Now, granted, some lower-budget and older film and video material can be barely considered featuring actual stereo, and applying this kind of low-end processing probably wouldn"t make it sound any worse (although, you never know). Synthesizing four additional channels from two just using simple filtering sounds nothing like a performance miked for and recorded in actual 5.1 surround. If you read the description of the IC, it contains a low-pass filter, the output of which is routed to a subwoofer, and bandpass filters to synthesize the center channel and back channels. That is _not_ 5.1 or even any resemblance to it. There are special decoder IC"s that transforms a (suitable) stereo signal into the six channels of a 5.1 "surround sound system". The ".1" means there is also a subwoofer (low frequency center speaker). The following picture was obtained using RealSR included in HybridĪnd finally this is the picture obtained using realesrgan-x4plusĪs it is possible to see it is not necessary to perform 4x zoom to observe the details, the difference is very clear and the up-scaled image with realesrgan-x4plus contains more details than the original movie.No, I"m really talking about 5.1, which has three speakers in the front, and two in the back. This is image is very near to the source (fake Blu-Ray) since I used for testing a clip of the movie "28 Days Later" that was filmed using a Canon XL-1s standard-definition camcorder at 720x576 resolution and probably the Blu-Ray was obtain by up-scaling with Lanczos (or something similar) the original movie. To test the model I downsized a short movie (20s) to the resolution of 640x346 (ratio 1.85:1) and then I resized it to 1920x1040. Given the compatibility with RealSR this was the easiest solution to use hybrid to test the model. To test the model i renamed "realesrgan-x4plus" in "x4" and then I override the files in "Hybrid\64bit\vsfilters\ResizeFilter\RealSR\models\models-DF2K_JPEG". The realesrgan-x4plus is probably the best model for upscaling videos (at cost of speed). Realesrnet-x4plus (model trained on DIV2K dataset) Realesrgan-x4plus-anime (for anime video upscaling) Realesrgan-x4plus (for real-life video upscaling) The models are included in the distribution:
The project is called Real-ESRGAN, and there is not a specific Vapoursynth filter, but the models are "compatible" with the models used by RealSR. Recently I found the following spin-off of the project RealSR:
0 kommentar(er)
