Setting Up a Real-Time Photo Workflow with FramePulse

What Is a Real-Time Photo Workflow?

A real-time photo workflow means your photos go from camera to finished result without manual intervention. You shoot, and within seconds the processed image appears on a second screen — ready for your client to see. No switching to Photoshop, no batch processing after the session. The entire pipeline runs automatically in the background.

FramePulse is the desktop application that makes this possible. It watches a folder on your computer, detects new photos the moment they arrive, sends them for AI processing, and saves the results to your output folder. Meanwhile, Live Feed displays each new generation on any screen in your studio.

Step 1: Configure Your Folders

FramePulse needs three folders:

• Watch Folder — Where new photos land. Point your camera's tethering software (Canon EOS Utility, Nikon Camera Control, Sony Imaging Edge, or Capture One) to save directly into this folder.
• Output Folder — Where processed results are saved. This is where you'll find the final images with replaced backgrounds.
• Backgrounds Folder — Optional. Store reference background images here. When you select one in FramePulse, it's sent alongside your photo to guide the AI generation.

Open FramePulse settings and set all three paths. The app creates the folders if they don't exist.

Step 2: Create Your Prompt Templates

Prompts describe the background you want the AI to generate. Good prompts are specific and descriptive:

• "Professional studio with soft gray gradient backdrop and subtle rim lighting"
• "Autumn park scene with golden oak trees, soft bokeh, and warm afternoon sunlight"
• "Modern minimalist office with floor-to-ceiling windows overlooking a city skyline"

Create several templates for different scenarios. Set one as your default — every new photo will use it automatically. You can switch prompts between shots without stopping the workflow.

Step 3: Configure Generation Settings

In the Settings panel, configure:

• Output Quality — Choose between 2K and 3K resolution. Higher resolution takes slightly longer but produces sharper results.
• Output Format — JPEG for smaller files and faster delivery, PNG for lossless quality.
• API Region — Select the region closest to your location for the fastest processing times.

Step 4: Enable the File Watcher

Toggle the File Watcher on in FramePulse. The app now monitors your Watch Folder for new files. Supported formats include standard images (JPG, PNG, TIFF, WebP) and RAW files (Canon CR2, Nikon NEF, Sony ARW, Adobe DNG, and more). RAW files are automatically converted on the server — no extra steps needed on your end.

Step 5: Open Live Feed

Live Feed is a real-time display that shows each new generation as it completes. Open it from FramePulse or from the navigation bar on the web app. You can view it on:

• A second monitor at your workstation
• A TV or large display in your studio for client presentations
• A tablet or phone for mobile preview

Live Feed supports fullscreen mode, thumbnail navigation, and automatic advancement to the latest image. Share the link with anyone — it works on any device with a browser.

Step 6: Start Shooting

With everything configured, start your photoshoot. Each photo follows this path:

1. Camera captures the image
2. Tethering software saves it to the Watch Folder
3. FramePulse detects the new file instantly
4. The photo is sent to the server with your prompt and optional background reference
5. AI generates the new background
6. The result is saved to your Output Folder
7. Live Feed updates with the new image

The entire process takes about a minute per image. Your client sees the result on the studio display shortly after you shoot.

Troubleshooting Common Issues

Photos not detected? Make sure the File Watcher is enabled and the Watch Folder path is correct. Check that your tethering software is saving to the exact folder FramePulse is monitoring.

Slow processing? Switch to a closer API region in settings. Reduce output quality to 2K for faster turnaround during the shoot — you can always reprocess select images at higher quality later.

Inconsistent results? Lock your camera exposure and white balance. Consistent input produces consistent AI output. Also check that your prompt matches your actual lighting setup.