<?xml version="1.0"?>
<feed xmlns="http://www.w3.org/2005/Atom" xml:lang="en">
	<id>https://wiki.selur.net/index.php?action=history&amp;feed=atom&amp;title=Vapoursynth_RVRT</id>
	<title>Vapoursynth RVRT - Revision history</title>
	<link rel="self" type="application/atom+xml" href="https://wiki.selur.net/index.php?action=history&amp;feed=atom&amp;title=Vapoursynth_RVRT"/>
	<link rel="alternate" type="text/html" href="https://wiki.selur.net/index.php?title=Vapoursynth_RVRT&amp;action=history"/>
	<updated>2026-11-19T04:55:09Z</updated>
	<subtitle>Revision history for this page on the wiki</subtitle>
	<generator>MediaWiki 1.46.2</generator>
	<entry>
		<id>https://wiki.selur.net/index.php?title=Vapoursynth_RVRT&amp;diff=593&amp;oldid=prev</id>
		<title>Selur: New filter page (via create-page on MediaWiki MCP Server)</title>
		<link rel="alternate" type="text/html" href="https://wiki.selur.net/index.php?title=Vapoursynth_RVRT&amp;diff=593&amp;oldid=prev"/>
		<updated>2026-11-16T08:51:33Z</updated>

		<summary type="html">&lt;p&gt;New filter page (via create-page on MediaWiki MCP Server)&lt;/p&gt;
&lt;p&gt;&lt;b&gt;New page&lt;/b&gt;&lt;/p&gt;&lt;div&gt;&amp;lt;div style=&amp;quot;float:right; margin:0 0 1em 1em;&amp;quot;&amp;gt;__TOC__&amp;lt;/div&amp;gt;&lt;br /&gt;
&amp;#039;&amp;#039;&amp;#039;RVRT&amp;#039;&amp;#039;&amp;#039; (Recurrent Video Restoration Transformer) restores video from many neighbor frames; this filter offers its denoising and deblurring models at the input size. It runs on an NVIDIA GPU through PyTorch, is very slow and needs a lot of GPU memory; the 4× super-resolution models are in [[Vapoursynth RVRT Resize]].&lt;br /&gt;
&lt;br /&gt;
&amp;#039;&amp;#039;Filtering → Vapoursynth → Other&amp;#039;&amp;#039;&lt;br /&gt;
&lt;br /&gt;
[[File:Vapoursynth_RVRT.png|none|RVRT controls in Hybrid]]&lt;br /&gt;
&lt;br /&gt;
== What it does ==&lt;br /&gt;
Frames are processed in groups; each group is aligned to its neighbors and refined recurrently, so the result of a frame depends on the frames around it. &amp;#039;&amp;#039;denoise&amp;#039;&amp;#039; removes noise of an assumed strength, &amp;#039;&amp;#039;deblur&amp;#039;&amp;#039; removes motion blur. Long clips are split into overlapping chunks.&lt;br /&gt;
&lt;br /&gt;
== Requirements ==&lt;br /&gt;
* Windows, Hybrid&amp;#039;s torch add-on (with its vsrvrt build) and an NVIDIA GPU with CUDA.&lt;br /&gt;
* The RVRT models are licensed for non-commercial use only.&lt;br /&gt;
* The filter is only shown when the RVRT resizer is available.&lt;br /&gt;
&lt;br /&gt;
== Settings ==&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Control !! Parameter !! Range / values !! Hybrid default !! Meaning&lt;br /&gt;
|-&lt;br /&gt;
| Mode || – || denoise, deblur || denoise || Denoising or deblurring model.&lt;br /&gt;
|-&lt;br /&gt;
| Sigma || sigma || 0 – 50 || 12 || denoise: assumed noise level.&lt;br /&gt;
|-&lt;br /&gt;
| Model || model || gopro, dvd || gopro || deblur: training set (GoPro or DVD).&lt;br /&gt;
|-&lt;br /&gt;
| FP16 || use_fp16 || on/off || on || Half precision: about twice as fast with half the memory.&lt;br /&gt;
|-&lt;br /&gt;
| Force Tile size || – || on/off || off || Enables the tile size; otherwise the plugin picks tiles from the free GPU memory.&lt;br /&gt;
|-&lt;br /&gt;
| Tile T / H / W || tile_size || T 0 – 1024, H/W 0 – 1024 in steps of 8 || 64 / 256 / 256 || Tile size in frames, height and width; 0/0/0 = no tiling.&lt;br /&gt;
|-&lt;br /&gt;
| Force Tile Overlap size || – || on/off || off || With Force Tile size: enables the tile overlap.&lt;br /&gt;
|-&lt;br /&gt;
| Overlap T / H / W || tile_overlap || 0 – 512 || 2 / 20 / 20 || Overlap of the tiles in frames, height and width.&lt;br /&gt;
|-&lt;br /&gt;
| Chunking || use_chunking || on/off || on || Processes the clip in chunks of frames; off processes the whole clip in one pass.&lt;br /&gt;
|-&lt;br /&gt;
| Chunk size / overlap || chunk_size, chunk_overlap || 0 – 127 / 0 – 63 || off, 64 / 16 || Forces the chunk size and overlap in frames; otherwise the plugin&amp;#039;s values are used.&lt;br /&gt;
|-&lt;br /&gt;
| Lazy chunking || preview_mode || on/off || off || Processes chunks on demand and independently; for quick previews.&lt;br /&gt;
|-&lt;br /&gt;
| GPU ID || device || 0 – 8 || 0 || CUDA device.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
== Example ==&lt;br /&gt;
Default settings on a 10-bit YUV 4:2:0 source (640x356):&lt;br /&gt;
&amp;lt;pre&amp;gt;&lt;br /&gt;
clip = core.std.AddBorders(clip=clip, left=0, right=0, top=2, bottom=2) # add borders to achieve mod 8 (vsRVRTFilter) - 640x360&lt;br /&gt;
# adjusting color space from YUV420P10 to RGBS for vsRVRTFilter&lt;br /&gt;
clip = core.resize.Bicubic(clip=clip, format=vs.RGBS, matrix_in_s=&amp;quot;470bg&amp;quot;, range_in_s=&amp;quot;limited&amp;quot;, range_s=&amp;quot;full&amp;quot;)&lt;br /&gt;
# Denoising using RVRT&lt;br /&gt;
import vsrvrt&lt;br /&gt;
clip = vsrvrt.Denoise(clip, device=&amp;quot;cuda:0&amp;quot;, use_chunking=True)&lt;br /&gt;
clip = core.std.Crop(clip=clip, left=0, right=0, top=2, bottom=2) # removing added borders from mod requirement (vsRVRTFilter) -  640x356&lt;br /&gt;
&amp;lt;/pre&amp;gt;&lt;br /&gt;
&lt;br /&gt;
== How Hybrid calls it ==&lt;br /&gt;
* The clip is converted to RGB and brought to mod 8; the output keeps the input size.&lt;br /&gt;
* The device (from GPU ID) and Chunking are always written; sigma, model and FP16 only when they differ from the defaults; tile size and overlap only when forced; chunk size and overlap only when forced with Chunking on.&lt;br /&gt;
* RVRT also offers the [[Vapoursynth common filter options|common options]] (Masked, ApplyOnlyTo, Merge, OnRegion, Gimmick).&lt;br /&gt;
&lt;br /&gt;
== Background ==&lt;br /&gt;
=== RVRT ===&lt;br /&gt;
RVRT (Liang et al., NeurIPS 2022) splits a video into short clips, processes the frames of a clip in parallel and passes features recurrently from clip to clip, aligned with guided deformable attention.&lt;br /&gt;
&lt;br /&gt;
=== vs-rvrt ===&lt;br /&gt;
Lyra-Vhess&amp;#039;s vs-rvrt offers RVRT denoising, deblurring and 4× super-resolution for VapourSynth. Its README suggests spatial tiles that are multiples of 8 (256 as a starting point) and chunks below 128 frames. Without an explicit setting it chunks on its own when a clip is long or would need too much memory; lazy chunking drops the recurrence across chunks.&lt;br /&gt;
&lt;br /&gt;
== Sources ==&lt;br /&gt;
* [https://github.com/Lyra-Vhess/vs-rvrt vs-rvrt]&lt;br /&gt;
* [https://github.com/JingyunLiang/RVRT RVRT]&lt;br /&gt;
* Hybrid&amp;#039;s tooltips&lt;br /&gt;
&lt;br /&gt;
[[Category:Vapoursynth Filters]]&lt;/div&gt;</summary>
		<author><name>Selur</name></author>
	</entry>
</feed>