Using Reference Images

How to match the color grading of any photo using AI analysis.

Reference image matching is available on paid plans only. View plans

What is Reference Image Matching?

Reference image matching lets you upload a photo whose look you love, and Hyperfocal's AI will analyze its color grading, tones, and mood to create a preset that recreates that aesthetic. It's like having a colorist study a photo and replicate its style for your images.

How to Use Reference Images

  1. In Hyperfocal, click "Upload reference"
  2. Upload your own photo that you want to apply the look to
  3. Upload a reference photo with the look you want to match
  4. Click Analyze & Generate
  5. The AI analyzes the reference image and creates a matching preset
Reference image upload interface in Hyperfocal
Upload any photo to match its color grading

What the AI Analyzes

When you upload a reference image, our AI examines:

  • Color temperature: Overall warmth or coolness of the image
  • Color grading: Tints in shadows, midtones, and highlights
  • Contrast & tonal range: How shadows and highlights are treated
  • Saturation levels: Overall vibrancy and individual color intensity
  • Color shifts: How specific colors (greens, blues, skin tones) are modified
  • Special characteristics: Fade effects, grain, vignettes, etc.

Best Practices for Reference Images

Choose Images With Clear Color Grading

The best reference images have a distinctive, intentional look. Photos that have been edited with a specific style will produce better results than unedited photos or those with subtle adjustments.

Good Reference Images

  • Photos with obvious color grading
  • Film photography or film simulations
  • Editorial/magazine-style photos
  • Photos with consistent lighting
  • Well-exposed images

Less Ideal References

  • Heavily composited images
  • Photos with extreme local adjustments
  • Very dark or overexposed images
  • Photos with strong HDR effects
  • Screenshots or low-quality images

Consider the Subject Matter

The AI works best when the reference image has similar content to what you'll be editing. A portrait reference will produce a preset optimized for skin tones, while a landscape reference will emphasize sky and nature colors.

Pro Tip

If you're trying to match a look from a movie or TV show, use a frame with good variety in tones (not too dark, not too bright) and clear color grading. Pause on a well-lit scene with skin tones visible for best results.

Privacy & Image Handling

Your reference images are processed in real-time and never stored on our servers. They're analyzed by the AI, used to generate your preset, and immediately discarded. We take your privacy seriously; your images remain yours.

When Reference Matching Works Best

Reference image matching excels in these scenarios:

  • Recreating film looks: Match the color science of your favorite film stocks
  • Matching a photographer's style: Analyze their edited work to understand their aesthetic
  • Consistent project editing: Create a preset from one edited photo to apply to a whole series
  • Inspiration from cinema: Recreate the color grading of movies and TV shows

Limitations to Know

While powerful, reference matching has some limitations:

  • Local adjustments can't be replicated: Presets apply globally; if the reference has localized edits (like a radial filter on just the subject), those won't transfer
  • Lighting differences matter: A preset made from a sunset photo may look different on a midday photo
  • Some effects are camera-specific: Lens characteristics, sensor color science, etc. can't be fully replicated

The generated preset is always a starting point. Use Tweaking & Customizing to fine-tune for your specific photos.

Related Guides

Still have questions?

Check out our FAQ or reach out to us.