VideoProc Converter AI
Version 8.11 · Digiarty Software · Windows 10 / 11 (64-bit) · ranked #166 in the library
Handles the jobs that surround an edit rather than the edit itself, converting between formats, compressing for delivery, capturing the screen, and running upscaling, frame interpolation and stabilisation over footage that needs rescuing.
Read the installation section below before you start. The order of the steps matters on this release.
About this release
Overview
VideoProc Converter covers the unglamorous half of working with video. Converting a folder of clips into something an editor will accept, compressing a finished export down to a size a client can actually receive, pulling the audio out as a separate file, trimming and merging without a full re-encode, capturing the screen and the webcam for a walkthrough, and ripping a disc into a file. None of that is creative work and all of it eats an afternoon if the tool is bad.
The AI section is the newer half. Super resolution enlarges footage by reconstructing detail rather than interpolating it, with separate models for real footage and for animation, which matters because the two fail in completely different ways when treated the same. Frame interpolation generates intermediate frames to convert a twenty four frame source to sixty, or to produce genuinely smooth slow motion from a normal frame rate. Stabilisation analyses camera motion across the clip and reframes each frame to remove the shake, with a crop margin you control.
Everything is built on hardware acceleration and the application is honest about when it is using it. Encoding and decoding run on the GPU where the hardware exposes an encoder, which on a modern card turns a long transcode into a short one, and the AI models run on the GPU too with a processor fallback for machines without one. Output presets cover the common delivery targets with resolution, frame rate, codec and bitrate already set, and every one of them can be edited and saved.
What it does
Feature breakdown
Format conversion and compression
A wide codec and container matrix in both directions, with quality, bitrate and two pass options, plus presets aimed at specific delivery destinations.
AI super resolution
Enlarge footage with detail reconstruction rather than interpolation, using separate models for live action and for animated sources.
Frame interpolation
Generate intermediate frames to raise the frame rate of a source or to produce smooth slow motion without the stutter of frame duplication.
Stabilisation
Analyse camera motion across a clip and reframe each frame to remove shake, with an adjustable crop margin and a preview of the correction.
Screen and webcam recording
Capture the screen, a region, the webcam or both together with system audio and microphone on separate tracks.
Hardware accelerated pipeline
GPU decode and encode where the hardware supports it, applied to conversion, compression and the AI models alike.
Changed in this build
Version 8.11
- Super resolution model updated with better handling of compressed source footage that carries block artefacts.
- Frame interpolation now handles scene cuts without generating a blended frame across the cut.
- Faster batch queue startup on lists of several hundred files.
- Stabilisation preview renders at a lower resolution first so the settings can be judged before committing to the full pass.
- Additional output presets added for vertical and square delivery formats.
Installation
Follow these in order
- Go offline before running the installer.
- Extract the archive to a local drive.
- Disable real time protection.
- Run the installer as administrator and complete the standard setup.
- Do not launch it from the finish screen. Apply the included activation step first.
- Open the application, confirm the export dialog offers the full resolution options, then re-enable protection.
Release notes
From whoever packed it
- Block the application in your firewall. An online check reverts it to the trial state with its length limit.
- The AI models are bundled here rather than downloaded on first use, so the AI section works with no connection.
- Set the output folder to a different drive from the source when converting large batches. It roughly halves the time on a mechanical disk.
System requirements
Minimum and recommended
| Operating system | Windows 10 version 20H2 or newer, or Windows 11, 64-bit |
|---|---|
| Processor | Intel or AMD quad core, 2 GHz or faster |
| Memory | 8 GB minimum, 16 GB recommended for 4K and for the AI models |
| Graphics | GPU with 4 GB VRAM for accelerated encoding and AI processing, 6 GB for 4K upscaling |
| Storage | 3 GB for the application plus working space for output files |
Questions about this release
Answered before you ask
- Does the trial length limit still apply?
- No. Exports in this build run to the full length of the source with no watermark.
- Will upscaling fix a heavily compressed clip?
- It helps, and the current model handles block artefacts better than it did, but it cannot invent detail that the compression already discarded.
- Can it convert without re-encoding?
- Yes, where the source and target containers share a codec it remuxes instead, which is close to instant and lossless.
Download
Pick a mirror