Still here? Okay. Let's get started.

In my last post, I described where standard RAW developers like Lightroom hit their limits when it comes to infrared photography. The takeaway was that automated processes optimized for normal photography get completely thrown off by the quirks of an IR image.

The first step of any image processing is what we call demosaicing. Even if you shoot straight to JPEG, this step still happens first, just directly inside the camera instead.

What is demosaicing anyway? The art of puzzling

A RAW file is not a finished image. You have to imagine it more like an incomplete mosaic. Every pixel on the camera sensor sits under a color filter - usually Red, Green, or Blue (the so-called Bayer filter). So a pixel under a red filter only knows the brightness value for red, and has absolutely no clue about green or blue. Demosaicing is the process where the software solves this puzzle. The algorithm looks at a pixel and its neighbors, then makes an educated guess to calculate the missing color values for every single pixel. Only after that do we get an actual color image.

How the software solves this puzzle depends on the motif, the situation, and the camera. It is less of a fixed mathematical equation and more of a strategic choice. This is where the wheat gets separated from the chaff. There are a bunch of different ways to solve this problem, and since most people want to deal with photos rather than algorithms, programs like RawTherapee have smart processes to choose for us. But they also let us make the choice ourselves - which can actually be better, since we as photographers usually know best what we want to focus on.

Here is a screenshot from RawTherapee showing a 100% crop of one of my RAW images with demosaicing deactivated. You can clearly see the structure of the sensor:

The exact same image, same crop. Only difference: the algorithm.

Broadly speaking, you can divide most algorithms into two philosophies:

1. The "detail-first" philosophy (The Optimists)

These algorithms assume that the image data is clean and that even the tiniest deviation between pixels represents a real detail. They analyze edges and patterns aggressively to reconstruct the sharpest possible image.

• Examples of these algorithms: AMaZE, RCD, DCB
• Strengths: They deliver perfect sharpness with clean, low-ISO images. Ideal for fabrics, architecture, or fine structures.
• Weaknesses: They are easily fooled. Noise looks like a fine pattern to them - and gets promptly "amplified". This leads to things called labyrinth artifacts. You can get false colors (color fringing) on hard edges, or moiré (rainbow patterns) on fine textures.

2. The "artifact-avoidance" philosophy (The Cautious)

These algorithms are more conservative. Their main goal is to avoid making mistakes, even if that means losing a tiny bit of sharpness. They assume the data might be noisy and put more value on smooth surfaces and clean colors.

• Examples: LMMSE, IGV, VNG4
• Strengths: They are ideal for high-ISO images or noisy data. Labyrinth artifacts or moiré are barely an issue here.
• Weaknesses: With an absolutely perfect image, they can look a bit "softer" in a direct comparison with their aggressive colleagues.

Useful parameters to know

Besides choosing the algorithm, there are a few sliders you can use to fine-tune your strategy. The following refers to RawTherapee, but you can find similar options in other tools too.

• Dual Demosaic (e.g., AMaZE+VNG4): A brilliant hybrid approach. You can use a "detail algorithm" for the sharp areas of the image and a "cautious algorithm" for the smooth surfaces (like the sky). The contrast threshold slider defines where that boundary lies.
• False Color Suppression: This tool applies a filter that affects only the color information, not the sharpness. It smooths out those colorful splotches that often appear along edges.
• Border: The pixels at the very edge of the sensor can't be perfectly reconstructed because they lack neighbors. Normally, RAW converters just crop them out. With this slider, you can force the software to try anyway - but at your own risk, as this can create artifacts.

With this knowledge, we can now choose the right strategy for our images on purpose. For standard shots, modern RAW developers are already pretty good at doing this for us. Low ISO shot? => AMaZE. High ISO? One of the others... For photographers, knowing this stuff becomes valuable when you deviate from the standard case. Shot at high ISO by accident? A special light spectrum or heavy filter use that creates a scenario the software can't guess? Choosing the algorithm yourself can really help!

One such scenario is infrared photography, so here is an explanation of what that looks like in practice.

Filter Group 1: "Super Color" / Low Wavelength (550nm, 585nm, 625nm)

• The Challenge: Here, a wild mix of strong infrared light and a portion of the visible spectrum hits the sensor. The result is very intense false colors and a high risk of artifacts. However, we still have usable detail information across several channels.
• The Strategy: We need a compromise. A "detail-first" algorithm is good, but it has to keep false colors under control.
• The Choice: I almost always start with DCB here. It's just as sharp as AMaZE, but it's better at handling false colors, especially for cameras without an AA filter (like my RP). An alternative is a Dual Demosaic approach to treat the sky and foliage separately. To help things along, I set the false color suppression slider to 2 or 3.

Filter Group 2: "Classic IR" / Mid Wavelength (720nm)

• The Challenge: The signal is now almost pure infrared. The red channel is the undisputed king, while the green and blue channels are weak and extremely noisy. The ISO 100 in your metadata is a flat-out lie - for G and B, the image is closer to ISO 12,800.
• The Strategy: We have to force the software to treat the image like a high-ISO shot. An "artifact-avoidance" strategy is mandatory so we don't interpret and amplify the noise as detail.
• The Choice: For false-color images, LMMSE or IGV are the top choices here. They handle the noisy channels gently and prevent those dreaded labyrinth patterns. With a 720nm filter, though, you might also just decide to go straight to a black-and-white conversion. Which brings us to our next option...

Filter Group 3: "Deep IR" / High Wavelength (850nm) & Black-and-White Conversion

• The Challenge: There is virtually no color information left here. The G and B channels are basically pure noise. Traditional demosaicing would be pointless and counterproductive here.
• The Strategy: We should focus on the strongest, cleanest signal we have and ignore the rest.
• The Choice: This is the perfect use case for the Mono algorithm. It is the "anti-demosaicer". It ignores the broken G and B channels and builds the black-and-white image directly from the data of the clean red channel. The result is the maximum possible sharpness and tonality without any interference.

So, that's it for today. At some point, I'll extend this series to cover file formats and color spaces, but that's enough for now. Ultimately, for the pragmatic case, you can always just switch your camera to "Sport" mode and trust the tech 😉

Anyone getting into infrared photography quickly notices: the usual image editing tools hit their limits here. Lightroom in particular, my tool of choice for years, is really not built for IR. It gets in your way at the very first step - setting the white balance. IR images are very warm (red, basically), and to compensate you have to push the white balance slider dramatically to the left into the "cold" range. Lightroom unfortunately has a hard-coded lower bound that is "too warm".

Resourceful users found a workaround: camera profiles! Turns out you can create and register camera-specific profiles in Lightroom. Getting there is a bit fiddly though, especially if you're trying to sidestep Lightroom's built-in limits. The solution? A free DNG tool from 2012 (!!) by Adobe, originally meant for converting manufacturer RAW files to DNG (more info here). Open a sample file in the tool, drag the color temperature all the way down under "Color Matrices", export the whole thing as a camera profile, drop it into Lightroom's profile directory, then either pick it manually or configure it as the default when this specific camera shows up as the author of a RAW file. Convoluted, sure (here's a tutorial), but at least you can now set a proper white balance for these files. One important thing: NEVER trust the auto setting. It's hard-tuned in LR. Great for normal shots, not great for special cases.

And the pain doesn't stop there - a lot of IR photographers eventually want to swap color channels. Lightroom is not exactly intuitive for that, and at the latest at this point you'll want to move to a "big" image editor.

The next issue: "intelligent" software like Lightroom often produces noisier IR images than necessary. Why? Because neither the development algorithms (in particular the demosaicing) nor Lightroom's famous denoising fit our use case. You don't always spot it immediately, especially not if you look at a downsized image instead of at 100%. But in practice, even when the image looks clean at first glance, you have much less headroom for corrections and less detail than you could have. And especially when adjusting color channels you'll hit limits and clip channels faster than you'd like.

Where does this come from? Simple: if you point a RAW developer at IR photos, you're basically forcing it to get confused. Unlike "normal" images, the image information is not evenly distributed across the four filters of a Bayer sensor. Instead the red channel is loaded with information while the green and blue channels are weak and noisy. The job of a RAW developer is to compute the "right" colors from those signal values using a suitable algorithm. Especially with high-ISO images, i.e. lots of noise, the software picks different algorithms than for low ISO. An IR image, particularly one taken during the day, often comes in at a low ISO (100, 200) but still has enormous noise in two of the three channels. So if you want to squeeze out the best result you need to tell the software up front how to develop the RAW. Many tools only allow that in a very limited way. Denoising is very similar: modern software comes with several specially trained AI models and picks them based on image metadata. Automatic denoising in Lightroom or in the DxO tools I usually love is a "hit and miss" affair. You just can't trust it blindly. In many of my shots these tools actually make the noise worse if I don't watch out.

All of these problems can of course be worked around in an image editor. Pros use channel-specific luminosity masks and do denoising the "classic" way, without AI tools. That gets tedious fast, and at least I sometimes lose track of what I'm even trying to do. Image editing gets mixed up with image development, and the result suffers.

So, what to do? I've been at this for a while and by now I've found decent solutions for a lot of these things. What I find interesting in hindsight is the journey I went on - from "blindly click through a tutorial and learn what to do to get a certain look" to "dig deeper and deeper into sensors, file formats, base processing". Especially around IR there's a lot written on the internet that is only half-true, or where the explanations don't quite match the effect. Hence this blog post. I don't claim to be error-free, but maybe these notes help someone who is past the simple basics and just wants to do it properly for once.

One thing to know up front: I don't see myself as a particularly strong image editor. Once my RAWs are properly developed I'm more of a tinkerer than a targeted editor. Photoshop & co are part of my toolbox, but 80% of the functions in there feel esoteric and dangerous to me. But at least the front end of the processing chain I think I've now understood well enough, so here's my workflow, as of 2025:

My workflow for maximum quality (my tools of choice: RawTherapee & DxO)

After a lot of trial and error I've built a workflow that combines the strengths of different programs. The basic idea: use RawTherapee for the technical groundwork on the RAW file, and only then hand it off to my favorite tool, DxO PhotoLab, for optical corrections and creative optimization. Both tools can of course be swapped for alternatives. The role of DxO PhotoLab in particular is interchangeable. But after years of experimenting I do consider RawTherapee the most powerful option, simply because the tool is basically arbitrarily configurable when it comes to the base algorithms. Close behind: Darktable. Personally I don't love DT, but I have to give it that it also offers lots of knobs and levers that are often optimized away or lumped together in the big productivity tools. And last but not least: my IR workflow is different from my workflow for "normal" images. For those I just stay in DxO from start to finish, and before that I mostly lived in LR. But back to the IR workflow.

The goal is to produce, in RawTherapee, a technically perfect but neutral 16-bit TIFF - a "digital negative". Only then does the actual editing begin. Why? Because the point is to interpret the sensor data in a controlled way and then lock that interpretation in. TIFFs are painfully large files, just as a word of warning. But they are lossless and give you maximum image information for further work.

Phase 1: The technical foundation in RawTherapee

This is where you make all the decisions that automated software would get wrong because of IR's quirks. We deliberately override the software's wrong assumptions.

Demosaicing - the single most important step

Here we decide how an image gets computed from the raw sensor data. At 100% zoom the differences become obvious.

• For 720nm filters (classic IR): I usually start with LMMSE or IGV. These algorithms are optimized for high noise and prevent the typical "maze artifacts" that noise in the G and B channels can produce. If you're aiming for a pure black-and-white image, Mono is worth a try - it often gives the sharpest results because it focuses on the clean red channel.
• For 550nm filters ("Super Color"): For me DCB has proven a good starting point. My EOS RP doesn't have an anti-aliasing filter, and the documentation says DCB is good at avoiding false colors on such sensors.
• False color suppression: For 550nm shots this slider is worth its weight in gold. A value of 2 or 3 removes the typical pink/green blotches without hurting sharpness. For 720nm I leave it at 0 or 1.

White balance - manual control

Auto is essentially useless here, because infrared is invisible by definition. The images we're working with are a form of false color photography, and it's up to the photographer which effect to aim for. That said, you can of course click the auto button and depending on the tool something you like might come out. Just make sure you don't accidentally clip one of the channels. That's exactly why I set WB manually as a rule.

• Eyedropper tool: A click on healthy green foliage (which in IR reflects almost white) or on clouds is usually my first try. But again: you decide what color shift you want.
• Fine tuning: Nudge with the Temperature and Tint sliders.
• The power tool: If the image still has a color cast, RawTherapee has a special tool that feels like it was made for us: the Red/Blue equalizer. This slider directly controls the ratio of red to blue and can eliminate stubborn casts. Why is it made for us? Because in IR the red channel is overrepresented at the expense of blue. The white balance shifts values between red and blue back and forth, and the equalizer lets you calibrate that ratio.

3. Exposure & noise reduction

• Exposure: I set exposure and black point so that the histogram clips neither on the left nor on the right. Important: I leave highlight recovery off, because it's optimized for visible light and produces weird results in IR. I also expose to the right when shooting (ETTR): I deliberately expose so bright that the highlights just barely don't clip. The reason: in IR the shadow detail drowns in noise very quickly. The only cure is to capture more signal, i.e. ideally to expose longer. Cranking up ISO is a bad idea, by the way, because that amplifies the noise too. So you want the lowest possible ISO with the most possible signal. Which means: aperture wide open, or long exposure time.
• Noise reduction: Here I focus exclusively on color noise (Chroma). Luminance noise I don't touch. A manual pass on the Chroma slider very effectively kills the color blotches in the weak channels.
• What NOT to do in this phase: sharpening, contrast, saturation, or any other creative tools. The goal is a "flat", neutral, but technically clean image.

Phase 2: Export as a digital negative

Once everything fits, the image gets prepped for handoff.

• File format: TIFF
• Bit depth: 16-bit (absolutely essential for quality!)
• Color profile: A wide color space like ProPhoto RGB or Rec2020. This makes sure none of the unique IR colors get lost on export.

Phase 3: The fun part in DxO PhotoLab (or another tool)

The exported TIFF is now a "well-behaved", clean image. The problematic IR artifacts have already been handled manually and correctly. Now DxO can play to its strengths without getting confused.

• Optical corrections: Apply DxO's first-class corrections for distortion, vignetting, etc.
• Contrast & clarity: Tools like ClearView Plus and the microcontrast slider now work beautifully.
• Creative work: If you want to swap channels for the "false color look", rotate the slider in the HSL tool by 180 degrees, or pass the TIFF over to Photoshop, Affinity or GIMP. Go wild.
• If some noise is still left: you'll notice right away - working from TIFF, all those almost magical AI denoising tools disappear, because they mostly work on RAW data and need it. But there are exceptions. The "classical" tools are still available of course, and some AI tools work too. Topaz Denoise AI, for example, handles TIFFs just fine, and many of the Photoshop automatics are still available.

This workflow may look convoluted at first, but it makes sure every decision is made by the tool best suited to make it. The result is visibly higher image quality and more flexibility for the creative side. It also optimizes for marginal utility. If you want to be pragmatic: take a look at the automatics and keep an eye on noise levels. Depending on the subject it may just not matter and a simple round with the auto sliders is enough. My final recommendation: if you shoot IR, seriously look at other RAW tools, even if your heart is with Lightroom. Even if you don't want to pick algorithms yourself, just the handling of white balance and IR-specific color shifts is often solved much better elsewhere. Tools like RawTherapee and Darktable are also free, so at least it doesn't hurt your wallet.

More reading?

• Demosaicing article on RawPedia
• Custom DNG Tutorial for setting a white balance for infrared photography
• Bayer sensor (Wikipedia)
• Advanced Infrared Photography Processing: The Power of Luminosity Masks

How do white balance and tint tools actually work?

Every pixel is a mix of a red, green, and blue value. To compute that value, demosaicing algorithms make assumptions about the usual distribution and combine the sensor values accordingly. White balance then shifts the balance between the R(ed) and B(lue) values toward cooler (toward blue) or warmer (toward red). Tint does the same thing but for the green value, which white balance actually leaves untouched.

If you want to play around, here's a little simulator that does this for ONE pixel. Just set the base color and mess with the sliders. Have fun 🙂