Transcription
If you only rely on LUTs or plugins to get your look, you're missing most of the process. In this video, I'm going to show you how to build a high-end color grade from scratch. No LUTs, no plugins, using only native tools in DaVinci Resolve. We're going level by level. Each stage unlocks a new layer of control. Light, color, tone, and finally, full look development. By the end, you won't be guessing anymore. You'll know exactly why every adjustment works.
So level 1, let's fix the light. At this level, we're only focusing on one thing – light. We want the highlights, midtones and shadows to sit in the right place, so nothing is clipped, nothing is crushed, and everything feels balanced. This is especially important if you don't have perfect lighting on set. Because here's the thing. If your light is off, your color will never look right. Now watch this, by adjusting light alone, without touching saturation, we can already bring back color. And to show this properly, I shot this in Hong Kong – busy streets, messy lighting, constantly changing conditions. By the way, if you want to practice with this footage, just let me know in the comments.
Let's break this down. This shot was filmed in S-Log3, so first we normalize it. All this step does is convert what the camera captured into a proper viewing space – Rec.709, the standard for web and video. That gives us a clean starting point. Make sure "Apply Forward OOTF" is enabled so the image looks natural on screen. If you're already shooting Rec.709, you can just skip this. Since I'm working with log footage, I place it into DaVinci Wide Gamut first. This gives us more room to work before bringing it back to Rec.709.
Now bring up the waveform. Turn off "Colorize", because at this level, we're only looking at luminance. This shot is bright, harsh, and washed out. So what we're gonna do is compress the tonal range and recover the image. Create a node for overall light adjustments. I'm using the HDR color wheels, set to DaVinci Wide Gamut. If you're working in Rec.709, use the primaries, but only adjust the luminance sliders. Right now there's too much highlights, so I bring it down. Lower the lights, and lift the shadows slightly. These adjustments can be subtle or aggressive. It depends on the image. But the goal is always the same. Push the light into the right tonal zone. Now it looks flat. That's expected. We'll fix that in a moment.
Next issue – White Balance. There's a slight warm tint, so we create a separate node for color balance. There's an effective and accurate method for both standard and log workflow. Switch your balance node's Gamma to Linear and use the Gain Wheel for balancing. They give you controlled, precise adjustments across the entire image. Here I'm just pulling it back slightly to neutralize the color bias.
Now, let's bring back contrast, but with control. Create a new node and go to Custom Curves. If you're on the Studio version, you can switch to HDR mode for finer control. Select the Y channel only because we're still working purely in luminance. Switch the histogram to Output so you can see the changes in real time. To build contrast properly, we need a pivot. Think of this as your anchor point – the middle gray. It is the midpoint between black and white, and where your camera naturally balances exposure. In DaVinci Wide Gamut, that sits at 0.336. In Rec.709, it's 0.435. Once you lock that in, you shape around it. To find it precisely in the free version, create a node and disable all the other nodes. Turn Contrast and Saturation down to 0. Set the pivot to 0.336. Then go to Curves and check the waveform. The spike you see, that's your pivot. Pin it. Reset Contrast and Saturation. Turn your nodes back on.
Now we build contrast. Enable editable splines. Then go to Waveform settings. First, switch to Data Level in Scopes. Then set soft boundaries for reference as it shows on screen. These are just guidelines to avoid clipping or crushing. For this shot, I bring the highlights down. Lift the shadows. And shape the curve around the pivot to control the midtones. And now, the image has depth again. And the true color starts to come through. At this point, we haven't even touched color. But the image already looks better. That's the power of fixing lights properly first.
Now let's take it further because even with correct exposure, uneven light distribution can still distract from your subject. So we fix this in two steps. First – Corrective Light Rebalance. Create a node before HDR. Go to Window. Use a gradient mask to reduce the brighter side. Then adjust only luminance in Gamma. Now the image feels more even. Second – Light Shaping. Now we guide attention. Create two layer nodes after Curves. In the lower node, select the subject and lift it slightly. In the upper node, do the opposite and darken the surrounding slightly. Now compare before and after. It's not perfect, but the focus is much clearer. So at Level 1, we're not adding style, we're building a solid foundation. And if you're not chasing a stylized look, this alone is often all you need.
But if you want your image to really stand out, now we move on to color. Most people think color grading is about adding more color and boosting saturation. But at a higher level, it's about restraint and control. Now that we've built a solid tonal foundation in Level 1, we move on to chroma. At this level, we're not just adjusting colors, we're fine-tuning the relationship between them. Take a look at this image. Technically, it looks fine. If you check the waveform, nothing is clipped, nothing is crushed, the light is balanced. But if you look at vectorscope, the colors are all over the place. There's no hierarchy. So let's fix that.
Create a node after your Curves or any targeted masking. Go to Hue vs. Saturation. First, green. In this scene, green is not the priority, so I click on the leaves and gently reduce the saturation. Not too much. We still want it to read as green. Most people would boost everything here. We're doing the opposite. Next, red and yellow. These lanterns should feel balanced, but right now the red is overpowering the yellow. So I bring red down slightly and lift yellow just enough until they sit at a similar intensity. It's a very subtle adjustment, but it immediately makes the image feel more balanced.
Now, let's refine the quality of the overall color. Create another node. Go to Luminosity vs Saturation. If we zoom in, the lanterns look a bit flat, almost plastic. So we're going to reshape it. To make colors feel richer, we reduce saturation in the shadows and highlight, and let the midtones carry the weight. This is what I call the "Hero Zone". Start with the shadows, click on the lower range and bring it down. Then slowly adjust the transition until it feels smooth. Now look at the lanterns, they instantly have more depth. The colors feel layered, not flat, not digital. This behavior is actually very similar to how film responds to color. Now we do the same for highlights, but more gently. I prefer a much softer roll-off here, so the highlights don't feel harsh or digital. And just from this, your image already starts to move away from that digital look, even before we touch on any stylization.
Let's quickly look at another example. In this street scene, there's a lot going on. Many competing colors. This time, instead of Hue vs Saturation, I use Color Slice. This tool feels more filmic in how it behaves. You can control hue, saturation and density within specific color ranges. Here, yellow isn't important, so I reduce it. The greenish-cyan scaffolding is distracting, so I lower both its saturation and density. And for the red taxi, just a small boost to bring it forward. Then again, I apply Luminosity vs Saturation to shape the depth of these colors. Now the image feels much more controlled and much more nuanced. If we want to go even further, we can clean things up surgically. Create a layer node. Use a qualifier to isolate the concrete road. There's a slight yellow tint here, so I removed just that. Now is this necessary? Not always, but I want you to see what's possible. So at level 2, we're not adding more color. We're organizing it. We're deciding what matters and what doesn't. Make sure you're comfortable with level 1 and level 2, because from here on, we start building actual looks.
Alright, this is where we start creating looks. Up until now, everything we've done has been about control. Now we start shaping style. We begin with color toning, also known as split toning. This is where we apply different color tints to different tonal zones, often with contrasting colors. As a matter of fact, this is one way to get the classic "Teal and Orange" look, where shadows lean toward teal and skin tones sit in warmer tones. But of course, color grading is much more than that. At a higher level, we use color toning to do three things. First, creates color contrast. Second, guide attention to the subject. And third, shape the mood of the scene.
Take a look at this example. This was taken at a street food vendor. This image is already color corrected. It looks fine. But it doesn't say much. So let's enhance it. Before we start, turn off the targeted colored adjustments from level 2. We'll bring them back later. Create a node after your light adjustment section. For this, we'll use a combination of Color Wheels and Log Wheels So we get a smooth and natural transition between tonal zones. Our goal here is simple – make the subject stand out. Bring up your vectorscope, set it to "Mid" so we can focus on midtones. Start with the Gamma wheel. Set the Luma Mix slider to 0, so your color adjustments won't affect the brightness of the image. Now gently push it toward a warm tone. Be subtle here. These controls are very sensitive. For this image, I'm keeping it close to center for a more natural feel.
Now we create separation. Add a new node. Switch to Log wheels. Change the vectorscope view to "Low". In the shadows, push slightly toward teal. Again, keep it subtle. Then adjust the shadow range to control how much of that tint affects the midtones. Now look at the before and after. The subject immediately stands out. Next, we refine the highlights. Switch the vectorscope to "High". In the Highlight Wheel, add a slight warm tint, just enough to move from pure white to a softer creamy white. Then adjust the range for a smooth transition. It's a very small change, but it adds a lot of polish. At this point, the image has direction. Now, you can do your level 2 color refinement adjustments. But as you can see, just from this technique alone, you can create many many many different looks. But we are not done yet.
Let's polish it further. This is where Soft Clip comes in. Soft Clip is a simple but powerful way to shape your highlight rolloff and soften shadows. It's often used for that slightly milky shadow look. I have mixed feelings about it, but it's very common in modern grading. Create a node after your color adjustment. Go to Custom Curves and find Soft Clip. Start with shadows. Bring up the waveform and turn off "Color". First set your “Low” threshold. Then adjust “Low Soft”. to control how much you compress the shadows. Be careful here. You still want to preserve detail. Now, highlights. Here I'm more generous. I prefer a softer roll-off, so I increase the highlight compression until it feels smooth. And now, the image feels more refined. And if you think this is powerful, Level 4 takes it even further.
Okay, this is the final level – the Boss Level. Up until now, we've been shaping light and organizing color. Now we're going to remap color itself. This is called color compression, or gamut compression. In level 1, we compressed light. Now we're compressing color. Instead of spreading colors across the entire spectrum, we start pulling them closer together, so the palette feels more cohesive and more unified. And for this, we use one of the most powerful tools in DaVinci Resolve – the RGB mixer. In here, we will use a technique called RGB Crosstalk, where information from one color channel is mixed into another. This is where things get a bit abstract. So instead of over-explaining it, let me simply show you.
Create a node after your light adjustment section. Go to RGB mixer, I'll explain it as I go. First, Green Output. Turn on "Auto Normalize". This keeps your white balance stable as we mix channels. Also enable "Preserve Luminance" so brightness stays consistent. Now watch this. Increase blue in the Green Output. This is the part most people don't even know exists. What's happening here is that we're blending color channels together. Blues and greens start shifting toward cyan. Yellows and reds start leaning towards orange. And just from this, look at the image. The palette already feels more cohesive. This move, blue into green, is a classic film style adjustment to mimic Fuji or Kodak film stocks. The more blue you add, the more compressed and stylized the image becomes. But now, look at the skin tone, it's getting a bit too red. So we correct that. Add some red into the green output, and gently rotate the hue back until the skin looks natural again. Now compare the before and after, the image feels tighter, more focused, less random.
We can push this further. Go to Blue Output. Turn on "Auto Normalize" again. Increase green. Then reduce red slightly to balance the overall hue. Now look at the vectorscope. The color distribution is even tighter, more compressed. And look at the image. The blues are warmer and those magenta patches in the skin are cleaned up. Skin tones look more flattering. And this becomes your new foundation. From here, you can bring back level 3 color toning level 2 color hierarchy, and refine everything on top of this. Just keep an eye on skin tones, our eyes are extremely sensitive to them. Every small shift can feel off, so adjust carefully.
If you're using the Studio version of DaVinci Resolve, you have access to a few more tools that can really elevate the image. This is all about texture and finishing. I won't go too deep here, but I'll show you my current node placements. Right after your first Color Space Transform, add Noise Reduction just enough to clean up luminance and color noise. Then, after building your look, you can start adding texture, Bloom or Glow to soften the light, Halation to add the subtle reddish edge around highlights. If needed, you can add Sharpening, but personally, my Sony footage is already too sharp, so I usually go the other way, and soften it slightly to reduce that digital edge. I also use Lens Blur to blur out the edges of the frame. And finally, Film Grain. This ties everything together and adds a more organic feel. None of this is required, but if you want that final layer of polish, this is where it comes from.
We've covered a lot, so let's simplify it. First, we normalize the image. Then, we fix the light. We balance exposure, shape contrast, and control where the viewer looks. Next, we shape color. First, we compress it to build a cohesive base. Then, we tone it to add separation and mood. And at the very end, we polish. So if you understand this workflow, you don't need LUTs, you don't need plugins. Got questions or you want to practice with my footage? Just let me know in the comments. We've got a lot more to explore and I'll see you in the next one.