Selecting AI models and Cloud Rendering in Topaz Video AI
Topaz Video AI
Demo summary
The user walks through the interface to select specific AI models like Starlight Mini and demonstrates the option to use cloud rendering to save local processing time.
Step-by-step
- Navigate to the models section in the interface
- Select the desired AI model based on the task, such as Denoise or Slow Motion
- Choose Starlight Mini for local rendering or select the Cloud option for faster processing
- Upload the file to the cloud to initiate cloud rendering
Options
- Fine-tune settings to recover more detail
- Use Starlight Mini for local rendering
- Use Cloud rendering for faster turnaround
Watch out for
- The Starlight model is extremely computer and graphics intensive
- Local rendering with Starlight Mini can take hours even on decent computers
- Your computer will be occupied for a long time during local rendering
Tips
- Use the cloud rendering option for a faster, easier, and more convenient experience
Highlights
“there is a faster, easier, more convenient option, which is uploading it to the cloud”
AI Video Enhancer
6:301:48Manual parameter tuning in Topaz Video AIThe demonstrator adjusts manual sliders for sharpening, de-haloing, and softening to refine the output of the Nyx model on an old film clip.Topaz Labs- 8:253:05Adding a second enhancement layerThe video shows how to enable and stack a second enhancement layer using the Thea model to add subtle details and fine-grain noise to the upscaled video.Topaz Labs
1:081:33Deinterlace VHS footage with Dione DVThe user demonstrates how to select the 'Interlace' enhancement and apply the Dione DV model with auto-detect field order to remove interlacing artifacts from VHS footage.Topaz Labs- 3:080:45Reduce dark edges with Dione DeHaloThe demonstrator applies the Dione DeHalo model in Topaz Video AI to specifically target and reduce dark edge artifacts on interlaced video.Topaz Labs
- 4:012:04Manual parameter adjustment with Iris LQThe user shows how to enable manual parameters and use the 'Estimate' feature with the Iris LQ model to fine-tune dehaloing and enhancement on low-quality footage.Topaz Labs
- 6:051:10Fix compression and sharpen with ProteusThe demonstrator uses the Proteus model to manually adjust compression fix and sharpening levels on a video clip within Topaz Video AI.Topaz Labs
- 7:461:04Process baked-in interlacing with Dione RobustThe user demonstrates the 'Interlace Progressive' setting using the Dione Robust model to fix footage where interlacing fields have been baked into a progressive render.Topaz Labs
- 9:153:34Denoise and enhance VHS with NyxThe demonstrator applies the Nyx model to denoise and sharpen non-interlaced VHS footage, showing both auto and manual parameter adjustments for grain and noise reduction.Topaz Labs
- 16:332:18Denoising with Artemis presetsThe demonstrator uses the Artemis model's low-quality preset to perform aggressive denoising and sharpening on old footage.Topaz Labs
- 22:461:58Cleaning 8mm film grain with ArtemisThe demonstrator applies Artemis Low Quality and Medium Quality presets to 8mm film footage to remove heavy grain and create a cleaner image.Topaz Labs
0:380:24Live rendering and side-by-side comparisonTopaz Video AI 6 is shown performing live rendering for instant previews and using tabs to swap between side-by-side or split-view comparisons.Topaz Labs- 4:100:49Selecting AI models and Cloud Rendering in Topaz Video AICurrentThe user walks through the interface to select specific AI models like Starlight Mini and demonstrates the option to use cloud rendering to save local processing time.Stefan Malloch