Actors can be converted into Instanced Actors via the IA plugin. During runtime and in the editor, an instanced actor’s visual is generated from the collection of StaticMeshComponents on the actor blueprint.

Problem

Any non-uniform transforms of StaticMeshComps on the actor blueprint aren’t correctly previewed in the editor. Instead, the IA representation in editor shows all meshes with the identity transform, albeit the IA’s location is correct. The relative transform from actor to mesh isn’t correctly previewed. See screenshots and repro project.

Suggested Fix

The relative transform should be correctly previewed.

Problem:
The problem is that when using "Texture Backed Render Target" as the Height Patch Source Mode everything seems to be working as intended when working in the editor at first, but when closing and opening the project again, Landscape patches are not applying properly the first time the level is opened.
The current workaround is to either reopen the level or recompile the blueprint containing the Landscape patch.
Note that the licensee noticed it seems to happen when "Waiting for texture resources to be ready for landscape named:Landscape" on a map startup.

This is because Movie Render Graph does not pass the Assembly object as a context object when evaluating “cat” tokens. This context should always be passed when a user renders a Cinematic Assembly. There is no need to offer an opt-out because always including the context should be considered expected behavior from the perspective of the CAT feature.

Texture size only available for Texture2D, TextureVirtual, Texture2DArray, VolumeTexture and SparseVolumeTexture, not unknown.

This is because Texture Collections that are Bindless mask their texture type with MCT_TextureCollection and FHLSLMaterialTranslator::TextureProperty doesn't take into account that mask, it just checks equality.

When the conditions mentioned in the Steps to Reproduce happens, the following happens:

• GameThread eventually get stuck on FlushRendering

• Workers (even including standbys when option is activated) and renderthread get stuck on FShaderCodeArchive::WaitForPreload

When going for the preload, an extra task that monitor the output of the IO completion event is created. The problem is that when the workers are oversubscribed, the extra tasks are never executed which cause the deadlock even though the IO for those preload requests are probably done.

See FShaderCodeArchive::WaitForPreload and FShaderCodeArchive::PreloadShader, the latter which create a FPreloadShaderTask with the read completion events from the storage as prerequisites.

Increasing the number of standby workers only delay the problem or in the case I studied did nothing at all. Disabling preload didn't seem like a great idea considering the slowness of zen storage (in comparison to straight SSD).

In the end, the solution seemed to be to bypass the prerequisites entirely and wait on the completion event directly (and the FPreloadShaderTask afterward). This enabled the tasks to work as intended possibly due to the task being able to run directly on the same thread during the wait period without having extra coordination through the prerequisite.

In Twinmotion 2026.1, when importing certain SKP files with grouped components, materials may render incorrectly on some grouped components and will appear with a dark, metallic color instead.

This is a regression from Twinmotion 2025.

This does not occur for Datasmith imports.

From testing, this appears to occur only when using the bilateral upsampling path (r.VolumetricRenderTarget.UpsamplingMode=4). Other upsampling modes do not produce the artifact.

I created new module and modular rig and by default it seemed to be working fine. Then I copied over some variables from my own module, added some nodes and the bug started happening also in the new test modular rig. I removed everything I did trying to find the culprit, but the modular rig stayed broken even with just single float variable.

In 5.7 it worked fine.

WORKAROUND: The issue can be temporarily resolved by loading the Rig Module into memory before opening up the Modular Rig. Opening the Rig Module or right-clicking it in the Content Browser should do this.

PhysicsSolverBase.cpp (line 122) is called when exiting a PIE session and might relate to the issue. I've attached it's Callstack to the report.

In a scene that only contains a GC fractured with Uniform and default configurations, I've got a memory allocation LLM Physics/ChaosUpdate during Editor time (valleys):
#PIE sessions|Memory
0|26.60KiB
1|45.29KiB
2|72.00KiB
3|82.66KiB
4|117.40KiB
5|120.04KiB

I won't attach the Trace because it's too big, it exceeds the size limits.

1. $ ldd UnrealTraceServer
2. ./UnrealTraceServer: /lib64/libstdc++.so.6: version `GLIBCXX_3.4.29' not found (required by ./UnrealTraceServer)
3. ./UnrealTraceServer: /lib64/libstdc++.so.6: version `GLIBCXX_3.4.26' not found (required by ./UnrealTraceServer)
4. ./UnrealTraceServer: /lib64/libc.so.6: version `GLIBC_2.33' not found (required by ./UnrealTraceServer)
5. ./UnrealTraceServer: /lib64/libc.so.6: version `GLIBC_2.34' not found (required by ./UnrealTraceServer)
6. ./UnrealTraceServer: /lib64/libc.so.6: version `GLIBC_2.32' not found (required by ./UnrealTraceServer)
7. linux-vdso.so.1 (0x00007ffe0615e000)
8. libstdc+.so.6 => /lib64/libstdc+.so.6 (0x00007f919bfe9000)
9. libm.so.6 => /lib64/libm.so.6 (0x00007f919bc67000)
10. libgcc_s.so.1 => /lib64/libgcc_s.so.1 (0x00007f919ba4f000)
11. libc.so.6 => /lib64/libc.so.6 (0x00007f919b679000)
12. /lib64/ld-linux-x86-64.so.2 (0x00007f919c37e000)

The licensee reports to change the code from :

bool bHasValidCategories = CurrentTagContainer.IsEmpty() || MetaDataCategories.IsEmpty() || CurrentTagContainer.HasAll(MetaDataCategories);

To:

bool bHasValidCategories = CurrentTagContainer.IsEmpty() || MetaDataCategories.IsEmpty() || CurrentTagContainer.HasAny(MetaDataCategories);

Audio that uses an Attenuation to set Spatialized Audio (just positioned, nothing to do with HRTFs or multichannel plugins) does not get rendered with the MRQ files. Initial discovery revealed that the listener is not getting updated correctly in MRQ code.

Slack thread: [Link Removed]

Saving FText "" which has been initialized from FString at cook time resave of source package /Game/MyUserStruct may fix issue.

This behavior is due to the way user-defined-structures initialize default values from a text representation. The parsed quoted string is used to initialize the FText using "FText::FromString()", resulting in the cook warning later on.

For the common case of default initialization, one possible simple fix is to change "FTextStringHelper::ReadFromBuffer()" so that the text is constructed with "FText::GetEmpty()" when the input string is empty, rather than "FText::FromString()".

Note: This warning repros from at least UE 5.4, but probably earlier (and up to mainline latest). Does not seem to be a regression.

They point out that CL 52639755 added window closing logic that is intended to close the picker on actor reinstancing, but that (according to the licensee) is triggering incorrectly when the actor reruns its construction scripts but without reinstancing.