Lydia Choy on Substance 3D Modeler's 1.8 Update & Beta Program

Substance 3D Modeler 1.8’s Main Features

There are many key features that are coming in the next two releases. In our newest release, 1.8, one of the main features is scene cameras and camera settings. You’ll be able to create cameras in your scene with specific settings and switch between them in both the modeling mode and render mode. This can be useful for composing the scene for specific views, whether for a screenshot or a demo or previsualizing a shot. You are also now able to switch between perspective and orthographic projection, the latter of which is useful for more precise editing of top, bottom, and side views of a scene or object. If you are using Modeler in VR mode, you can also enable a Steadicam mode that will make the desktop preview of your viewpoint in VR more stable, which can be very handy for presentations or demos.

In 1.8, there is also a display options panel to toggle the visibility of different elements in your scene, IBL lighting quality improvements, depth of field in the Render mode (previously called Capture mode), the ability to rename materials, ambient occlusion in the Modeling mode, and many UX improvements that our community has been requesting for a long time. These UX improvements include saving your tool, buildup alpha, and stamp settings across Modeler sessions, more keyboard shortcuts, the ability to directly jump scopes, and the ability to rotate your IBL using the same shortcuts as Substance 3D Painter.

We are also working on a scene outliner, which will let you visualize and organize your scene as a hierarchy of node names, as well as primitive layers, which will let you place SDF primitives in your scene that you can Boolean together and edit even after you have placed them. These are two major features that our community has been anticipating for a long time and we hope to have these ready in 1.9.

Beta Testing Opportunities

Our primary strategy is prioritized around listening to our core community’s feedback and addressing their needs. In addition to this, we are actively working on a Mac port, and there will likely be some kind of beta program for this when we are ready to get feedback. We are also paying close attention to how VR hardware and platforms evolve and have been having a lot of discussions about how we should evolve with these industry developments.

There are a lot of research initiatives at Adobe that we want to leverage in Modeler, and we’ve been exploring integrating research in the areas of physics, mesh processing, and machine learning. Getting research breakthroughs into a product often requires a lot of iteration between developers and their customers, and while we do often iterate internally with our artists who work at Adobe, we want to reduce the friction of surfacing these more experimental features to our community. Because of this, we are now offering a public beta on Creative Cloud that will let anyone with an Adobe account download and use a beta version of Modeler. Over time, we hope these experimental features do end up in our stable retail version with the key step of iterating with our customers in the wild who are excited about trying out these features even if they might not be ready for production. This beta program will help us get these experiments into more hands and more industries and will help us figure out how to package them up as first-class features. You will find this beta version of Modeler in the Beta section of Creative Cloud. We have a special Modeler Beta channel on the Substance 3D Discord server to gather feedback from these builds.

The beta versions of Substance 3D Modeler will serve a specific purpose. They are a platform for innovation where users can experiment with new features and workflows in development. Their feedback will be used to refine and validate new workflows that may be considered for inclusion in the release version of Substance 3D Modeler. It’s important to note that these beta releases are not intended for production use, as there may be bugs or workflows being tested that do not make it into the shipping version of Modeler. Users requiring a stable, production-ready version should continue using Modeler as part of the Substance 3D Collection. The beta is for those who want to be at the forefront of testing new capabilities and shaping the evolution of Modeler.

The Role of Feedback

We rely heavily on the feedback of our communities and mostly engage with them on Discord (Modeler channels on the Substance 3D server), the Modeler and Medium Artists Facebook group, and our private beta program and internal artists. Sometimes, we establish closer relationships with various teams and studios and schedule meetings to interview them or walk them through new features or the roadmap to get their opinions. We also often work with our user research team to run studies among different types of customers in areas where we want to grow adoption.

For the partnership with Epic, for example, our user research team interviewed many different types of Unreal Engine creators, and we bubbled up features that were priorities in a real-time asset creation workflow. This included many improvements to our mesh decimation and export pipeline and a lot of refinement to our desktop experience since a lot of those types of creators do not have VR headsets. We will continue to improve in that direction as the real-time experience community adopts the tool and gives us more feedback.

Modeler Open Beta’s New Tech

The main goal of the Modeler Open Beta program is to surface experimental features from our research teams sooner rather than later so that we can gather public feedback and iterate on these features with a broader community. There is a breadth of research that we would like to test out in our workflow around mesh processing, physics, machine learning, and asset categorization and search, and a lot of it is experimental enough that we are not sure if the use cases are broad enough to build upon. Many of these features will be focused on our scene assembly and kit-bashing workflows and may also be too unstable for people who are using the tool for their professional work and may not ship in our retail build for quite some time. We didn’t want these barriers to keep us from being able to gather the vital feedback needed to refine these experiments and ultimately ship them in a polished manner.

Improvements

A lot of the recent focus has been on filling out functionality around sculpting and surfacing (for both desktop and VR) and bringing the desktop interface to a more mature state so that customers without VR headsets will also have a more creatively-flowy experience. We’ve also improved our export functionality and artists will find that they may not need to export to an intermediate tool as often to improve their topology before using it in their downstream tools. This will continue to get better over time as we integrate more work from research teams who are working on faster and better mesh processing. We will also continue to refine our sculpting tools as we get community feedback while shifting our main development towards scene assembly more because that will ultimately be the area where we plan to layer on more complex and experimental workflows. The goal is to allow the user to manage complex assemblies, whether they are complex objects or characters or elaborate scenes and environments.

Plans For Future Releases

There have been some process changes that we’ve implemented on our team with the goal of more frequent releases without sacrificing testing, stability, and overall quality of Modeler. The goal is to be able to release at a roughly monthly cadence for both the live app and the public beta. We hope this increases engagement with our community and gives us the ability to respond more quickly to feedback by shipping improvements sooner rather than later. Occasionally we may delay releasing if major changes are coming online that require more testing or iteration before it is ready, or if we are working towards a wider announcement or press event. Otherwise, it shouldn’t be too difficult to hold to the monthly cadence during most of the year.

Long-Term Vision for Modeler

Our vision is to build a tool that lets you create and design in 3D without having to think heavily about the technical constraints during the creative process, and to automate the interop between other tools as much as possible. We believe that when an artist is in a state of creative flow they do their best work, and we also want to unlock the best “editing surface” for the task at hand. Sometimes, this means using 6-DoF controllers or a VR headset, and sometimes it means using a desktop interface with pen, mouse, and keyboard, and we hope to unlock more surfaces and input modalities in the future. Our starting point has been sculpting and modeling, and this year we are shifting our focus to scene assembly and kit-bashing. We will see how our community evolves as we ship more experimental features in this area.

Using Customer Feedback For Future Developments

Our development process now roughly follows our release cadence, so we have 4-week sprints. At the beginning of each sprint, we have a planning phase where we bubble up items from the backlog and organize any recent feedback we have gotten from artists. Whether or not a feature or a change gets scheduled for the sprint is a function of how important or urgent the feature may be, how long it may take (or how open-ended this estimate may be), the level of risk or dependencies, the engineers who are available and engaged to work on that feature, and how well it fits into an overall product narrative or our strategic roadmap. Sometimes if an artist’s request is easy to do, we try to squeeze it into the schedule even if it may not fit into the overall development arc. If a request is difficult, then there is a lot of discussion as a team if it is the right time to take on the task. At the end of each sprint, sometimes larger tasks carry over to the next sprint, and we reassess where we are while also updating our stakeholders to make sure they agree with our direction.

For these discussions to go smoothly, every member of the team must feel aligned towards our overall roadmap goals. If we need more time to find alignment, then often we have these discussions in the background as the sprint progresses and development is happening in areas where we do have agreement. On the team, we have leads in the areas of product, design, engineering, and QA, and we all need to be on the same page for healthy development velocity.

Conclusion

If you are not part of the community already, please join our Discord server where you can find Modeler (and other Substance 3D app) channels. There are many helpful community members there who can answer questions, and the Modeler development team reads the messages there carefully and will also respond and engage in discussion.

• Modeler documentation can be found here.
• Release notes can be found here.
• Our video tutorials can be found here.

Over the next few months, we will be refreshing our video tutorials and also releasing more content that will highlight various techniques to achieve a wide range of visual targets.

Lydia Choy, Director of Product/Engineering of Substance 3D Modeler

Artwork by Joshua Eiten: Senior Staff Designer, and Gio Nakpil: Head of Visual Development, Substance 3D Evangelism