Skip to content

DLSS 5 Visual Enhancer v3.0

Choose a tag to compare

View all tags
@Merserk Merserk released this 01 Sep 22:53
· 10 commits to main since this release
3.0
c12102a
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Changelog

  • Added sequential batch video processing with reorderable multi-video input.
  • Added ordered per-video results, downloadable successful outputs, overall batch progress, and JSON batch manifests.
  • Added fault-tolerant batch rendering so one failed video does not prevent later videos from being processed.
  • Added reliable batch cancellation that stops the active render, removes only its incomplete output, retains completed outputs, and marks remaining videos as skipped.
  • Preserved input and output video previews for single-video processing while hiding both previews in batch mode.
  • Added shared output renaming controls for images and videos:
    • Auto retains the existing timestamped DLSS5 naming format.
    • Copy keeps the original base filename without adding a suffix.
    • Custom appends a user-defined suffix to every output.
  • Added collision protection so existing output files are never overwritten accidentally.
  • Fixed image processing failures caused by EXIF/TIFF IFDRational and other metadata values that cannot be serialized directly to JSON.
  • Preserved original image metadata for output encoding while converting only diagnostic-report metadata into a JSON-safe representation.
  • Added a synchronized DLSS Model Preset setting to the Image and Video tabs with Default, J, K, L, and M options.
  • Added a new DLSS 5 Settings section directly below the neural-rendering settings in both tabs.
  • Made Default preserve NVIDIA's normal mode-specific behavior without applying a render-preset hint override.
  • Added forced J/K/L/M preset support across DLAA, Quality, Balanced, Performance, Ultra Performance, and Ultra Quality hint parameters.
  • Added native validation and read-back verification for every forced DLSS model-preset hint.
  • Added rejection of unsupported preset values, incompatible legacy workers, and mismatched worker acknowledgements.
  • Added requested and applied DLSS model-preset information to image and video diagnostic reports and native logs.
  • Updated renodx-dlss5.addon64 from RenoDX DLSS5 4.60 to 4.70.
Assets 3
Loading