Haha...nice one! I could guess people see first what MACHIN3 did then i took inspiration and did yesterday my own version and now it is in Modo too.
The only limitation i found is that boolean base and cutters must have around the same size of faces/edges density otherwise if the base is high poly and cutting object low poly it will give unexpected result if it moves all the edges. In that case, i did an update with a threshold so that it doesn't merge all the edges but, if you want, just the closest ones within certain threshold depending on the edge lengths.
Nice. Since it seems like you guys have already covered boolean cleanup for Blender, Modo, and Maya, I went ahead and scripted it for Max.
@Klaudio2U Your gif is 8mb, that's crazy. I don't know how you're recording, but maybe consider something like LICEcap to record instead, or some sort of gif optimizer.
I'm also not tech artist and I've been doing lots of scripts but using MEL in Maya which is nice some simple things. I do know some very basic Python too and lately I've been looking into switching for good from MEL to Python so i was guessing that you guys are using Python but what i was curious about is how much in your scripts is "pure" Python and how much is using software specific built-in commands to do certain things. For now, there are few tools i would like to do like some selections tools what you William are doing but then also some other like MACHIN3 chamfer/boolean cleanup similar scripts shared here on the previous page.
The way you interface with the 3d app is what matters. Python is just the language you speak, literally. Operators in blender are classes formated in a special way with certain functions defined(execute, invoke, modal, draw, etc, depending on what you wan to do).
In blender you can script in a way that just calls existing Blender tools, that's also what the Blender UI does. It's a great way to get started. A lot of DECALmachine is doing that.
But then you can also go deeper and work directly on the data blocks, do vector and matrix math for geometry manipulation. MESHmachine does a lot of that. It also turns out this is much easier than I ever thought it would be), in large parts due to the very excellent mathutils library made for Bender but usable outside it as well AFAIK.
In terms of python it's just a lot of loops and list comprehensions for the most part. and figuring out good ways to store and access your data so you can work with it. Lists of dictionaries are often handy.
The way you interface with the 3d app is what matters. Python is just the language you speak, literally. Operators in blender are classes formated in a special way with certain functions defined(execute, invoke, modal, draw, etc, depending on what you wan to do). ..
Thanks for the info! Long story short, basically, Python is just a way to go because it is way more robust do work with then, in my case MEL. Another good thing is that with Python there is always that possibility to directly access Maya API + libraries, modules, etc... all this is not possible with MEL so it often comes to some big workarounds to do things which then additionally slow down performance which was already slow, to begin with.
especially once you hit more general purpose usecases like .. calling other processes, handling files, processing data like xml/json, network communication etc. things will be indefinitely smoother with python - because it has so much included, and there exist at ton of libs for pretty much anything.
No Boolean feature from my side, sorry (great job guys, really inspiring to see what you are doing!).
I am slowly wrapping up the next version of my ZBrush plugin and wanted to show a synergy I think is quite handy: You will be able to choose a color either via the (customizable) color palette, by using the foreground color, or by a random generated one and apply them to a subtool by hovering over it and pressing of a hotkey.
CRYENGINE has no way to make text in game in the editor so I made a hacky little system using the python api they have in Sandbox.
Works reasonably well.
It's made using single designer objects mapped to a font face texture. So at the end you can merge the designer objects and export them as a cgf which will cut down on the number of cgf's and drawcalls.
Hello! I am a beginner looking for some critique on my pyro. Simulation in houdini rendered in mantra. Getting indie this weekend to render at full resolution still messing with my gas forces and seeing what gives me the best quality. I am not set on my current shader because I want to try redshift I see it gives much nicer results.
I have been working on smoothing algorithms for meshes in blender. You can define sharp edges as rails that will be used by verts to slide on. Volume is more or less preserved.
Trying to create modular buildings generator for Unity
Looks incredible!
---
I've been working on a tool to turn mirror mods into
real geometry, with proper origins and orientations and proper custom
normals.
I consider this the final missing piece in my custom normal pipeline.
I've got creation and symmetrize covered. Now mirror is done as well.
This tool is also essential if you are doing object space texturing, as
the mirror mod in blender messes with this as well (understandably) as shown in the second part of
the video.
I have been working on smoothing algorithms for meshes in blender. You can define sharp edges as rails that will be used by verts to slide on. Volume is more or less preserved.
Magical, I've been looking for a good way to relax topology forever.
Hi, cool script. I tried to follow you on gumroad so I can purchase this later when I need it, but you never filled out your page so I can't follow you on gumroad.
I've been working on a Modo plugin for speeding up the process of baking and exporting to Painter. Also has support from quality presets and bake morphs (for exploding things).
Porting Stephen Hill's and K. Narkowicz's ACES approximations over to Toolbag. I'm very new to writing code and it's been a fun challenge. I have the color space conversions working, but not exactly sure where they fit around the ODT so it might not be 100% perfect yet.
I created a small script to prepare my meshes in 3ds Max for 'match by name' baking in substance. Have been using it for quite some time now and decided to make it available for everyone who is interested in having a look at it.
I'm determined to find a way to make Active/Optical Camouflage as an opaque material setup - so something that would function on the horizon, doesn't quite hold up close-range. Got something like high-tech frosted glass so far :P
Hi, I'm working on some scripts for faster workflow in Substance Painter. I would like someone to test them first before I release them. If you are interested give me an e-mail at buliarca@yahoo.com This is a small screen grab from it: you can check this link to see more: https://gammy-balances.000webhostapp.com/qAct/help.html
more boolean clean up..... path creation tool using a spline to mark the path which is then cut into the mesh (still needs work, quads for the path and path aligned uv's in another map channel and a few spurious situations or places where my routines fail and an absolute pain to track down)
One week ago I found what my team spend a lot of time for reviewing
UV and padding. So I decided to wrote script for Maya for checking
padding. Main condition was - Maya LT, so I can use only MEL…
I solved the problem. In few words: calculate hypotenuse from Point C
(vertex) to A and B points (edge vertices) using minimum padding (small
cathetus CD) and compare (AD+DB) with real edge length (AB). And some
simple (or not
) conditions, exceptions, inspections… And a lot of optimization,
because at first time all worked great… but spent a a lot of time…
Also at begun I checked harden/soften edges+saved them in list, so at the end I assign them back.
small GIF presents result, Pinned vertexes have small padding to near UV shell:
The gif is too fast, is it scaling the shells up so there is only 8 pixels of padding between all of them? And if so that would be really helpful I think you should sell the script on the gumroad if you can get that working.
more boolean clean up..... path creation tool using a spline to mark the path which is then cut into the mesh (still needs work, quads for the path and path aligned uv's in another map channel and a few spurious situations or places where my routines fail and an absolute pain to track down)
this is looking awesome, would be a huge time saver for certain modeling techniques. I'm currently doing this manually via shapemerge and it's a p.i.t.a. scripting with shapemerge is impossible as well so i've been at a roadblock for a while. Does your method scale well with higher density meshes? how about multiple splines per mesh? (i.e 10 - 20) i know you're in the early stages with this but it looks very promising for organic hardsurface panel seams.
I just released a mini plugin for ZBrush, that helps to speed up Quicksaving time and Autosave stability by deleting all unselected Tools.
These get stored while saving as well and are a bit tedious to clean manually.
this is looking awesome, would be a huge time saver for certain modeling
techniques. I'm currently doing this manually via shapemerge and it's a
p.i.t.a. scripting with shapemerge is impossible as well so i've been
at a roadblock for a while. Does your method scale well with higher
density meshes? how about multiple splines per mesh? (i.e 10 - 20) i
know you're in the early stages with this but it looks very promising
for organic hardsurface panel seams.
turned into a dead end, it was based on some poly mesh boolean routines in the SDK which turn out too be far to flakey and unreliable and unpredictable to use in any kind of tool. The whole of editable poly in max is PITA bug ridden mess probably why there are so few quality tools doing this sort of thing.
this is looking awesome, would be a huge time saver for certain modeling
techniques. I'm currently doing this manually via shapemerge and it's a
p.i.t.a. scripting with shapemerge is impossible as well so i've been
at a roadblock for a while. Does your method scale well with higher
density meshes? how about multiple splines per mesh? (i.e 10 - 20) i
know you're in the early stages with this but it looks very promising
for organic hardsurface panel seams.
turned into a dead end, it was based on some poly mesh boolean routines in the SDK which turn out too be far to flakey and unreliable and unpredictable to use in any kind of tool. The whole of editable poly in max is PITA bug ridden mess probably why there are so few quality tools doing this sort of thing.
that's cool, i'm rooting for ya. I'm not a programmer, but i've always wondered if it's possible to write a class adjacent to editable poly that's geared more towards booleans and "next gen" modeling techniques seen in other packages. something that maybe has less functionally than editable poly, but is faster when i comes to a non-destructive workflow with compound objects. That's a monumental task tho I'm sure.
manage to get to a, I hope reliable, cutting/projection routine that can handle all the border conditions (after about 5 attempts). It uses a mesh "traverse" algorithm so only uses one mesh ray intersect so it's pretty quick (6-7 times faster than uses max's poly boolean which is always nice )
I had a revelation/epiphany about the clean up (it wasn't going well ! ) the whole process will be a tad slower but i've got frame rate to burn
damn this is looking awesome, look forward to seeing where it goes. do you think custom uv mapping would be an option? ie mapping follows spline and user can set the uv coords ( mapping region of tiling texture ). also let's see it on a mesh that's 10 000 polys
Replies
The only limitation i found is that boolean base and cutters must have around the same size of faces/edges density otherwise if the base is high poly and cutting object low poly it will give unexpected result if it moves all the edges.
In that case, i did an update with a threshold so that it doesn't merge all the edges but, if you want, just the closest ones within certain threshold depending on the edge lengths.
Since it seems like you guys have already covered boolean cleanup for Blender, Modo, and Maya, I went ahead and scripted it for Max.
@Klaudio2U Your gif is 8mb, that's crazy. I don't know how you're recording, but maybe consider something like LICEcap to record instead, or some sort of gif optimizer.
nice work guys!
Thanks for the info!
Long story short, basically, Python is just a way to go because it is way more robust do work with then, in my case MEL. Another good thing is that with Python there is always that possibility to directly access Maya API + libraries, modules, etc... all this is not possible with MEL so it often comes to some big workarounds to do things which then additionally slow down performance which was already slow, to begin with.
I am slowly wrapping up the next version of my ZBrush plugin and wanted to show a synergy I think is quite handy:
You will be able to choose a color either via the (customizable) color palette, by using the foreground color, or by a random generated one and apply them to a subtool by hovering over it and pressing of a hotkey.
Works reasonably well.
It's made using single designer objects mapped to a font face texture. So at the end you can merge the designer objects and export them as a cgf which will cut down on the number of cgf's and drawcalls.
FireWithDisturb from Katchawan MGD64 on Vimeo.
MOPS HOUDINI from Katchawan MGD64 on Vimeo.
Volume is more or less preserved.
"Search Overlap" for Maya
- Select all overlapped UV shells
- Shift them to the side (so they will not take part in the baking process)
- During this process symmetry will be also taken into account
Demo video:this plug-in is on:
https://gumroad.com/l/SearchOverlap
I never posted here before, but it's time to:
So currently on this one: Octopus for 3dsmax
Also has support from quality presets and bake morphs (for exploding things).
ACES LUT for Painter (crosspost)
Porting Stephen Hill's and K. Narkowicz's ACES approximations over to Toolbag. I'm very new to writing code and it's been a fun challenge. I have the color space conversions working, but not exactly sure where they fit around the ODT so it might not be 100% perfect yet.
EDIT:
Updated to make it much more like the intended material (thin metal fuselage):
I hope to record part 5 tomorrow evening.
This is a small screen grab from it:
you can check this link to see more:
https://gammy-balances.000webhostapp.com/qAct/help.html
Hi everyone.
One week ago I found what my team spend a lot of time for reviewing UV and padding. So I decided to wrote script for Maya for checking padding. Main condition was - Maya LT, so I can use only MEL…
I solved the problem. In few words: calculate hypotenuse from Point C (vertex) to A and B points (edge vertices) using minimum padding (small cathetus CD) and compare (AD+DB) with real edge length (AB). And some simple (or not
) conditions, exceptions, inspections… And a lot of optimization,
because at first time all worked great… but spent a a lot of time…
Also at begun I checked harden/soften edges+saved them in list, so at the end I assign them back.
small GIF presents result, Pinned vertexes have small padding to near UV shell:
I just released a mini plugin for ZBrush, that helps to speed up Quicksaving time and Autosave stability by deleting all unselected Tools. These get stored while saving as well and are a bit tedious to clean manually.
You can download it here for free: https://gum.co/HBLvy
Info on all the scripts
Teaser:
And they're available on my Gumroad and Cubebrush