How to Use the Desktop Software

1.1 System Requirements

1) Minimum / Recommended Hardware

2) Supported GPUs

The AI PIXELL PRO desktop app uses TensorRT acceleration on the GPU, so an NVIDIA GPU meeting the conditions below is required.

Required

• NVIDIA GPU
• TensorRT acceleration support
• Windows 10 / 11 (64-bit)

3) Unsupported / Caveats

• NVIDIA GTX series (e.g. GTX 10/16xx): Classified as unsupported because TensorRT acceleration conditions may not be met or may be limited.
• AMD / Intel GPUs: Unsupported
• External GPUs (eGPU): Not recommended as issues can occur depending on connection type, bandwidth, and driver setup.

4) How to Check Your GPU (Windows)

• Task Manager → Performance tab → check the model name under GPU
• Or check Device Manager → Display adapters

5) Troubleshooting

• If errors keep occurring during launch or processing, first check the following.
  • Update the NVIDIA driver (Studio Driver recommended)
  • On laptops, set the app to run on the external GPU (High-performance NVIDIA processor)
  • If GPU VRAM is insufficient (large videos / high-ratio upscale), lower the resolution / concurrent jobs and retry

1.2 Install & Run

1) Purchasing the Program

1. Visit the aipixell.com page.
2. Click the Buy button on the home page.

   

3. On the login page, create an account via Sign up and log in.

   

4. Accept the payment terms and complete payment. (AI PIXELL PRO is a perpetual license.)

   

5. Payment supports card, bank transfer, and easy pay.

   

2) Downloading the Program

1. On My Page, click the Download button for your purchased license.

   

3) Changing the License PC

1. On My Page, click the Change PC Request button to deactivate the license on the current PC.
   • License transfer is allowed only once per year.

   

4) Installing AI PIXELL PRO

1. Run the setup.exe file downloaded from the website.

   

2. When the setup wizard starts, click [Next] to continue.

   

3. Use the Browse button to set the install path. The default path is the Desktop.

   

4. Choose additional options like shortcuts, then click [Next].

   

5. Installation will proceed. Please wait until it completes.

   

6. After installation, launch AI PIXELL PRO.

   

2.1 Screen Layout

1) Import Panel

An area where you load and manage video / project files for processing. Added files appear in the work queue for downstream processing (model application / output settings).

Key Features

• Add Files: Select a folder from local storage to list video files.
• Multi-select / Bulk Import: Select multiple files at once and add them to the queue.

Information shown in the file list

• Duration: total length (min / sec)
• File name: name of the loaded file
• Resolution / Frame rate: basic specs of the input video

2) Work Queue / Watch Folder Panel

An area for managing files to process as a task list and operating the overall processing flow.

Core Concepts

• Work Queue: A list where files / tasks to process accumulate. Each item has its own settings (model / options / output path) and is processed sequentially based on the run mode.

• Watch Folder: Monitors a specified folder so new / copied files are automatically registered in the work queue.

Key Features

• Add Task
  • Add files selected in the Import panel to the queue
  • Auto-add new files arriving in the watch folder
• Order / Priority Management
  • Reorder queue items to prioritize important tasks
  • Group selected files and process them as a group
• Output Path Settings
  • Set the output folder where AI-processed results will be saved
• Run / Stop / Retry
  • Run selected tasks
  • Stop running tasks (where possible)
  • Retry failed tasks (including after modifying settings)
• Per-task Settings View
  • Check resolution, source/output metadata, save path, applied model, and status

Tips / Cautions

• AI PIXELL PRO currently supports only MP4, MOV, MKV, AVI, WebM, MXF. Check the file format before processing.
• Watching network or external drives may have detection delays. Local, stable paths are recommended when possible.
• If failures repeat, first check the input file for corruption, supported format, output path permission / disk space, then review logs / error messages.

3) Preview Panel

An area where you can compare the source clip with the AI-processed result and optionally export only a selected segment.

Layout

• Left (Source view): Shows the input source video / clip.
• Right (Output view): Shows the result with the selected AI model applied.
• Bottom (Timeline / Storyboard): Shows the clip as a timeline storyboard.

Key Features

• Source vs Output Comparison
  • Compare before / after results quickly via left / right views.
• In/Out Range and Range Export
  • Use the bottom timeline to set In/Out and export only the desired range.
• Playback Controls
  • Play / pause and other basic controls to inspect range and results.

4) Output Settings Panel

A menu to choose the AI model (Upscale / Sharpening / Denoise, etc.) for the selected task, configure output format, codec, save location, and then start the job (encode / render).

Key Features

• AI Model Selection: Choose a model that fits the purpose (resolution boost / sharpness / denoise).
• Auto De-Interlace: For interlaced files, the De-Interlace menu is enabled automatically.
• Output Format & Codec: Pick a container format (MP4 / MOV, etc.) and codec (H.264 / H.265 / ProRes, etc.).
• Output Quality / Bitrate: Use a preset or custom values for bitrate mode / value.
• Resolution / Frame Rate: Keep source or change.
• Start Job: Confirm settings and start the job. Progress is shown in the work queue.
• Save Preset: Save output settings as presets for reuse.

Tips / Cautions

• Codec / format combinations can drastically change file size and processing time — use the defaults that match your purpose first.
• Compare source and result in the preview panel by range (In/Out) first, then export the full clip.
• Jobs may fail if the output path has disk space / permission issues. Check the save location before running.

3.1 Add Files to the Work Queue

1. Select a folder in the Import panel.
2. Files inside the selected folder appear in the file list.
3. Select files to process from the list. (Multi-select supported.)
4. Use the checkbox at the top of the file list to select all files at once.
5. Press Add to Queue to add to the queue.
6. Confirm that items have been added to the queue.

3.2 Selecting a Task File

1) Single File Select

1. In the work queue, click the left checkbox of the file you want to process.

   

2) Group Multiple Files

1. Click the left checkbox of every file you want to group together.

   

2. Click the Group button at the top — `Group1` is created and the selected clips are added to it.

   

   

3. Drag a file into an existing group to add it to that group.

   

4. To apply an AI model to a group, click the checkbox to the left of the group name to select it.

   

5. Use the [Ungroup] button to dissolve a group.

   

3) Watch Folder Setup

1. Click the [+] button at the right of the Watch Folder tab to specify a watch folder.

   

2. Verify the information of the selected watch folder.

   

3. Watch folders can be removed via the [Delete Watch Folder] button.

   

3.3 Selecting an AI Model

In the Output Settings panel, choose the AI model (Upscale / Sharpening / Denoise) that matches your purpose.

Upscale 2x

Uses AI to upscale the source resolution by 2x while improving detail and sharpness and reducing noise.

Sharpening 1x

Keeps the resolution while boosting sharpness for a crisper image.

Denoise 1x

Keeps the resolution while removing unwanted noise. You can pick the noise strength (Light / Strong); for milder noise, applying Light first is recommended.

Auto De-Interlace

For interlaced sources, Auto De-Interlace is enabled automatically to remove interlacing.

3.4 Output Settings

1) Encoding Support Range

2) File Output Settings

Resolution

• With Upscale model selected: output resolution is automatically upscaled by 2x.
• With other models (Sharpening / Denoise): defaults to the original resolution.
• Custom resolution: you can change the resolution, but without selecting the Upscale model the change does not apply AI upscaling — it is a simple resize.

Video Codec

• Choose a codec from the codec list (supports H.264 · H.265 · VP9 · Apple ProRes · DNxHD · DNxHR).

File Format

• Choose a file format (e.g. MP4 / MOV) that matches the selected codec.

3) Video Output Settings

Frame Rate

• Defaults to the source frame rate.
• You can keep the original frame rate or change it to a desired fps for your purpose.

Profile

• High: a quality-focused profile with higher compression efficiency. Better for detail preservation at the same bitrate, but compatibility may be limited.
• Main: a compatibility-focused profile suited to a variety of playback environments. Recommended when stable playback is needed.

Quality

• Best — quality first
  • Lower compression (= larger files) · longer encode time
  • Preserves detail / grain; suited for post-production / archives
• Good — balanced (recommended default)
  • Suited for general online uploads / internal sharing
• Low — size / speed first
  • Higher compression (= smaller files) · faster encode
  • Detail loss; block noise / banding may be more visible

Timecode Mode

• NDF (Non-Drop Frame): timecode increases sequentially with frames. (Suited for frame-based editing / post-production.)
• DF (Drop Frame): at 29.97 / 59.94fps, skips some numbers so the displayed timecode matches real time. (Suited for broadcast / real-time runtime alignment.)

Bitrate

Bitrate is the amount of data per second of video (Mbps) — higher values improve quality but increase file size and encoding time.

• CBR (Constant Bitrate): keeps the bitrate near-constant throughout. (Best when bandwidth must be steady — streaming, broadcast.)
• VBR (Variable Bitrate): distributes the bitrate based on scene complexity. (Better quality-to-size efficiency at the same target.)
• Target Mbps: the average / baseline bitrate the encoder aims for. Higher = better quality and larger size.
• Max Mbps: the peak cap that the encoder must not exceed even on complex scenes. Mostly meaningful for VBR / ABR.

4) Audio Output Settings

Audio Codec

• Defaults to the source audio codec.
• Output codec supports AAC / OPUS / PCM depending on the video codec. (Other codecs not supported.)

Audio Channels

• Defaults to Stereo; Mono is also available if needed.
• Channel options are enabled only when AAC / OPUS / PCM is selected.

Sample Rate

• 44.1 KHz / 48 KHz / 96 KHz supported.
• Sample rate options apply identically regardless of the selected audio codec.

3.5 Creating & Managing Presets

A preset stores frequently used output settings (model / codec / resolution / bitrate, etc.) and lets you apply them all at once to future jobs.

1) Creating a Preset

• Manage: Click the manage menu to open the preset management window.
• Click the New Preset button to open the preset settings menu.
• Name the preset, finish the AI model / export settings, then click Save.
• Created presets can be edited · copied · deleted in the management window.

2) Applying a Preset

• Select Preset: Choose a saved preset to apply to the current job.

3.6 Running Jobs & Status

1) Output Path

• In the work queue list, click the output path link and select a folder to save to.

2) Output Range

• By default, the entire video range is set.
• To set a range, move the left / right cursor-shaped bars to specify the range.

3) File Output

Start Job

• Press the [Start Job] button at the bottom of the settings panel to begin output (rendering).

Status

• Once started, you can check progress (%) in the status pane.
• The estimated remaining time is shown in the center of the preview panel.
• When finished, the status message shows ‘Complete’.
• Settings panels are locked while a job is running.

4) Reviewing Results

When output completes, the generated result is automatically loaded into the preview panel's right side (Output view). You can compare Left = Source / Right = Output simultaneously without reopening the file.

• Left (Source view): the input source clip
• Right (Output view): the just-finished result, loaded automatically
• Use Split (left / right split) or Toggle (source / output switch) to inspect before / after differences more easily.

3.7 Reviewing Output

• Completed jobs produce files in the specified output folder.
• If failures repeat, check in the following order.
  • Input file corruption
  • Supported formats
  • Output path permissions / disk space
  • Error messages / logs

4.1 Interface Settings

Change background color

• From the Settings menu, change the work interface's background color.

Change language

• From the Settings menu, change the work interface's display language.

5.1 Billing · Refund

The desktop software is a $159 perpetual license — pay once, use forever. There are no monthly subscriptions or credit top-ups.

To request a refund

1. Prepare the email and order number (receipt) used for purchase.
2. Submit your refund request and reason via Billing · Refund inquiry below.
3. After review, refunds are issued via the original payment method within 3–5 business days.