Feature Requests

the node input/output labels(that appear when hovering a node) color is the same color as node connections. could you change the node labels color to black so it can be seen when a label is directly over a connection? regards,phil

could you add a method so we can obtain a child comment's graph position? i have this plugin that when placing entities(nodes/objects) in scene, it first checks if an entity is selected, if so, it offsets the new entity from that. however when it comes to offsetting something from a selected comment thats a child of a node, then the newly placed entity is badly offset due to the child comment's pos being based off the nodes coordinate system rather than the graphs. regards,phil

would be neat to have a hotkey to disable frames that are behind entities from actually moving said ents when frame is being moved. so the hotkey essentially temporarily disables the frame's "stickiness". regards,phil

basically, when we selected a portal node, the dashed wireless connection becomes visible, i know why this is and its a nice feature. but any chance of adding a btn to the permanent graph toolbar where we can toggle the wireless connections from appearing cause when graphs are large and portal nodes in graph is many, then wanting to shuffle things about, the vast wireless connections being on-view is quite obstructive. regards,phil

I have a plugin that utilises stock instance nodes. but at some point it would end up adding old graphs(nodes) to scene if we arent ever made aware that stock instance packages have updated as the plugin requires users to update the stored graph name as and when it gets updated by you guys. So if you could make this info publicly available that'll be great! regards,phil

just noticed that the graph object(s) classes dont have this method: .getIdentifier() like the SDCompNode class does. This method is great cause it returns a unique set of nums(id) which is handy for when needing to compare entities. regards,phil

Ive noticed that in the uiMgr class, only methods exist for obtaining graph entities(nodes &/ objects) if theyre selected.It'll be great to have a method that simply returns all graph entities whether selected or not. Regard,Phil

please consider adding this stuff to the api, thanks```All of this i know is pretty much GUI stuff. But its deffo cool stuff and would give us a lot more variety when letting the creative juices run wild.Im sure someone else would probs like other things ive not mentioned/not generally related to GUI.  Some things i think we should be able to do: 1) subclass nodes, graph scenewhy?so we can access the following methods:mouseMoveEventMousePressEventMouseReleaseEventetc..or at least provide a callback for this. 2) map mouse cursor to scene(graph) posWhen i have a node follow the mouse cursor, theyre not aligned.Its due to mouse cursor going off monitor coordinate system and node going off graph coordinate system so need a method like mapCursorToScenePos() which handles the offset conversion for us. 3) change node outline color,   mark a node as selected so the outline actually appears,   change a connections color / thickness.   chang

So when we have a few missing  images we could select  their folder only once.  Not doing it for each and every missing resources.  Perhaps as an option.  

I've got a number of images that share the exact same transforms. Using the transform widget in 2d view for my scenario is the fastest way to position and scale multiple images. Instead of manually updating each one I want to instance their values. I setup an exposed parameter then duplicated the transform nodes. However, the widget is no longer useable even if I click on "Edit in 2d view".    I couldn't figure out a way to expose outputs from a Transform 2d so they can be daisy chained using custom inputs, so I assume exposing to the main graph is the only method to instance a transform 2d? Below is my instinct on how this should be possible.    

I really want an input node in my custom graphs that acts as the input node on the levels and blend nodes. A node that dynamically detects the input type and then outputs grayscale and/or color if needed.It would save me a lot of hassle when building custom nodes if I don't need to make them using a boolean to toggle input and output visibility. I just want run everything though the same graph, independent of input type without using conversion nodes.

When adding a new node, you have to type long strings for longer node names, or scroll through many options if you don't want to type long names. What if you only had to type part of each word of the node's name? For example, if I want to add a "Multi Directional Warp Grayscale" node, it would be much faster to type "mul dir gray". Other node-based softwares do this. It means less typing and less scrolling. Adding nodes feels slow and distracting when you have to sort through options every time. There are many nodes that start with "gradient" - so either I type the full word "gradient" followed by another word, or I have to read through all the nodes that start with "gradient" until I find the one I was looking for. Or I can type "gra lin" and press enter to add the "Gradient Linear 1" node in less than a second. This would also be helpful if you don't remember what a node is called, or you don't remember what order the words are in. Right now, the search doesn't show a

Is there a way to reorder the input list for the new portal node? I tried labeling them with number e.g. "01 Base Height", "02 Height w Cracks" but they don't sort alphabetically for me.

Similar to a new feature introduced in Unreal Engine 5's material graphs, reroute nodes that lets you plug an input into a node, and then grab that as an output elsewhere in the graph without the need to have tons of connecting lines going all over the place.So instead of this:You'd get something like this: Obviously it's a limited example, but it'll let you produce much more readable and clean graphs, and can help with optimisation as you could grab and use a single noise rather than reusing a new Clouds generator each and every time, while keeping everything tidy and organised. Bonus points if the nodes were visually distinct, and maybe didn't have the thumbnail to save computation time, as all the nodes need to do is point to where data should go like a connector.

I know there are aliases   but wouldn't it be much more convenient to just link and re-link resource folder ?   Not each bitmap one by one but just folder.   Same way evry 3d package does. Those aliases make me problems.  I have a huge list of them .   If I want to  do just new texture  version  using  new set of bitmaps  exported  from another sbs  I have to relink the alias  and  figure out which one  exactly.   It turns to the total mess so quickly.      Please do it standard way  where we could just re-linlk the folder. 

Wish we can add a node that can cache previously computed result so that the image caches before this node is not stored in memory. I know the new version of SD has node image cache option on the menu, but many times I found its really the best to cache some nodes while others are not. Please make this come true thank you!

I know I am supposed to use Painter.  But gosh how  inconvenient  and  pain in someone  ... it is  comparing  it to Designer nodes.     Always non stop troubles with some hidden checkboxes   and gazillion sliders I have no understanding what they really do math wise and no way to see or quickly  modify it.  Or why they work odd when in Designer evrything is just fine.     It's miles  easier when you do your own "filters"  and understand what's actually happening .  With Painter I wasting days and weeks  debugging  things.   Also anything I do in Painter  is a tad bit blurrier . Had to downscale at export .  While in Designer  I use my own transform node  that works in pixel precision  . i.e moves something exactly  5 pix left for example. Not 5,345 that instantly adds more blur.   Tried to do a same  "filter" for Painter  and it neve

In the pixel processor, I would like to be able to store the result of a part of a function in an array/temporary image (up to the same size as image size) and have that stored array/temporary image available for sampling, modifying and re-storing in subsequent runs of the processor (i.e. for every pixel processed).   This would increase the power of the processor, for example to build a mask as each pixel is processed and use and modify that mask when the next pixel is processed.   If the final array(s) could be output alongside the normal pixel processor image output - even better   Dave  

The Adobe Substance products empowers 3D design and allow you to experience intuitive workflows, to design, connect, import, and view Substance materials on your favorite visualization applications.   To continue improving your experience and to tailor our products to your needs, we have created this survey.   Please note that the survey is anonymous and completely optional. We are grateful for your answers! This will guide us in creating a better experience for you.    

I would like the ability to draw inside a mask layer and use that as a bitmap within the node graph. So the painted mask can impact the designer graph. For example I have created a vein material that uses a single mask and goes through a scatter node to add small branching veins. I would like the ability to paint the original vein mask in substance painter on an object and have that automatically get the branching veins and fit within the material made within the graph.

A one compatible with modern renderers. So we could import and composite procedural and rendered elsewhere things together. For complex things scattering/population it’s actually easier and speedier in something like Redshift or Octane. With better immediate visual feedback, better AA. Proper intersecting physics. Much easier to do erosion effects etc. That color ID node is a pain to use comparing to proper cryptomatte. Especially for small few pixels details. 

Hello, I am new to Substance Designer, I usually use Mari software for texturing, and I was exploring different ways to create Masks/Gradients and to control them in SD.I noticed that there is an option to create a gradient using 3D linear gradientand using worldposition to orient/manipulate it,I was wondering if there is a way to reproduce sort of 3D Radial gradient, based on an empty object location as gradient centre, and controlling the radius, falloff and feather from that point forward?If not, I think the software would benefit from such a node.

For U and V independently .  Or just a scale gizmo  for a test object.  It's so annoying to make and load  new objects for each of texture output factor .And please make zero height scale a default material  setting.  It's so tiresome to scrol through material settings to find it evry time. 

I mean to share some parameter  you have  to expose it  and onse you did it  the parameter it's locked . You have to click on background , look through parameters, change the default settings ,  click , click, click etc It's so monstrously inconvenient and tiresome .  Especially if it's levels  or  transform nodde whree graphical representation is always better than just sliders. Please let  it be editable,  or do something color codded  Blender does with drivers and driven, or let us have a widget sliders on top of  the main node pane .I know I can make value inputs and  connect nodes by means of this   but it's so much extra pain in my a.. too.       Just  make exposed  parameters editable  please.  

I'd like to be able to right-click on a node/graph/subgraph and do "Open Dependencies". Today we can do that from the package itself, but this sometimes opens many more dependencies than we need.