Getting Started
System Requirements
anelo runs on your local GPU. Processing speed scales directly with GPU capability — faster hardware produces faster results, but the same output quality is available on entry-level hardware.
Minimum requirements
The minimum configuration runs all pipeline stages but at reduced speed. Expect roughly 4x longer processing times compared to the recommended configuration.
Recommended requirements
The recommended configuration processes a full hour of 4K content in approximately 8 minutes, including depth estimation and stereo conversion.
GPU memory considerations
VRAM is the most common bottleneck. At 4K resolution, depth estimation requires approximately 4GB of VRAM. Upscaling with Real-ESRGAN requires approximately 6GB. If your GPU has less than 8GB, the pipeline automatically tiles large frames to fit within available memory at the cost of some speed.
Disk space
anelo itself requires approximately 2GB of disk space. Processing generates intermediate frame sequences (PNG format) that can be large. A 1-hour 4K video produces approximately 50GB of intermediate frames. Ensure you have sufficient free space on your output drive.
Settings reference
| Setting | Type | Default | Description |
|---|---|---|---|
| GPU (minimum) | hardware | NVIDIA GTX 1060 6GB / AMD RX 580 8GB | CUDA 11.8+ or ROCm 5.4+ required |
| GPU (recommended) | hardware | NVIDIA RTX 3060 12GB or better | RTX 4090 24GB for fastest 4K processing |
| CPU | hardware | 4-core x86_64 | Used for frame extraction and encoding. GPU does the heavy lifting. |
| RAM | hardware | 8GB minimum, 16GB recommended | More RAM helps with large batch queues and 4K frame buffering |
| Disk | hardware | 2GB for app, 10GB+ working space | SSD recommended for intermediate frame I/O |
| OS | software | Windows 10+, macOS 12+, Ubuntu 22.04+ | Other Linux distros work but are not officially tested |
GPU (minimum)
hardwareNVIDIA GTX 1060 6GB / AMD RX 580 8GB
CUDA 11.8+ or ROCm 5.4+ required
GPU (recommended)
hardwareNVIDIA RTX 3060 12GB or better
RTX 4090 24GB for fastest 4K processing
CPU
hardware4-core x86_64
Used for frame extraction and encoding. GPU does the heavy lifting.
RAM
hardware8GB minimum, 16GB recommended
More RAM helps with large batch queues and 4K frame buffering
Disk
hardware2GB for app, 10GB+ working space
SSD recommended for intermediate frame I/O
OS
softwareWindows 10+, macOS 12+, Ubuntu 22.04+
Other Linux distros work but are not officially tested