What does Comfy's own upscaling handbook say about choosing and preparing upscalers for image and video, and how strong is that evidence?

Problem. Choosing between conservative and creative upscalers and avoiding wasted passes. Conditions. Comfy blog post dated 2026-02-06; vendor-authored, qualitative side-by-side comparisons, includes partner-node (paid API) and local open models. Method. The post's takeaways: choose by use case (portrait skin, product photography needing faithful labels and edges, landscapes, illustration, archival video, CGI); conservative models for faithful materials and text; do not rely on upscaling to fix common AI artifacts; for SeedVR2 downscaling to about 0.35 megapixels with ImageScaleToTotalPixels before upscaling gives better results (its claim); video options named include SeedVR2, FlashVSR, HitPaw, Topaz Astra and Wan 2.2 upscale, with workflow downloads provided. Failure modes. Hallucinated detail from creative models; generation time varies strongly by model and hardware (the post lists per-model times without hardware or n for most items). Limitations. No objective metrics (PSNR/LPIPS) are reported in the text read; comparisons are visual and the vendor has partner-node interests. Treat as creator-grade claims. Version and date. Post dated Feb 06, 2026; fetched 2026-10-23. Curator qualifications: All cited documentation excerpts located on the live pages by curator GET; no video sources. Limitations: Qualitative claims by the platform vendor; no replication. Evidence status: statements located in the cited sources by a correlated desk review (Claude); not executed; not independently reviewed.

{
  "citations": [
    {
      "url": "https://blog.comfy.org/p/upscaling-in-comfyui",
      "note": "Image Upscale TL;DR; Image Upscale TL;DR"
    }
  ]
}