|20th Mar 2018, 02:39 PM||#1|
LightWave - Brief Lightwave 2018 Review
With the reduced upgrade price of $295 set to expire at the end of this month, I’m sure there are a few Lightwave users sitting on the fence trying to decide whether to make the switch to LW2018. Using 2015 at work, I held off upgrading to 2015 on my home computer since the changes were minimal and I couldn’t justify the dual-upgrade cost from 11.6.3. for two upgrades during 2015's run. The $295 price was financially agreeable.
2018 has received the most significant changes since version 9 was released over ten years ago. I decided to upgrade my home computer first, and learn the software on my own time, rather than scratching my head at work where deadlines are cast in carbonite.
First, a disclaimer. The many 2018 changes would be hard to digest in a single post, so this will touch on some of the major aspects used on a daily basis, like documentation, modeling, lighting, and rendering.
Documentation has been thoroughly overhauled and Newtek did an excellent job. The manual is wiki-based for continuous editing. Hyperlinks and video examples abound, so for those of you who learn best by reading, you’ll appreciate the new methodology. It's still a work in progress, with a lot of topics to add as more feedback is received.
Modeling changes are minimal. As usual, Modeler received the least love in this update, but what small changes occurred, I like.
One of the new features I really appreciate is the Rotate tool incorporates a smaller angular method of interactively rotating an object using the RMB. Previously, I used the Numeric panel to set a movement of .1 degrees after establishing the pivot point and kept pressing the Apply button until I got the rotation just the way I wanted it. The new method is most welcome. There are a couple of other additions to Modeler to, but nothing major. A new window option allows the OpenGL lighting from the Layout camera to be seen within Modeler.
Here’s the show-stopper for many: If you aren’t a fan of Nodes, you might not be enthusiastic about upgrading. There’s no getting around it, nodes are only getting more integrated into Lightwave’s workflow, and it’s causing a lot of headaches for many seasoned artists who aren’t used to them. Even lights are using nodes now. Nodes aren’t mandatory to using 2018, but you can’t take advantage of the latest version without investing the time to understanding how they affect different aspects of the output renders.
A thorough understanding of material properties is required for texturing in 2018. Gone is Transparency; its called Transmission now. Reflection is history too – replaced by Specularity. Gloss is gone as well as Translucence. There are a few preset nodes, and I expect many more will become available. At work, I rely on a lot of carbon fiber and machined metal textures when rendering hardware: Aluminum for the minor components, up to stainless and titanium, with different anodized coatings. Using the old texturing system, I had a bunch of pre-made textures to choose from and it took moments to texture something based on its orientation and size. Now I have to rebuild a library of node-based textures using many interconnects to stay current. It will be worth it when it’s done, but it will take time to get it looking as good, (and hopefully better) than what I use now.
Lighting is one of the hardest skills to master and makes or breaks a render, so there’s no way around having to become proficient at it. 2018 totally revamps lighting and I feel like I’m starting over in some ways, but it’s immediately apparent how much better the new lighting system is once the new settings become more familiar. Being physically-based, the lighting system uses new terminology that reflects industry-wide standards to describe former LW descriptions. No more tweaking property percentages until a surface looks good. It’s much more technical now, based on the way the material realistically reacts to light.
Whether you're lighting or texturing you'll need to learn the nodal system. It isn't so much that nodes are sometimes intimidating --we've all seen tangled spider webs of incomprehensible interconnects, but nodes are for the most part unintuitive to many. To successfully make use of them, it takes a lot background knowledge, (outside of traditional 3D subjects) but once that's mastered, it's easy to experiment to achieve the desired look. I've fought with them in the past, mainly using dielectrics for realistic glass, but there's no choice but to dig in and dive deep now.
Lights have their normal Properties panel, but Volumetrics controls are located in a new Render Panel. Gone too is Background Illumination --which I have always set to zero anyway. The Dome light is deceased – with improvements to the Distant light to cover that function. New to Layout is the Environmental light, used for background lighting. Some lights that were previously invisible to the camera are now visible.
Render options are revamped as well to accommodate PBR. There are a lot of changes to learn to become proficient.
Lightwave has spent the last decade trying to find its way, after a meteoric rise as an affordable, easy-to-use app. The talent and technology drain from the Modo spinoff by some of the original Newtek principles hammered it.
Lightwave's evolution has been interesting to watch. (I started with 1.0 on the Amiga.) LW9 promised a “re-invention”. LW10 was a hastily-rushed embarrassment to look busy as Newtek reassessed it's future, after the parting of the Modo team. Genoma was better than what it supplemented, but for rigging, Maya is still a benchmark. Then came CORE, which demonstrated how lost Newtek really was. LW11 – 2015 refined LW10 with some significant rigging, rendering, and shading improvements but Modeler was left to languish. While Modo made yearly significant changes, Lightwave plodded along, with a heavy reliance on plugins to get the job done. Compare the modeling tools in both programs and it’s obvious why so many well-known artists have moved on. Even the DAVE school gave up on LW, since their job is to place students in the industry with the tools they'll likely use.
Just try to model in a perspective window in LW; it’s easy in Modo. Going silent for the last two years while 2018 was in development drove even more users away, as they wondered if anything was happening at all. The Unified Mesh System in 2018 makes implementation of major changes a lot less painful, so I'm optimistic in two or three years, Lightwave will make more artists sit up and take notice. This may well be Newtek's last chance to claw back some of the market share they've been steadily giving up for so long.
Marketing and innovation used to be two of Newtek's stongpoints, (I'm reminded of the DigiPaint days in the late-80's when they made it seem cool to use their cutting-edge products -and it was!) They need to get back to the fundamentals that made them stand out to attract new users before its too late. It's still the best bang for the buck, but (free) Blender is catching up each year.
Modeling tools have remained stagnant in LW for far too long, and its my hope Newtek makes a major effort to bring meaningfull enhancements to the modeling workflow ASAP. The recent modeling survey conducted by Newtek gives me hope. You can take it here: https://docs.google.com/forms/d/e/1F...XVGig/viewform
I watch the CAD jockeys at work perform modeling functions that take immense effort to duplicate using Lightwave. Modeling eats up precious man-hours in my workflow, and as fast as I am, I’m often presented with multi-million poly CAD models requiring significant reduction using the original as a background reference. Sometimes this can take a week or more to complete a mesh. Texturing, lighting and rendering are pleasant afterthoughts in comparison. Once the model is ready, I can go from request to delivery of a custom render in as little as fifteen minutes. I’d like enhanced tools to model even faster with better results.
QuickBolt is included in the 2018 plugins directory, which was a pleasant surprise (It may have already been included 2015.) It's a real time-saver if bolt threads need to be included. Most of the time, I only model the heads depending on the application, but leaving a handful of bolts adjacent to a hand tool in a scene adds a nice detail sometimes.
The decision to upgrade to 2018 is more about intent than just money. If you're a hobbyist making spaceships on the weekends for fun, it may not be worth it due to the significant amount of time and effort required to flatten the learning curve for the new lighting and shading systems, but you will be rewarded as LW evolves. I hope the particle system evolves to realistically produce fluids similar to RealFlow; not the pudding-looking plugins available now that some users are raving about. Realistic fluid generation is difficult to pull off.
Old models can remain textured as they are, but to get the most out of the render engine, you’ll need to re-surface them as PBR objects.
The following link from the Newtek forums offer two pdf volumes of PBR information that are helpful to get started with the new terminology and the reasoning behind the changes in LW 2018.
--For those of you who've upgraded, share your opinions. What do you love, hate, or want to see improved next?
If there's no such thing as a stupid question, what kind of questions do stupid people ask?
|22nd Mar 2018, 10:01 AM||#3|
Thanks for this, I would agree with you absolutely on your major points. I really appreciate the PBR workflow, the overall modernisation towards nodes and the signs of a general improvement overall. I´m sure we´re all aware that Newtek let the ball drop for many years now and it will take more effort to get people like Allegorithmic, Pixologic, and others to embrace Lightwave again. I upgraded simply because the cost was very low but I think Lightwave suffers from being on the outside now, others just aren´t supporting it. I´d like to see that situation change so I can use other 3D programs without hassle.
my two cents
Testing, testing one, two ... is this thing on?
|3rd Apr 2018, 07:13 AM||#4|
PBR or die.
Not lightwave but Blender user here.
And yes, PBR or die.
My YouTube Channel !
|Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)|
All times are GMT -6. The time now is 01:50 AM.