Xeno CX

Version

Target: Lab Pixel Sprite Editor (L.P.S)

Version: LPS v26.02.28 Alpha

Known Issue

  • No Report During This Period

New

  • User can export any projects up to 1000% of its Original Size (For Free) without using the EGU

[Developer Note]

This is the initial implementation of the EGU in LPS Editor. The current integration uses LPS Zoom 32x8k Quad 1640 Prototype 6 GB. While this does not represent the final version currently in development, it will be included in the upcoming LPS Editor v26.02.28 Alpha.

Testing was conducted on NVIDIA GTX 1660 Ti with 6 GB of VRAM. During recording, Pix. P. was reduced to 50% due to VRAM constraints introduced by the recording process. Under normal conditions, the GTX 1660 Ti operates at 100% Pix. P. without issues. Systems with less than 6 GB of VRAM require a reduced Pix. P. setting.

 

Estimated Pix. P. Requirements
The following values are projections and should not be considered definitive benchmarks:

  • 51%-100%: Stable on 6 GB VRAM or higher

  • 50%: Stable on 4 GB VRAM or higher

  • 25%-49%: Intended for systems below 4 GB VRAM

Values below 24 percent are not recommended, as the EGU may fail to initialize or execute properly.

 

Compatibility Notice
It is acknowledged that the EGU may not function on certain systems. This is due to its reliance on virtualization through the Oblinesis Engine Virtual PC. The Virtual PC environment enforces strict hardware compatibility requirements, which may prevent execution on unsupported configurations.

 

EGU used in the video:

LPS Zoom 32x8k Quad 1640 Prototype 6 GB, configured with Half Pix. P. enabled and Maximum QT applied. This EGU operates on LSS 1.5.

  • Added EGU Support
  • Added VRM Support.
  • Added Virtual PC Compatibility
  • Added LPS Zoom 32x8k Quad 1640 Prototype 6 GB as default in adept and master tiers.

Improved

  • Loading screen have been improved
  • Loading screen have been expanded
  • Xelab AI Detection improve for mismatch or incorrect input during LPS Zoom Process
  • Xelab AI has been improved

Adjusted

  • Vixel Cost per generation has been reduced significantly.

Removed

  • Xelab AI Models v1 will no longer be supported as we switch to Xelab AI v2 models.

  • LPS Zoom 16x8k – Fine Detail (No Longer Supported)

  • LPS Zoom 16x8k – Pixel Sharp (No Longer Supported)

  • LPS Zoom 16x8k – Soft Blend (No Longer Supported)

  • LPS Zoom 16x4k – Fine Detail (No Longer Supported)

  • LPS Zoom 16x4k – Pixel Sharp (No Longer Supported)

  • LPS Zoom 16x4k – Soft Blend (No Longer Supported)

  • LPS Zoom 8k – Pixel Exact (No Longer Supported)

  • LPS Zoom 4k – Pixel Exact (No Longer Supported)

  • Vixels in store have been removed. Vixel can now only be purchased in Xenoame Launcher

Bug Fixed

  • Resizing images with Bilinear makes the image darker

  • Export warning is always showing even without using Enhancer

  • LPS Zoom has incorrect input but this visually doesnt affect the image output because the Xelab AI was correcting them. (This was fixed and now it is getting correct input and improved Xelab AI to detect more errors)