Postshot just released their V1.1. When we covered Postshot V1.0.324 last month, many of the final features were being implemented, such as expanding format compatibility and addressing edge cases in the capture stack. V1.1 makes the jump from incremental updates to a genuine feature release. Three additions in particular stand out as changing what Postshot can actually do, not just how well it does it.
Postshot now supports previewing scenes directly in a connected VR or XR headset without leaving the application. Postshot has also added an Observer Scale preference, letting you control the physical scale relationship between the viewer and the scene.
The selection system has been substantially expanded. The previous depth-brush approach has been supplemented, and for most workflows probably superseded, by a lasso select and rectangle select tool, plus explicit Add, Subtract, and Intersect selection modes accessible via toolbar buttons. The Escape key behavior has also been refined so it aborts only the current drag operation rather than wiping the entire selection. For scene editing and crop/cleanup work, this brings Postshot’s selection toolkit closer to what you’d expect from a dedicated 3D application.
Photometric Compensation, which we noted as an emerging feature in March, is now formally shipped in V1.1. The implementation handles auto exposure, white point correction, and vignetting, all of which affect how the reconstruction interprets input imagery. Scenes captured under varying exposure conditions or with significant lens vignetting have historically been harder to reconstruct cleanly; the compensation step allows Postshot to account for these photometric artifacts as part of training rather than relying on inputs to be already normalized. The result should be better color consistency and fewer floaters in scenes that would previously have required external preprocessing.
The SfM pipeline gains a Pose Quality option in the training configuration, giving users control over how aggressively the system filters camera poses during reconstruction. This is meaningful for challenging capture scenarios — sequences with motion blur, partially occluded cameras, or difficult lighting — where the default pose estimation may include unreliable cameras that degrade the final splat quality.
On the format side, V1.1 adds the ability to load sequences of SPZ files. Combined with support for Pix4D OPF data, RealityScan 2.1, and improved COLMAP database handling, the import surface continues to expand. Unreal Engine 5.7 is now supported on the plugin side, alongside a series of depth compositing and asset path improvements that make the UE integration more reliable in production configurations.
The V1.1 build is available for Windows now at jawset.com.
