Convert FFF to JXL

Why Convert FFF to JXL?

Converting Hasselblad FFF RAW files to JXL creates the optimal balance between preserving medium format image quality and achieving practical file sizes for delivery and archival. FFF files from medium format sensors can exceed 200 MB each — enormous files that are impractical for client delivery, web display, or long-term storage at scale. JXL's exceptional compression reduces these to 20-50 MB lossless or 5-15 MB lossy, while maintaining quality that does justice to the medium format sensor's resolving power.

JXL is uniquely suited for medium format photography because of its support for high bit-depth (up to 32-bit float), HDR, and wide color gamuts. Medium format sensors from Hasselblad capture 14-16 bit per channel data with exceptional dynamic range — detail that JPEG (8-bit, sRGB) simply cannot represent. JXL preserves the full tonal range, the Hasselblad HNCS color rendering, and the subtle gradations that make medium format photographs distinctive, in a format that is universally viewable.

For commercial studios that produce thousands of medium format captures per project, the storage implications are transformative. A fashion shoot might generate 2,000 FFF files totaling 300+ GB. After processing, the JXL lossless archive might be 60 GB — an 80% reduction. For advertising agencies and stock libraries maintaining millions of medium format images, JXL enables affordable long-term storage without compromising the quality their clients demand.

The conversion process involves demosaicing the FFF RAW data (ideally in Hasselblad Phocus or Capture One for optimal color rendering), then encoding the processed output as JXL. Apply your desired processing — HNCS color profiles, exposure adjustments, sharpening — before conversion. The JXL output is your finalized, delivery-ready image. Always preserve the original FFF files for potential future re-processing.

Key Benefits of Converting FFF to JXL:

• Massive Size Reduction: 80-90% smaller than TIFF exports from FFF
• Medium Format Fidelity: 32-bit float preserves every tonal nuance
• Client Delivery: Practical file sizes for high-resolution handoff
• Web Portfolio: Progressive loading for high-resolution galleries
• HDR Preservation: Full dynamic range and wide gamut maintained
• Universal Access: View without Hasselblad-specific software
• ISO Standard: Future-proof format for long-term archival

Practical Examples

Example 1: Fashion Campaign Client Delivery

Scenario: A fashion photographer shot 1,500 images on a Hasselblad H6D-100c for a luxury brand campaign. The client needs high-resolution processed files for print and digital use, but FFF files are too large and require specialized software.

Source: campaign_look12_047.fff (180 MB, 11600x8700px, 16-bit)
Conversion: FFF → JXL (lossless, 16-bit output)
Result: campaign_look12_047.jxl (38 MB, 11600x8700px, lossless)

Client delivery:
1. Process selects in Capture One with HNCS profiles
2. Export 16-bit and encode to JXL lossless
3. Deliver via file sharing service
✓ Client receives 38 MB instead of 180 MB per image
✓ 100-megapixel detail preserved for billboard printing
✓ No specialized software needed to view on client side
✓ 1,500 images: 270 GB → 57 GB delivery package

Example 2: Museum Art Reproduction Archive

Scenario: A museum uses Hasselblad medium format cameras for high-fidelity reproductions of paintings. The FFF originals need to be archived in a standardized format that will be accessible for decades.

Source: rembrandt_nightwatch_detail_003.fff (210 MB, 11600x8700, 16-bit)
Conversion: FFF → JXL (lossless, 16-bit, ICC embedded)
Result: rembrandt_nightwatch_detail_003.jxl (45 MB, lossless)

Museum archive:
✓ ISO-standard format for institutional permanence
✓ Color-managed with embedded ICC profiles
✓ 78% storage reduction per reproduction
✓ Brushstroke-level detail preserved losslessly
✓ Accessible without proprietary Hasselblad software
✓ 50,000 reproductions: 10.5 TB → 2.25 TB archive

Example 3: Architectural Photography Web Portfolio

Scenario: An architectural photographer showcases medium format work online. The FFF-sourced images need to load quickly in web galleries while demonstrating the resolution and quality that justifies medium format.

Source: skyscraper_twilight_pano.fff (195 MB, 11600x8700, 16-bit)
Conversion: FFF → JXL (lossy, quality 94)
Result: skyscraper_twilight_pano.jxl (4.2 MB, 11600x8700)

Web portfolio:
✓ 100MP image loads progressively in seconds
✓ Visitors can zoom into architectural details
✓ 4.2 MB manageable for mobile data connections
✓ Quality demonstrates medium format capabilities
✓ Progressive decode shows image outline immediately

Frequently Asked Questions (FAQ)

Q: What is the difference between FFF and 3FR?

A: Both are Hasselblad RAW formats. FFF (Flexible File Format) originated with Imacon digital backs and was used by early Hasselblad digital systems. 3FR is the newer Hasselblad RAW format used by current camera models like the X and H series. Both store unprocessed sensor data, but 3FR uses a different container structure. The conversion to JXL works the same way for both formats.

Q: Will the Hasselblad HNCS color science be preserved?

A: The color rendering from Hasselblad's Natural Colour Solution is applied during RAW processing and is baked into the converted JXL image. Process your FFF files in Phocus or Capture One with HNCS profiles for the most authentic color reproduction. JXL's high bit-depth and wide gamut support ensures no color banding or precision loss in the conversion.

Q: How long does it take to convert large FFF files?

A: FFF files from 100-megapixel sensors are 150-200+ MB each. RAW demosaicing takes 5-15 seconds per image, and JXL encoding at lossless quality adds another 10-30 seconds depending on effort level and hardware. Total conversion time is typically 20-45 seconds per 100MP image. Batch processing overnight is practical for large projects.

Q: Can JXL handle 100-megapixel images from medium format sensors?

A: Yes. JXL supports images up to 1 billion x 1 billion pixels — far beyond any current camera sensor. Its progressive decoding is particularly valuable for 100MP images, as viewers can see a preview within milliseconds while the full resolution loads. JXL's compression is also more efficient at larger resolutions, where spatial redundancy increases.

Q: Is JXL quality sufficient for large-format gallery prints?

A: In lossless mode, JXL is pixel-identical to the source — it is mathematically perfect for print production at any size. Even JXL lossy at quality 95+ is indistinguishable from lossless in print, as the lossy artifacts are below the threshold of visibility at any viewing distance. JXL with ICC profiles provides print-ready files that any professional print service can work with.

Q: Should I use 8-bit or 16-bit JXL for FFF conversions?

A: For archival and print production, use 16-bit to capture the full dynamic range of the medium format sensor. For web delivery and client proofing, 8-bit JXL at quality 90-95 produces excellent results at much smaller file sizes. JXL's superior compression means 16-bit lossless JXL is often smaller than 8-bit PNG, so there is less reason to compromise on bit depth than with older formats.

Q: Can I convert FFF files without Hasselblad Phocus software?

A: Yes. Our converter uses rawpy/LibRaw to process FFF files, which does not require Hasselblad's proprietary Phocus software. However, for the most accurate color reproduction matching Hasselblad's HNCS profiles, processing in Phocus (free download from Hasselblad) or Capture One before conversion is recommended. The generic rawpy processing produces excellent results for most purposes.

Q: How does JXL compare to HEIF for medium format output?

A: JXL offers better compression than HEIF at equivalent quality, particularly for lossless content. JXL also supports true lossless compression (HEIF lossless is less efficient), progressive decoding (HEIF does not), and is an open ISO standard without patent concerns. For medium format photography where quality is paramount, JXL is the technically superior choice for output delivery format.