Infognition Super Resolution plugin for VirtualDub resizes video using Super Resolution approach: by accumulating information from previous frames it can provide much more details than any other resizing method.
Here are some examples of 4x upsizing with Super Resolution:
13 Dec 2013: version 2.0 released. Changes:
- Works 1.5 – 3 times faster than 1.0 (reaches real time in YV12 and fast mode).
- Slightly higher quality (+0.5 dB PSNR).
- Natively supports YV12, YUY2 as well as RGB24 and RGB32. The YUV modes work 1.5 times faster than RGB, so when used as a first filter in a chain and decoding source video to YUV works much faster than older version (~2 times faster).
- Adjustable parallelism, simpler quality settings.
Just download the archive, unpack it and select superres32.vdf file if you’re using 32-bit version of VirtualDub, or superres64.vdf file if you’re using VeeDub64. Place that .vdf file to/Program Files/VirtualDub/plugins/ or “plugins64” folder correspondingly.
VirtualDub 1.9.1 or higher. Windows XP SP2 or higher (Vista, 7, 8). CPU with SSE2 support (virtually any CPU since 2001). Recommended: 4 cores or more to get HD video in real time.
The plugin can resize video to any desired width and height. However the SR engine upsizes video exactly 2 times in both dimensions. If you set output size less than original size, SR is not used and a high quality image interpolation method resizes the frames. When output size is the same as input size, nothing is done at all. If new width, new height or both is bigger than original value, SR engine upsizes video two times and then if new dimensions are not exactly twice bigger than original ones, an image interpolation method is used to resize video from 200% to desired width and height. Two consequences: upsizing to 200% is the fastest upsize mode. If you need to want more than 200% the best quality is obtained by making a chain of two Super Resolution filters: the first one upsizing to 200% and the other upsizing further.
We’ve compared this technology with different plugins for VirtualDub, Adobe After Effects and AviSynth as well as several stand-alone applications like vReveal and CinemaHD. See Video Resize Shootout. This plugin shares the Super Resolution engine with Video Enhancer, so in the comparison see the results of Video Enhancer.
That comparison demonstrates strong advantage of our super resolution implementation over all available video resizing tools.
The engine is suited for sequential processing, frame by frame. The more frames are processed sequentially the more details are accumulated. When saving result to a file video is processed from start to end in normal order and maximum quality is obtained. When seeking to random places in preview mode, there is no good history of processed frames and accumulated information, so the plugin takes 5 source frames (N-4 … N) and processes them before showing you the result. Visual results in preview mode when jumping to random frames can be worse than when rendering to a file.
Version 2.0 natively supports RGB24, RGB32, YUY2 and YV12 color modes. When working in other color modes VirtualDub will convert automatically. Most videos decompress to some YUV modes and if you apply this filter before any RGB-only filters (and don’t force VD to work in RGB) then it will work in YV12 or YUY2 which is substantially faster than RGB modes.
Because of high computational complexity of the technology Super Resolution plugin is not very fast, but it’s highly parallelized, providing a good speed boost on multicore and multiprocessor systems. On a Quad Core 2.33 GHz PC upsizing 960×540 to 1920×1080 is done with 14-19 frames per second depending on quality settings, and some lower resolutions get processed in real time or faster. (Speed depends on image area, so 1280×720 is 2.25 times faster than 1920×1080).
Super Resolution plugin works in VirtualDub 1.9.1 and later versions. Older versions didn’t support necessary functionality to get previous frames. For the same reason it doesn’t work in Video Enhancer as a VD filter, and there’s no sense in loading it in VE since the super resolution engine is already built into Video Enhancer.