> ## Content Index
> Fetch the complete content index at: https://blog.tonalphoto.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# Convert to FFF Changes the Raw Data, and Lightroom Renders It Differently
- URL: https://blog.tonalphoto.com/3fr-to-fff-conversion-raw-data-lightroom/
- Published: 2026-11-21T10:25:00.000Z
- Updated: 2026-11-21T10:24:59.000Z
- Author: Konrad Michels
- Tags: Phocus

If you convert your 3FRs to FFF when you import into Phocus and later open those FFFs in Lightroom Classic, Lightroom is not working from the data it would have got from the 3FR. It reads a different "As Shot" white balance from the FFF, its "Auto" button chooses different settings, and the same settings give a measurably different picture. The conversion is one-way, so if the FFF is the only copy you kept, there is no way back to the file the camera wrote.

A reader on Reddit mentioned this week that they convert to FFF when they import into Phocus. That surprised me, because the reader was a new Hasselblad owner/Phocus user, and Phocus by default imports as 3FR unless explicitly set to import as FFF: the "RAW File" popup in the Import dialog starts on "Keep 3FR". My guess is that they ran across one of the FFF versus 3FR debates somewhere on the interwebs and came away thinking FFF is the better choice. I suspect a lot of people who import as FFF do it because the files are smaller.

And it does save space. Converted from my own X2D II files, an FFF comes out 25% to 47% smaller than the 3FR it was made from. What the Import dialog does not tell you is how this space saving is achieved. The FFF stores the raw data as a lossless JPEG, where the 3FR stores it uncompressed, and that step loses nothing. But the conversion also rescales the raw data before compressing it.

If the space saving is what you are after, compressing the 3FR with xz and a delta filter gets you a smaller file than the FFF, and it decompresses to exactly what the camera wrote.

💡

****A note on support:** This post represents my personal exploration and testing, not official technical support or guidance from Hasselblad. If you need assistance with your Hasselblad equipment, please contact Hasselblad directly: [customersupport@hasselblad.com](mailto:customersupport@hasselblad.com) for global support, [support.us@hasselblad.com](mailto:support.us@hasselblad.com) for the Americas, or visit hasselblad.com/support for regional options.

> **Key finding:** On my Hasselblad X2D II, converting a 3FR to FFF in Phocus for Mac 4.2.3 changed the raw data in all five frames, and Phocus cannot convert it back. Lightroom Classic 16.0 reads a different white balance from the FFF, renders it measurably differently and picks different "Auto" settings. To save space, compress the 3FR instead.

|                                 | 3FR from the camera    | FFF converted by Phocus 4.2.3 |
| ------------------------------- | ---------------------- | ----------------------------- |
| Raw data                        | as recorded            | green x 1.0015, blue x 1.0078 |
| File size, my five frames       | 211 to 213 MB          | 112 to 159 MB                 |
| Embedded preview                | 3888 x 2918 JPEG       | 971 x 728 thumbnails          |
| Phocus export, same adjustments | reference              | channel averages within 0.01% |
| Lightroom export, as imported   | reference              | averages differ by up to 0.7% |
| Lightroom export, same settings | reference              | blue 0.24% to 0.83% lower     |
| Lightroom Auto, ISO 400 frame   | Whites +36, Blacks -33 | Whites +41, Blacks -35        |
| Way back to the other format    | none needed            | none in Phocus                |

## What does Phocus for Mac's Convert 3FR to FFF change in the raw data?

Phocus 4.2.3 for Mac rescales two of the three color channels. Red comes through untouched. Every green value is multiplied by 1.0015 and every blue value by 1.0078, with a small fixed amount taken off each. The result is then stored with lossless compression, where the camera's 3FR stores its raw data uncompressed.

I took five 3FRs from my X2D II, one each at ISO 50, 100, 400, 1600 and 6400, imported them into Phocus 4.2.3 with "RAW File" set to "Convert 3FR to FFF", and compared the raw values in each pair. The multipliers were the same on all five frames. Raw values are whole numbers, so the rescaled ones are rounded, and that is why the original values cannot be recovered exactly from an FFF.

![Diagram of one raw file converted from 3FR to FFF in Phocus 4.2.3, with three result cards: the raw data is changed (red untouched, green multiplied by 1.0015, blue by 1.0078), Phocus 4.2.3 exports the same picture from both files, and Lightroom Classic 16.0 exports a different picture, with a different As Shot white balance, blue 0.2% to 0.8% lower and different Auto settings](https://storage.ghost.io/c/f5/ac/f5aca107-db6c-464b-b8b7-7b834c3c955f/content/images/2026/11/fig1-what-convert-to-fff-does.png)

Figure 1\. The same frame as a 3FR and as its FFF, in the three places I compared them.

In July a reader passed on a comment from a RawDigger developer that FFF files are normalized from the 3FR, and pointed me to a Hasselblad webinar on Reproduction mode in which, as the reader described it, a pipeline slide lists the camera's 3FR and the Phocus FFF as separate stages.¹ I measured one pair that month, converted by Phocus 4.2.1, and got the same multipliers. What this new post adds is the size of that normalization on one camera, across five frames and a newer Phocus.

It is not clear what the scaling is for.

## Phocus renders the two files the same

Give a 3FR and its FFF the same adjustments in Phocus 4.2.3, export both, and the two TIFFs have the same color and brightness: on all five frames the red, green and blue averages of the FFF export are within 0.01% of the 3FR export.

For this I applied the "Standard" preset to each of the ten files and exported 16-bit TIFFs in Adobe RGB. "Standard" is not the state a 3FR opens in. "Standard" pushes Vignetting from 50 to 100, which is why it went onto both files of each pair and not just one. As a control I exported one 3FR twice, and the two TIFFs were byte for byte the same, so Phocus does not vary from one export to the next.

The pairs are not identical pixel for pixel. On four of the five frames nearly every sample differs, typically by 11 to 61 on a scale of 0 to 65,535, which is a fraction of how much neighboring pixels in the same frame differ from each other. The ISO 100 pair is the exception, at 97.9% identical. I have no explanation for that one yet.

So Phocus either applies the same scaling to a 3FR when it renders it, or allows for the scaling when it reads an FFF. From the outside it is impossible to tell which. If Phocus is where your files stay and are exported from, the folks who say an FFF gives you the same output have these numbers on their side.

Researching and writing the blog and page posts I produce takes up a lot of my time and costs me money for software subscriptions, tooling, and other incidentals related to running a business.

Please consider a paid membership to help me keep the lights on and pay the bills - it is only a fraction of the cost of an X2D II and a lens. Click on the button to go to the membership page to see the options available.

[Support the blog. ](https://blog.tonalphoto.com/membership/) 

## Lightroom Classic renders them differently

Lightroom Classic 16.0, however, does not treat the pair as the same file. It reads a different "As Shot" white balance from the FFF, and when every develop setting is forced to match it still renders the FFF with less blue and more red.

I added all ten files to a catalog and changed nothing. Lightroom gave every one its "Camera Standard" profile, and under it every 3FR came in at "As Shot" 5500 / +10 and every FFF at "As Shot" 5450 / +11\. Under "Adobe Standard" the temperatures matched at 5100 and the tint differed by one.

Then I set all ten to the "Daylight" white balance, so that every develop setting matched within each pair, and exported 16-bit TIFFs. Change in the FFF export against the 3FR export:

| Frame    | ISO  | Camera Standard: red | green  | blue    | Adobe Standard: red | green  | blue    |
| -------- | ---- | -------------------- | ------ | ------- | ------------------- | ------ | ------- |
| B0031524 | 50   | +0.21%               | +0.11% | \-0.83% | +0.53%              | +0.44% | \-0.24% |
| B0021315 | 100  | +0.07%               | +0.01% | \-0.46% | +0.07%              | +0.02% | \-0.51% |
| B0021134 | 400  | +0.14%               | +0.07% | \-0.48% | +0.13%              | +0.07% | \-0.48% |
| B0031107 | 1600 | +0.16%               | +0.27% | \-0.45% | +0.63%              | +0.16% | \-0.43% |
| B0021255 | 6400 | +0.12%               | +0.08% | \-0.33% | +0.12%              | +0.06% | \-0.35% |

Blue is lower in the FFF export on every frame under both profiles, and red is higher on every frame.

![Two bar charts, one for the Camera Standard profile and one for Adobe Standard, showing for five frames from ISO 50 to ISO 6400 how the red and blue channel averages of the FFF export differ from the 3FR export in Lightroom Classic 16.0 with identical settings. Red is 0.07% to 0.63% higher and blue 0.24% to 0.83% lower in every case](https://storage.ghost.io/c/f5/ac/f5aca107-db6c-464b-b8b7-7b834c3c955f/content/images/2026/11/fig2-lightroom-shift-by-frame.png)

Figure 2\. Both files at "Daylight" with every develop setting matched. Green, not drawn, is also higher in the FFF export on every frame.

The direction is odd. The FFF's raw data has more blue than the 3FR's, yet Lightroom renders it with less. So Lightroom is not simply passing the rescaled numbers through. These tests do not show why its two renders differ.

Pull the two files straight in and change nothing, and the renders differ as well. Each file is then at its own "As Shot" white balance:

| Frame    | ISO  | Camera Standard: red | green   | blue   | Adobe Standard: red | green   | blue    |
| -------- | ---- | -------------------- | ------- | ------ | ------------------- | ------- | ------- |
| B0031524 | 50   | \-0.06%              | \-0.07% | +0.04% | +0.65%              | +0.34%  | \-0.01% |
| B0021315 | 100  | \-0.04%              | \-0.06% | +0.13% | +0.13%              | \-0.03% | \-0.20% |
| B0021134 | 400  | \-0.07%              | \-0.07% | +0.20% | +0.21%              | \-0.03% | \-0.18% |
| B0031107 | 1600 | \-0.04%              | \-0.07% | +0.19% | +0.68%              | +0.02%  | \-0.17% |
| B0021255 | 6400 | \-0.06%              | \-0.05% | +0.12% | +0.20%              | \-0.01% | \-0.15% |

Under "Camera Standard" the different "As Shot" reading turns the blue shift around: the FFF now has slightly more blue, 0.04% to 0.20%, and slightly less red and green. Under "Adobe Standard" the as-imported renders look much like the matched ones, with more red and less blue in the FFF.

### "Auto" picks different settings for the two files

Press "Auto" in the Basic panel on each file and Lightroom lands on different sliders. I did this by hand on the ISO 400 frame in Lightroom Classic 16.0.

| Slider     | 3FR   | FFF, as imported | FFF, both files at "Daylight" |
| ---------- | ----- | ---------------- | ----------------------------- |
| Exposure   | +0.70 | +0.71            | +0.69                         |
| Highlights | \-62  | \-62             | \-60                          |
| Whites     | +36   | +41              | +36                           |
| Blacks     | \-33  | \-35             | \-32                          |

Contrast, Shadows, Vibrance and Saturation came out the same every time, at +6, +72, +20 and +4\. The 3FR got the same values in both states, which makes sense: its "As Shot" reading is already 5500 / +10, the same as "Daylight". I pressed "Auto" a second time to check it does not wander, and it gave the same numbers.

"Auto" works from the picture, so if it chooses differently, it is not unreasonable to assume that Lightroom is not seeing the same picture. As imported, the FFF gets five more points of Whites and two fewer of Blacks. With the white balance matched the gap shrinks to a point or two on Exposure, Highlights and Blacks, but it does not close. In that state, both at "Daylight" with "Auto" applied, I flicked between the two files and could watch the blue highlights in the histogram move. With both files reset and no "Auto", I could not see the histogram move at all.

These are small moves. Put the other way round: the same settings give two slightly different pictures, so getting the same picture out of both files would take two slightly different sets of settings.

Have you seen the guide? I've published Essential Phocus 4.x for Mac - 85 topics across 8 sections and 251 pages covering everything from HNCS color science to HDR workflows. It's the reference manual Hasselblad hasn't updated since 3.8\. It's $49, and updates are included.

[Get it here ](https://payhip.com/b/6pqAt?ref=blog.tonalphoto.com) 

### Can you see it?

Only just. In a blind test I picked the difference out in two of six pairs, which puts it at the very limit for my eyes.

From the ISO 400 frame I cut the same two 2400 x 1600 areas, one of water and one of sky, out of each exported TIFF, at "As Shot" and at "Daylight". Those went into ten pairs with coded names: six were a 3FR render against an FFF render, and four were the same file twice. I flicked between the two files of each pair in Quick Look without knowing which was which.

On the four identical pairs I called no color difference. On the six real pairs I called a difference twice, one file "a touch warmer" each time, and both calls went the way the numbers say they should. On the other four I could not tell the files apart, and one of those four was a pair I had called correctly, shown in the opposite order.

Measured, the differences in those crops are 0.14 to 0.40 dE2000\. That is the score for how different two colors look that I used in the [Lightroom camera profile post](https://blog.tonalphoto.com/lightroom-camera-profile-phocus-color/), where about 1 is the edge of what you can see side by side. And this was a flick test, with the two versions swapping in the same spot on the screen.

## Saving the space without converting

If you're using FFF to save space, a 3FR compressed with xz and a delta filter comes out smaller than the FFF, on all five of these frames, and decompresses to exactly the file the camera wrote.

Given what storage costs at the moment, I understand why a quarter or more off every file is attractive, and next to an uncompressed 3FR the FFF may well be the better choice on that count. But for keeping the data the camera recorded, I'd rather compress the 3FRs I've archived.

| Frame    | ISO  | 3FR      | FFF      | 3FR with ZIP-style compression | 3FR with xz and a delta filter |
| -------- | ---- | -------- | -------- | ------------------------------ | ------------------------------ |
| B0031524 | 50   | 211.0 MB | 112.4 MB | 131.8 MB                       | 93.8 MB                        |
| B0021315 | 100  | 211.2 MB | 133.9 MB | 172.8 MB                       | 126.4 MB                       |
| B0021134 | 400  | 212.4 MB | 151.5 MB | 178.3 MB                       | 147.0 MB                       |
| B0031107 | 1600 | 211.9 MB | 158.9 MB | 180.7 MB                       | 151.1 MB                       |
| B0021255 | 6400 | 212.7 MB | 155.9 MB | 180.6 MB                       | 149.8 MB                       |

![Horizontal bar chart of file sizes for five X2D II frames from ISO 50 to ISO 6400. For each frame, the 3FR from the camera is 211 to 213 MB, the FFF from Phocus 4.2.3 is 112 to 159 MB, and the 3FR compressed with xz and a delta filter is 94 to 151 MB, smaller than the FFF every time](https://storage.ghost.io/c/f5/ac/f5aca107-db6c-464b-b8b7-7b834c3c955f/content/images/2026/11/fig3-file-sizes.png)

Figure 3\. The same five frames three ways. The xz file decompresses to the identical 3FR.

Ordinary ZIP compression does poorly here, 15% to 18% on four of the five, because it is not built for a file that is one long run of 16-bit sensor values. xz is a stronger general-purpose compressor. The delta filter is what gets it past the FFF: before compressing, it replaces each value with its difference from the value of the same color two photosites earlier, and those differences are small numbers that compress well. Without the filter, xz was still larger than the FFF on three of the five frames.

The command, in Terminal:

```
xz -k -T0 --delta=dist=4 --lzma2=preset=6 B0031107.3FR
```

That writes `B0031107.3FR.xz` next to the original and, because of `-k`, leaves the original in place. Before you delete an original, check the archive against it:

```
xz -dc B0031107.3FR.xz | cmp - B0031107.3FR
```

No output means the two are identical. To get a 3FR back later:

```
xz -dk B0031107.3FR.xz
```

That one refuses to run if a file called `B0031107.3FR` is already in the folder, so it will not overwrite anything.

xz is not part of macOS. Homebrew has it (`brew install xz`). On an M4 Pro Mac mini each file took 12 to 17 seconds to compress, and the restore I timed took 0.7 seconds. The restored file matched the original byte for byte.

The cost is that nothing opens a `.xz`. Phocus, Lightroom and Finder previews all need the file decompressed first, so this is for the archive, not for the folder you are working in.

## What else the conversion changes

Checked on all five of the 4.2.3 conversions to FFF:

- **The embedded preview shrinks.** A 3FR carries a 3888 x 2918 JPEG written by the camera. The FFF has 971 x 728 thumbnails in its place. Anything that shows the embedded preview instead of rendering the raw has much less to work with.
- **A color calibration tag is gone.** The 3FR has a ColorMatrix1 tag and the FFF does not. Phocus and Lightroom bring their own color for this camera, so I would not expect either to miss it, but I have not tested software that reads it.
- **The camera name in the file changes**, from `X2D II 100C` to `Hasselblad X2D II 100C`. It is the same change a Phocus 4.2 import made to 3FRs, after which [DxO PureRaw rejected them](https://blog.tonalphoto.com/dxo-pureraw-3fr-phocus-4-2-import/).

The conversion from 3FR to FFF is also one-way. Phocus has no Convert FFF to 3FR, and [renaming an FFF does not turn it back into a 3FR](https://blog.tonalphoto.com/hasselblad-fff-vs-3fr-why-renaming-doesnt-work-and-why-thats-good/). If the FFF is the only copy you kept, the rescaled data is the data you have.

Tethered shooting writes FFF directly, with no 3FR at all. In July I checked one tethered FFF from this camera and it showed the same scaling as a converted one.

## What I have not tested

- **Capture One.** It opens 3FRs from the X2D II natively. I have not run the pair through it.
- **A second camera body.** Everything here is one X2D II.
- **Other Adobe profiles.** "Camera Standard" and "Adobe Standard" only.
- **A tethered FFF in Lightroom.**

I also can't explain the ISO 100 pair that came out 97.9% identical in Phocus. ISO 100 is the last step before the [X2D II's readout gain changes at ISO 200](https://blog.tonalphoto.com/iso-invariance-hasselblad-x2d-ii/), which may or may not have anything to do with it. With one frame per ISO, each of a different scene, I can't separate the ISO from the picture. One scene shot at ISO 50, 100 and 200 would.

The method, the tools and the per-frame data behind every number here are on a separate page: [3FR to FFF Conversion in Phocus: Methods and Data](https://blog.tonalphoto.com/3fr-to-fff-conversion-methods-data/).

## Sign up for The Tech Behind the Frame - Hasselblad edition

Practical Hasselblad knowledge the official docs don't cover. Workflows, file behaviour, and real-world insights for Phocus and X2D II users.

Subscribe 

Email sent! Check your inbox to complete your signup. 

No spam. Unsubscribe anytime.

## References

1. [Phocus Reproduction Mode Walkthrough, Hasselblad webinar](https://www.youtube.com/watch?v=lZthFo98HTk&t=1242s&ref=blog.tonalphoto.com), pipeline slide at about 20:42