For the first time, a full path tracer which enables near photo-realistic rendering is available in Notch, along with several other features and improvements, such as industry-leading denoising and upscaling, new deformer effects, significantly increased video playback performance and over 200 other updates and fixes.
Update to Notch Builder 0.9.23 and see the full changelog here!
We are extremely happy to announce our new software update (download and see the full changelog here), which among other things includes a full path tracer. We want to thank our user community, especially our beta testers who have provided invaluable feedback on this release for the past few months.
Top tip! We’ve written a series of blog posts on the new features in this release, specifically the path tracer, denoisers and upscaling features. Check them out!
In this blog post, we’ll dive a little deeper into the main news of this release, but we strongly encourage you to read the changelog to catch up as well.
Introducing Path Tracing
An industry-standard technique, path tracing is used by many leading high-quality renderers. It simulates the paths of real light including the way light bounces, reflects and refracts through a 3D scene. It also takes into account the materials it interacts with, making near photo-realistic results possible.
The new GPU path tracer is seamlessly integrated into the existing Notch toolkit. The same real-time WYSIWYG environment and nodes our users know and love are all unified with the path tracer. This includes particles, procedurals, cloners, lights, post-processing and compositing. Available as a drop-in component, path tracing can be introduced to scenes and toggled on and off when needed.
The increase in render quality using the new path tracer node makes Notch viable for a whole new range of use-cases. Offering entirely new possibilities for motion designers and artists.
Note: path tracing is a big topic, and we’ve written a whole separate blog post about it which goes deep into the details of the technique, where it’s applicable and how to use it.
We’ve put up a few new tutorials on our YouTube channel which explains the path tracer, and how to use it with different existing Notch techniques like cloners, particles and procedurals. Here’s the first one:
While path tracing is the shiny (literally) new feature in this release, we also have a whole long list of new features, updates and fixes, and you should really check out the full changelog here and catch up on what’s new.
For convenience, we’re listing a few of the top-tier news here as well:
IMPORTANT: 32-bit End-of-Life
This release marks the end of 32-bit support in Notch. We started the journey to 64-bit back in August of 2018, and 0.9.23 is the first release that ships exclusively as a 64-bit executable. This is important to note as it also affects standalone and block exports.
64-bit allows for significantly larger data sets, such as 3D models and animation, textures, images and audio. The hard limit of 3 GB for 32-bit applications meant users were limited in the size of projects they could create and the assets they could use. Trying to exceed that limit would cause problems and instability, which running in 64-bit solves.
IMPORTANT: Time base Change
We have changed the internal time base for Notch projects to a higher granularity. This allows keys and time bars at all common frame rates (12, 15, 25, 30, 50, 60, 100 etc) to be represented accurately. (Previously only 25 and 50 fps were able to be stored accurately.)
This is fully backwards-compatible: projects or bins saved in previous versions of Notch will load and convert to the new timebase without issue. However: it is NOT forwards-compatible: projects saved from 0.9.23 will not load properly in older versions of Notch.
AI Denoising and Upscaling
Denoising is an important part of an efficient path tracing workflow, and in this release, we’re shipping with two industry-leading denoisers: NVIDIA OptiX (GPU accelerated), and Intel OIDN (CPU). They come with different advantages and disadvantages, and we recommend you check out this blog post that goes into more detail.
AI Upscaling using machine learning is another intriguing technique that can be used on both images and videos in Notch. We’ve written a blog post about that too.
Key & Fill for Standalone
A highly requested feature is being able to do classic key + fill when running Notch in standalone mode. Key + fill is an output mode suitable for broadcast where one output renders the RGB signal and the other the alpha. This is incredibly useful for broadcasts where you want to take transparent renders and apply overlays, such as lower-third title graphics etc. We’re happy to say that this is now implemented in Notch 0.9.23. See the manual for usage.
Updated SDKs
Notch supports a large number of 3rd party SDKs and formats, and in 0.9.23 we’ve updated almost all of them; the most important being Cinema4D (v20.004), Spout (3.8), NDI (v2.006) and HAP (April 5th 2019).
Faster Video Playback
We’ve made a number of improvements to our multi-threaded video playback engine, resulting in significantly faster playback for video. You can now bring in larger (and more) video files into your Notch projects without worrying about frame drops or other issues associated with slow playback.
4k Playback
Previously, Builder Pro has been able to act as a Playback license up to 1080p for both blocks and standalone. Today, we’re increasing that to 4K (4096x2160), giving you even more playback power when your Builder Pro dongle is present in your machine.
Just a week ago, we announced new Playback licensing options for leasing and permanent Playback setups — you should check that out if you haven’t already!
