Kostack Studio

Every perspective has its own sound.

One unusual aspect of this project is that changing a camera perspective doesn't just mean rendering the images again—the audio has to be regenerated as well.

Just like in Part 1, the sound is generated individually for each shot, taking the camera position into account. The system analyzes the movement of the simulated geometry and derives collision events from it, then turns these into sounds with distance attenuation, frequency filtering, and stereo positioning.

This means each perspective gets its own spatially matching soundtrack. A change in camera position can therefore require both the visuals and the corresponding audio to be regenerated.

The good news is that all perspectives are now done. What's left is mostly polishing and final cut.

5 hours ago | [YT] | 49

Kostack Studio

Working with a really big simulation model.

The model has over 21 million vertices—almost three times as many as the first simulation. Its point-cache file alone is 460 GB in size.

The model grew considerably because it had to include the complete geometry resulting from the first simulation, along with WTC 7 and the adjacent pedestrian bridge.

This also created a practical problem: even under ideal conditions, changing a single frame took at least four seconds. That was simply too slow for working comfortably with the scene.

To make the project more manageable, a proxy model had to be created containing only faces above a specified minimum size, along with its own animation cache. This reduced the point-cache size to 76 GB and brought frame changes down to around half a second, making the scene much easier to work with.

The image shows the proxy model in the viewport. What's surprising is how complete it still looks, even though a large portion of the original geometry is no longer there. At first glance, there are hardly any visible gaps.

The final video will, of course, be rendered using the complete model. The proxy is only used while working on the scene—for things like setting up cameras, lighting, and other scene work.

1 week ago | [YT] | 275

Kostack Studio

A seamless continuation.

The new simulation begins exactly where the first part of WTC Simulation 2025 ended. The final simulation result of the South Tower was transferred directly into the North Tower simulation, becoming part of its starting geometry and allowing the debris and surrounding scene to continue seamlessly into the next phase.

And there's another interesting detail: the debris from the first simulation remains fully dynamic, continuing to react and interact with the collapse of the North Tower.

This image shows the starting point of the new simulation.

As for the progress: the simulation is complete, and the full video is now in the final rendering stage. It should only take a few more days before it's ready.

Channel members can already watch a preview featuring the first finished minute and a complete simulated sequence from one perspective. The finished video, as usual, will be available to everyone.

1 week ago | [YT] | 167

Kostack Studio

The North Tower simulation is complete—yay!

The video is still rendering, so I’ll likely need a few more days before the full version is ready.

For the truly impatient channel members, I’m uploading the first finished minute of the new video right now, featuring a complete simulated collapse sequence from one perspective.

The full video, including all perspectives, will be available to everyone in a few days—as soon as the rendering is finished.

1 week ago | [YT] | 262

Kostack Studio

A new documentary is coming soon: 32 Frames: A 9/11 Mystery.

The film follows the investigation into 32 unreleased photographs recovered from the World Trade Center rubble after 9/11. I had the opportunity to contribute a small amount of footage to the project, and here is the first trailer.

According to the filmmakers, the documentary is expected to be released before the anniversary.

1 month ago | [YT] | 100

Kostack Studio

The new Blender Open Movie "Singularity" is out. The Blender Studio went fully stylized this time, utilizing Geometry Nodes and custom brushstroke tools for a unique watercolor look. 🌌

Like all Blender Open Movies, the entire production files, assets, and software updates are shared openly with the community, driving the development of the tool itself.

4 months ago | [YT] | 11

Kostack Studio

The creator behind the YouTube channel DzungBlenderDestruction, from Vietnam, used our BCB software to run simulations of this triple-tower skyscraper concept. The model includes multiple connecting bridges at different heights. This video shows the results, exposing the structure to various earthquakes and an airplane impact. You can find more of their simulation videos on their channel—feel free to check them out.

#BCB #Simulation #StructuralEngineering

4 months ago | [YT] | 17

Kostack Studio

Some people have asked why there is now a high-tier membership on the channel, so I wanted to briefly explain the idea behind it.

Most of the destruction and collapse simulations on this channel are created using Bullet Constraints Builder (BCB), an open-source simulation framework for Blender that Oliver and I originally developed between 2015 and 2018 as part of an EU-funded research project.

Since then, the software has remained publicly available and has continued to be maintained and improved over many years, despite having almost no direct funding.

The new “BCB Dev Sponsor” membership is intended for viewers, professionals or studios who would like to support the long-term maintenance and future development of the project.

This includes ongoing compatibility work, technical improvements and eventually future platform updates such as Blender 5 support.

The regular memberships remain completely unchanged and are still the main way to get access to exclusive simulation previews and behind-the-scenes content.

Thank you to everyone supporting the channel and the software project.

inachuslaurea.wordpress.com/

4 months ago | [YT] | 15