Topics by tag:
Recently published articles by Dave G:
So, after finishing up some jobs and adding some features to these while working on those contracts, I have rolled in some new things to both V-Ray Tuner and Facer. There’s quite a bit that’s added in each and I’ve included some features of one script in the other just because it made sense for things like UVs, which affect both modelling and rendering.V-Ray Tuner 3.0 new features:
As you can see from the way I start a linear workflow scene from scratch, this lets you click the LWFMe button to initialize all the right parameters for a linear workflow but then you disable the LWF toggle to work with manually linearized swatches. The Gamma Correct me script for swatches currently only works on V-Ray Materials (not FastSSS2, Light Mats, etc). I’ll add support for those later.
Often when you work, you’re lazy about using instances to save memory so this lets you just do it all in one step for big memory savings at render time. When you render a bunch of those heads with copies, it takes over 6GB of memory (click for high res):
Use the script to replace those copies with instances and your render memory usage drops to 1.6GB (click for high res):
So, I hope you enjoy these updated scripts – the download links are in the new feature headers for both. If you get a lot of use out of them, maybe consider picking up a copy of the ePUB, Kindle, iPad iBook or PDF version of 101 Autodesk Maya Tips, which is still only $2.99. In the words of Yakov Smirnoff – IS BARGAIN!
2 notes | Permalink
A while ago, I wrote a tip for sharing assets between Mac OS X and Linux, by mimicking paths with symlinks on slave machines. It seems that workflow is no longer needed and, what’s better, you can now use Windows machines as slaves without having to worry about C: drive paths screwing up everything. The latest nightly builds of V-Ray for Maya let you transfer missing assets without worrying about network shares. Here’s my MacBook Pro getting a boost from my Windows 7 gaming rig:
I’ve tested it and it also works with Linux and I’m guessing it works with any combination of platform as slave/master.
1 note | Permalink
I just finished up a project that involved making an intravenous bag filled with blood and I thought I’d share the workflow for how to go about modelling something like this, since it can be tricky. If you’re rendering a liquid inside of a transparent object, there’s no easy way to fake the thin wall and liquid – they have to be modelled so that they render correctly. While I won’t go into the specifics of rendering these shapes, there is a good tutorial here for V-Ray and an explanation of how dielectric interfaces work in V-Ray, which is the crux of accurate refractions between transparent objects.
So here is my first talkie tutorial movie for you to enjoy. But, in the spirit of silent films, my girlfriend is playing background piano to liven it up. She’d be horrified if she knew her practicing was being recorded.
As you can tell from the video, I have a pretty specific set of tools for my modelling workflow and most are gathered in my Facer utility. You can do this without it but you might want to install Facer before starting since it makes following along a lot easier, and it’s a very handy toolbox, if I can pat myself on the back like that. Other time-saving MEL scripts used include Zentools’ Flood Select and alignVertices.
3 notes | Permalink
Version 2.7.1 of V-Ray Tuner is up and has a few new features and a bug fix that might be affecting certain Windows users:
V-Ray Tuner’s reached 4000+ downloads and counting and it’s got 4.5 stars as a rating on Creative Crash. Glad people people are enjoying it and find it useful.
2 notes | Permalink
V-Ray Tuner has a “sleep at end of render” script that works on all three platforms but I realized recently that most Linux distros require that you run pm-suspend as root, nullifying the script. This can be addressed by letting users run pm-suspend without sudo. Just use visudo to add these lines to your user’s sudo permissions to prevent a prompt for password when using sudo:username ALL=NOPASSWD: /usr/sbin/pm-suspend
Until I post an update, Linux V-Ray Tuner users will have to find and replace “pm-suspend” with “sudo pm-suspend” in the V-Ray Tuner MEL script.
If you have been following V-Ray Tuner for a while, you’ll know I use the per-light render feature quite a bit to make flexible composites that can be rendered once and adjusted in post. While working on my last job for Reader’s Digest, I added per-light render support for the self-illumination values of V-Ray Materials. The self-illumination value in the V-Ray Material can be an uncorrected colour swatch, a gamma correct with anything plugged into it or it can be a file texture. I made sure that this script works fine with a mix of meshes with V-Ray Light Materials and meshes with V-Ray Materials with self-illumination:
It took some work to set that up but it’s bulletproof from what I can tell – it’s only noisy because of the quick render settings I used. I’d like to eventually add the exposure tweaking feature to material-based lights but that will have to wait. The workflow is straightforward: select your meshes that have either type of material, run PerLightRender and you’ve got a very flexible set of images for compositing and your scene is left untouched. Keep in mind that you should be working in a linear workflow for those to add correctly in Nuke or Photoshop.
The other updated feature of V-Ray Tuner 2.7 is to the RT Multi-frame Render script, that lets you use V-Ray RT as an animation renderer for quickly previewing animation renders. I’ve added an option to use the Max Paths Per Pixel option to determine the max render time. This will give you more consistent quality across frames vs. using the RT Max Time Parameter:
There is now an “RT max paths per pixel” slider in the V-Ray RT Settings foldout of V-Ray Tuner so this is easily tweaked and you can change that at any time during the animation render.
Other small changes:
Here’s the V-Ray Tuner download link.
All the Mac/Linux/Windows versions of V-Ray 2.3 are on the Chaos Group support server if you’ve been waiting on this to move to Maya 2013.
If you’re running my Automator action for nightly unrar and installs, open the workflow file and change 2012 to 2013 in the file path and you’re good to go.
I made this Automator service a while ago for making V-Ray Proxies and I added a prompt for the face count.
I think a lot of people will like the new render utility preview added in V-Ray Tuner 2.6.2:
When you build complex shaders, it becomes increasingly difficult to tell what a stage looks like by looking at the end result, so that’s why I made the render utility preview. I recommend using the non-OpenCL RT for this feature since RT GPU has very limited support for shader network types. The facing ratio used above is one that doesn’t work in OpenCL mode.
The render utility preview works with pretty much anything you throw at it (utility or procedural texture) if V-Ray RT supports it. It just maps the utility to a surface shader, which has no shading properties so it shows you just what the utility looks like, unshaded. As usual, all temporary surface shaders, render layers and objects are removed on hitting the Cleanup Material Elements button. This won’t make garbage in your scene that you have to manually remove.
Other changes in 2.6.2:
After manually switching the V-Ray for Maya distributed rendering settings a bunch, I found out that these settings are stored in plain text files that can be changed separately. On OS X, there are two files. The first:
That has the slave addresses (10.0.1.x, for example). I’m not sure where it is stored in Windows or Linux but look in the user level maya folder and it should be there. And the second file sets the slave status (Enabled/Disabled):
All you do is toggle the Disabled/Enabled text for the corresponding IP address:
And you’ve just set the slaves without opening Maya.