. It's insane that we need scripts for such a basic functionality in one of the worlds oldest and most used 3D software. Show More

The Spline system inside 3ds max is very good, but there are two very useful and fundamental part missing, which can be found in other 3d software: A way to control Twist and Radius/Scale per spline knot and not only allow for non-linear thickness and twist at object base level, but use such data to drive geometry generated by modifiers like Sweep and Path Deform.   Similar to how Poly Meshes have vertex/edge data channels which can hold data such as Weight, Crease, etc that can be passed on to other modifiers, spline objects should have channels for Twist and Radius/Scale data.   Current situation: Let's say we have a geometry which we want to use with Path Deform. Any change to twist and scale has to be done via curves representing change over the whole spline, a solution which is imprecise and tedious to work with for this reason. Suggested solution in detail: Similar to Editable Poly, when in Knots (or Vertex) sub-level for splines, selecting one or several knots lets you set values for Twist and Radius/Scale.   Twist could be angle away from what is the "up" direction according to the scene node of the spline, or the "up" direction relative to the previous knot. I acknowledge that compromises might be needed for certain situations to calculate the correct twist angle, for example if you use the scene node Z axis as the "up" vector and the spline is perfectly parallell to that axis, that will produce a highly unstable twist angle. Perhaps the user can select which scene node axis to use as "up" vector? If only the first knot is dependent on the scene node axis, all other knots could potentially rely on the previous knot twist angle for calculations, or if an array of "up" vectors are calculated for each step along the spline segments. I am sure it can be solved somehow since other 3d softwares have implemented this already. Radius/Scale could either be a radius value away from the knot position, which could be a in world units, or a scale value which is a multiplier of the spline's general Thickness value. Since a modifier such as Path Deform works with arbitrarily sized geometry, the concept of "radius" might make less sense there as opposed to scale. With that in mind, perhaps the simplest solution is to only have a Scale value and let it be a multiplier of the general spline Thickness. One could easily just use a general spline Thickness of 1 world unit, and the Scale value would become world units that way. A visualization of these data channels could be displayed as follows: Each knot will have a circle with an arrow perpendicular to the tangent of the spline path. The radius of the circle shows the radius/scale of the knot. Inside the circle an arrow shows the direction of the twist angle. A configurable number of "spines" on the segments between the knots shows the interpolation of the twist and radius/scale between them.     It should of course be possible to toggle the visibility of these visual feedback elements, perhaps even individually so that you can view the radius/scale only, without the angle arrow or the "splines" showing, if you're only interested in the radius/scale for example.   Some unclear aspects of this idea: How would be the best way to control the interpolation of twist and radius/scale between the knots? A linear interpolation is the simplest, doesn't require any additional parameters or interfaces, but gives the least artistic control. Would a bias parameter that pushes the "midpoint" of the interpolation towards or away from either point be useful? Would that be something you control by selecting knots (meaning the selected knot will control the bias in the segment ahead of it), or would you select a segment and control bias there? Curves provide the most flexible solution but might be cumbersome if a lot of knots need to be changed. I imagine one curve representing one segment, where a linear curve between 0-1 means twist and radius/scale is linearly interpolated between the two knots. Quick ways to manipulate many curves at once would be highly useful. Again, perhaps this is something that would be more intuitive to control by selecting segments rather than knots since what we are changing is what is happening between two knots. Should twist and radius/scale have independent set of curves so that you can control the interpolation of the two independently? It gives more creative freedom, but means more things to consider and control. Since the twist value is one-dimensional you only need one parameter to change it, which is why I think it would be more intuitive rotating the twist angle by a spinner in the command panel rather than rotating it with the transform gizmo in the viewport. Nevertheless, with this new data channels, not only could modifiers dealing with spline data greatly benefit and open up new ways of modelling, but it can also be used by already great simulation systems like tyFlow, likely leading to better artistic precision relating to angle and size around simulation elements driven by splines.     Show More

Hello, I design 3D Architectural Renderings. I'm looking for some feedback and advice on how to take my renderings to the next level. I've been using 3ds max for close to 5 years. I use Vray with it to render. My modelling seems to be good. I've been trying to get better at lighting, shadows, and adding more people and objects to my models.    -Does anyone favor a rendering engine over vray? I've looked into corona. Not sure what others may be superior. -Based on my attached images how could I take these to the next level? Any links or resources would help. - and finally, what is the best source for buying quality add in 3d blocks? People, Trees/Plants, Cars, Furniture, etc.   Any feedback would be greatly appreciated! I love 3DS Max and my skills have improved with each use but I intend to achieve higher quality in photo-realism Show More

Can you add MOV and MP4 support to the Bitmap texture? Its weird that it is not available. So now I have to additionally convert my videos into file sequences. Show More

Lots of things in 3ds max are obsolete and double. Leading users to confusions and making max bigger and messy.   1. Why do we keep the Video Post for? Its so obsolete. Last time I used it was in over 22 years ago.   2. We have the new PBR material.  Great! So right now we have like 4 pbr materials. Physical Material, metal and glossy pbr materials. Can we just focus on one? Its so confusing.   3. I notice that there is in Material editor a library called Autodesk material or something. I think it was working back then when max had mental ray - you cant even render that materials out.   4. In 2025.3 you moved two menu options to other places. Thats ok but why you leaved the old tabs in the old places.  I guess only for information whete to find it but again it makes 3ds max Messy and duplicates the things. I think we could temove that tabs and infos in new release.   5. Why the Art Renderer is still available? XD come on its soo obsolete.    What other things you find that are: - old/obsolete - doubled - messed up - ...   Show More

• The problem: Aligning array elements to meshes can be difficult. The current (and only) projection mode when distributing with splines paths only assume a projection direction according to the Z axis of the spline node. Relatively simple situations like the one below becomes very difficult to align the individual array elements to the normals of the surface:    Almost all array elements misses the surface, and the ones that hits it ends up in the wrong place. The current solution only works for surfaces that are close to flat. The Conform modifier isn't really useful to solve this situation either.    • The Proposed Solution: For stitch-like details like you usually want to use topology-independent paths like spline nodes to guide the stiches. The spline paths are usually placed near the mesh surface where you want the array elements to sit.   What is needed is a projection mode where the closest point on the picked surface is identified by each array element, positioned there, with the option to align an axis of each array element to the surface normal. Example using a custom MAXscript solution.   This would allow for procedural placement of details like stiches, rivets, etc on rounded surfaces that essentially loops around and you are not limited to mainly-flat surfaces.   Show More

We need high quality and advanced tools for sculpting. Tool that are available now are not enough - they are very basic. We have be able to create and sculpt high density models as Blender can. Along with the texture painting. Right now the current canvas paint work to slow. Show More

Hey Autodesk 3ds Max Team,   It's been a long time since the biped features have not been upgraded. Could you please consider the following point for animators to have these feature?   Could you upgrade biped rig features to include squash and stretch capabilities for legs, heads, and spines, without causing skew issues in game engines?   We are very limited to for this feature to make our character animations looks best for games. Show More

We need more viewport / quicksilver improcements. Implementing RTX features, blurry reflections, reflections between the objects, global illumination, reflecting the lights shapes not only as point lights, etc. Show More

There's always a phase of developing a shot where I am moving the camera target and adjusting the camera parameters - I have to keep switching back and forth between the two and selecting and deselecting. It's especially annoying because this is the phase where I like to lean back from my desk, put my feet up, and take time to consider how shots look. I see no reason why the camera parameters cannot be edited while the target object is selected.   Being able to edit the camera parameters while the camera's target is selected would make life easier. Show More

I suggest the following:   1. When copying maps and materials, it should behave more similarly to when copying objects. That it doesn't resat all naming to map #1927813674263, but instead has the same name or appends a 01, 02 etc. So shift-dragging a map named "BaseCol Mix" would be copied as "BaseCol Mix 01".   2. When drag-n-dropping bitmaps on SME, it should name the map the filename of the bitmap. So instead of "Map #8945734589734589" it gets named "Wood_Normal.png" for example.   Quality of life, 4 lyfe! Please make it happen! 😄 Show More

Just like there is a modifier that dynamically welds vertices on geometry objects there should be a modifier that welds spline knots the same way. This will allow for more procedural non-destructive workflows when creating complex spline shapes, for example when using the Array modifier on a spline and want to keep each clone connected. Without welding (note the visible seams from disconnected edges)   With knots welded     Show More

The Bridge operation in Editable Poly is quite basic, but it would be very helpful if this operation can be used to make more than just straight connections.   It is often the case that I want to Bridge two selected surfaces but I do not wish to to have a linear connection. This non-linear path can be controlled by a spline object, like in the mockup below:     The Extrude Along Spline operation will give you sort of a similar result, but the end does not connect or conform to the end surface, which requires a lot of manual work to make look right. To have this spline guide directly in Bridge would lead to much quicker and more accurate results.   Since the curvature of a path is used there can be an Adaptive Mode that inserts segments based on the curvature of the spline, for optimization.   I also have a suggestion for a new type of Bias when creating the Bridge. The current Bias setting only controls where the center of the tapering sits relative to the bridge. But i would like to see another Bias that controls the general segment distribution that can "pinch" segment distribution closer or away from the center. The benefit of this is that it is often the case that you need more segments near the connections of your bridge, and by pushing segments closer to these connections provides better segment distribution for things like relaxing the borders of the Bridge. Mockups below:     An improvement that could be implemented, preferably alongside the ability to use spline paths, would be automatic curvature of the bridge based on the face normals and distance of the extruded surfaces. A lot of other 3d modelling softwares has it, so why not 3ds max? Show More

Based on this thread, CAT as it currently stands won't be receiving any new major updates or new features. This is very unfortunate because CAT is a great concept and it's extremely powerful and there's nothing else really like it on the market. There are some 3rd party tools for 3ds Max such as InstantRig and LH Auto-Rig that replicate the modular rigging of CAT, but they don't have the extensive animation tools. Likewise, most other 3d packages don't really have an equivalent--Cinema 4d has CMotion but that doesn't have all the other great features like foot placement, motion extraction node, etc.   The problem with CAT is that it's a bit outdated. It's buggy in places and can cause crashes. Some things are counter intuitive--for example, Motion Extraction Node is a broken mess unless you select the pelvis and enable "Additive to Setup Pose" in the Link Info, then it magically works for the most part. The interface is dated and clunky. CAT could use additions like an equivalent to Biped's Motion Mixer--manually blending animation layers in CAT is well and good but it's not the same as a timeline editor. CAT can also be slow when working with multiple characters.   So my idea is, if it isn't practical to give current CAT a major overhaul, then build a new version of it from the ground up. Completely new codebase. Something that basically does the exact same thing as CAT and has similar features (CAT motion layer, adjustment layers, motion extraction, etc) but does so with a fresh, clean, and modern codebase that can be maintained and regularly updated. Something that can have a slick new interface that meets the needs of modern animators. Something that can take the lessons learned from CAT and Biped and expand on them and improve them. Something that can be designed from the ground up for modern hardware and make use of multi-core processing or even GPU processing. One benefit is that this could be a way to unify the character animation systems of 3ds Max and streamline the workflow and interface. Right now it's disjointed, fragmented. First you have the regular bone system which has a bunch of tools. Then you have Biped which has its own set of tools. And then you have CAT, which also has its own set of tools. But then they sometimes share features/concepts using their own implementations (animation layers vs biped layers vs CAT layers). This makes animating in 3ds Max a bit rough or confusing for newer and younger users since there's so many systems to learn, each with their own pros and cons.   Another benefit is that this could provide a framework to more easily add other modern features later down the line in future updates. For example, maybe the new tool could get built-in lip sync in a future update. Something similar to FaceFX? What about something like Clovis Gay's F.A.S.T. plugin for 3ds Max? Maybe the tool could get a node-based scripting system like Houdini to allow for procedural animations beyond CAT motion layer. Maybe a built in crowd system? Maybe the ability to script AI so animated characters can "react" to stuff without the need for manual keyframing? What about a timeline editor like the one I suggested here? I'm just spitballing of course, but you get the idea--stuff like this could potentially be added in future updates, which may or may not be practical with current legacy CAT.   So to summarize, make a brand new rigging/animation tool, CAT 2.0 or some other name, built from the ground up to take all the good stuff and lessons learned from CAT, Biped, and bone system, and which can be updated regularly. One centralized system to rule them all, which can be the sole focus for future development.  Show More

It is good that we in 2024 can finally control the Viewport Clipping values parametrically via MAXScript. Handling Viewport Clipping in previous versions have been one of the biggest annoyances inside 3ds max.   However, not all users know that this is possible to change with actual values, or where to find scripts that provide an interface.  I suggest that parametric control for the near and far Clipping Planes should be added to the Active Viewport Settings so that any user can find and control this in the most logical place to look without the need for external solutions.   And now that we can control this with using actual numbers we do not need to rely on that slider gizmo thing on the view of the viewport, so please let the user toggle whether this gizmo should be visible in the viewport while Viewport Clipping is enabled. And please expose this to MAXScript so it can be toggled by custom tools. It is often the case that when you need to make screenshots of the viewport you need to have Viewport Clipping enabled to minimize z-fighting, and the gizmo will always be there as a distraction. Let users be able to choose if they want to include it or not.     Show More