Drone Racing

The Location and What I Was Going for

I drove out to this reservoir about 40 minutes from my house. I'd driven past it a dozen times and always thought it would make good footage - there's a concrete dam on one end, a lot of open water, and some pine trees along the ridgeline that catch the light really nicely in the late afternoon. My original idea was to do something that felt moody and slow, kind of like those travel films you see on YouTube where everything is grade and smooth and there's some ambient music underneath.

I was shooting with a DJI Mini 3 Pro. I know, I know, it's not the Mavic 3 Cine or anything high end, but it's what I have and honestly the footage it puts out in good light is genuinely solid. I had it set to 4K 30fps with D-Log M enabled so I'd have some room to color grade later. This was actually my first time shooting in D-Log M for a real project and that alone caused me some grief, which I'll get to.

I Actually Tried to Storyboard This One

This was new for me. Usually I just fly around and grab whatever looks cool and then try to edit something out of it afterward. That works for personal stuff but it ends up being this random pile of clips that don't really connect. I watched a few videos about shot planning and decided to sketch out a rough sequence on paper before I left the house.

My shot list ended up being something like: wide establishing pull-back over the water, slow orbit around the dam structure, low pass along the tree line, rise-up reveal of the full reservoir, and a final slow push toward the horizon. Pretty simple stuff. The goal was that each shot would connect to the next logically so the edit would feel like it was going somewhere.

Having that list actually made a huge difference. I wasn't just hovering around trying to figure out what to do next. I had a purpose for each battery. I brought three batteries and it turned out I needed most of them.

The Flying Itself - Smoother Than I Expected, Mostly

One thing I've been working on is slowing down my stick inputs. I have a habit from FPV of being pretty aggressive with the controls and that looks terrible on a camera drone. Like, the footage comes out looking twitchy and rushed. So for this shoot I was deliberately trying to think about moving in long, gradual arcs instead of pointing and going.

The orbit around the dam was the hardest. Getting the subject centered while moving laterally and maintaining altitude is way more coordination than it sounds. On the Mini 3 Pro you can use the automated circle mode, but I wanted to do it manually because I wanted to be able to adjust speed and radius mid-shot. I ended up doing maybe five or six attempts at that shot before I got one that felt controlled all the way through. The early ones had this little wobble where I was overcorrecting.

The low pass along the tree line came out better than I expected. I was flying maybe 15 feet off the ground going pretty slow and the parallax of the trees moving past the camera actually looked really cinematic. That shot ended up being my favorite in the edit.

Where Everything Kind of Fell Apart

Okay so the D-Log M thing. When I loaded the footage into DaVinci Resolve it looked completely washed out and flat, which is normal, but I had no real idea how to grade it. I'd watched some tutorials but watching and doing are very different things. I spent probably three hours trying to get the colors to look right and most of that time I was just pushing things around randomly and making it worse.

Eventually I downloaded a free LUT pack someone had shared in a Mini 3 Pro Facebook group and applied one of those as a starting point. That got me closer. Then I adjusted the exposure and pulled up some of the shadows and it started looking more like what I had in my head. It's not a perfect grade by any stretch but it doesn't look like garbage either, which I'm counting as a win for a first attempt.

The other problem was that I had way too much footage. I shot probably 25 minutes of clips across three batteries and the final video is a minute and forty seconds. That sounds like a good ratio I guess, but the editing process was brutal because I kept second-guessing myself about which version of each shot to use. Note to self: next time take better notes during the shoot about which takes actually felt good in the moment.

The Edit and the Music Problem

I cut everything in DaVinci Resolve. I'm not super advanced in it but I've been using it long enough that I'm comfortable with basic cuts and transitions. I kept the editing pretty simple - mostly straight cuts, one or two very subtle cross dissolves where the pacing needed it.

Finding music was its own whole thing. I needed something royalty free if I was going to post it anywhere and I went through probably two hours of browsing on Pixabay and some other free music sites. A lot of it is just generic and doesn't fit the vibe. I finally found an ambient instrumental track that had the right slow, open feeling and that made a big difference to how the whole thing came together.

The final video is up on my YouTube channel with about eleven views, most of which are probably me checking it. But honestly I'm less interested in views and more interested in the fact that I actually finished something intentional for once.

What I'd Do Differently

Practice color grading before the real shoot, not after. Spend more time on the orbit technique because that's a useful shot I want to get better at. Maybe shoot a shorter project next time so the edit doesn't feel so overwhelming. And I'd probably do a location scout before the actual shoot day just to see how the light falls at different times.

If you're in a similar spot where you fly a lot but haven't really tried to make something with purpose yet, I'd say just try it. The planning part felt awkward and kind of overthought but it genuinely helped. And even when stuff doesn't go perfectly, you learn more from actually finishing a project than from just collecting raw clips forever.

I'll probably post the video link here once I stop cringing every time I watch it.

The Plan That Made Sense in My Head

I picked a local state park about 45 minutes from my house. Nice open meadow, a small pond, some tree lines. I went out there the week before just to walk around and scout it without the drone, which I actually think was a good call. I took notes in my phone about where the sun would be hitting in the morning and basically sketched out a rough shot list. Nothing fancy, just like six or seven ideas. Reveal shot coming over the tree line, low glide across the water, that kind of thing.

I also watched probably three hours of videos about storyboarding for drone footage. Most of it was aimed at people with way more experience than me but I picked up the basic idea — you want shots that connect, not just random pretty stuff. A beginning, middle, end. Establish where you are, move through the space, end on something that feels like a conclusion. Got it. Simple enough.

I set my alarm for 5:15am to get golden hour light. This part actually worked. The light was incredible when I got there.

Where It Started Falling Apart

First problem was wind. I checked the forecast the night before and it said 7 mph which is totally fine for the Mini 3. But out at the meadow it was gusting more than that, probably 15 or 16 mph at times. Nothing dangerous but enough that my slow, smooth pulls and pushes looked really jittery in the footage. I had the drone fighting the gusts while trying to execute a slow reverse pull and you could just see it stuttering. Not great.

Second problem was I had my settings wrong. I've read about the 180-degree shutter rule like four times and I still managed to show up with my shutter speed at 1/1000 instead of 1/60 to match my 30fps. So my footage looks weirdly sharp and choppy during movement instead of having that slight natural motion blur that makes things look smooth and filmic. I didn't realize it until I got home and pulled the clips into DaVinci Resolve. That was a painful moment.

I didn't bring ND filters either. I have a set, they were sitting on my desk at home. I knew about the shutter rule, I knew I'd need them in bright conditions, and I still left without them. I genuinely don't know what happened there.

The Shots That Actually Worked

Even with all that, there were a couple of clips I'm actually happy with. The low glide across the pond turned out really nice — I was maybe four or five feet off the water, moving super slowly, and the reflection of the sky in the water looked amazing. I had to do it like six times to get one where the wind didn't push me sideways, but the final one is smooth. That one's going in the video for sure.

I also got a descent shot that came out well. Started high looking down at the tree canopy and slowly descended while rotating slightly. Took my time with the sticks and just tried to be as gentle as possible. That cinematic Look came through on that one. So I know I can do it when conditions cooperate.

Editing Was a Whole Other Lesson

I use DaVinci Resolve Free and I'm still learning it. I'd watched some tutorials on color grading drone footage and I shot in D-Log so I had to do a log conversion before anything else. I used a free LUT I downloaded and then tried to push it further myself and made everything look orange and weird. Ended up backing way off and just doing a lighter touch with the color wheels.

The bigger editing issue was pacing. I put the whole thing to a song I liked and just started cutting clips to the beat, which seemed logical. But when I watched it back, the shots felt disconnected. You'd see the pond, then suddenly you're looking at a tree line from above, then back to the meadow. There was no visual logic to it. I think this goes back to the storyboard problem — I had a list of shots but I didn't really think about how they'd flow together in the edit. That's something I need to plan better before I even fly next time, not figure out in post.

One thing that did help was cutting on movement. Like if the drone is pulling back in one clip, I tried to cut to another clip where there's also backward or lateral movement happening. That made the transitions feel a little smoother even when the locations changed. Small thing but it helped.

What I'm Doing Differently Next Time

Honestly the list is long but here's what I'm focused on:

• ND filters go in the bag the night before. Not optional.
• Check my camera settings before takeoff every single time. I'm going to tape a little cheat card to my controller.
• Storyboard with transition logic, not just a shot list. Think about how clip A leads into clip B visually.
• Scout for wind protection. That meadow was too exposed. Next time I want locations with natural windbreaks.
• Do more practice flights with no filming goal, just working on stick control and smooth movement. I think my inputs are still a little jerky.

I also want to try flying with just one or two shot types per outing instead of trying to get everything. Like go out with the specific goal of nailing a good orbit, or a specific goal of doing slow push-ins. Get really comfortable with one move before layering in six different ones.

Final Thoughts from a Total Beginner

I didn't post the video. Maybe I will eventually just as a before/after kind of thing once I have something I'm actually proud of. But I don't regret making it because I learned more in that one morning than I did in weeks of reading about aerial videography theory. There's no substitute for just going out and failing at it in the real world.

If you're at a similar stage and thinking about trying your first real cinematic project, just do it. Go make the bad video. You'll figure out what you need to fix way faster than if you keep waiting until you feel ready. I definitely didn't feel ready and I still don't, but at least now I know specifically what I need to work on instead of just vaguely knowing I'm not good enough yet.

More updates coming as I figure this stuff out. Hopefully the next one goes better.

Then I sat down at my desk, pulled the clips into DaVinci Resolve, and just stared at this flat, washed-out, kind of grey-green mess on my monitor. That was my wake-up call to the whole color grading thing. Turns out, shooting in D-Log M on the Mini 4 Pro and then doing nothing about it in post is not a workflow. It's just... sad files on a hard drive.

I want to talk about what I've learned over the past few months trying to actually grade my drone footage, because honestly this part of the process completely blindsided me. The flying stuff you can kind of feel out. Color grading is a whole separate skill that I didn't respect at all until now.

Why I Even Shoot in D-Log M

Quick context for anyone who's newer than me: D-Log M is a flat color profile on DJI drones that compresses the highlights and shadows to retain more dynamic range. The idea is you get more information in your file that you can then shape in post. The sky doesn't blow out as fast, dark areas hold more detail. It sounds great in theory.

The tradeoff is that the footage straight out of the drone looks genuinely awful. Like, your parents would watch that and think something's wrong with your camera. You have to do work in post to make it look like anything. For a while I was just slapping a LUT on it and calling it done, which, okay, that's better than nothing, but I was getting inconsistent results and I didn't really understand why.

The LUT Problem

LUTs — lookup tables — are basically color presets that try to interpret your flat log footage and transform it into something that looks cinematic or at least normal. There are free ones everywhere and paid packs that cost anywhere from ten bucks to ridiculous amounts. I bought a pack early on because someone on a Facebook drone group swore by them.

Here's what I didn't get: LUTs are not magic fixes. They're a starting point. And they're designed with specific exposure levels in mind. If your footage is slightly overexposed or your white balance was a bit off when you flew, the LUT is going to fight you. I kept getting these clips where the sky turned this unnatural teal color or the greens on trees looked almost neon. I blamed the LUT. The problem was actually that my exposure was inconsistent clip to clip.

I spent a lot of time thinking I just needed better LUTs. What I actually needed was to be more careful with my histogram while flying. On the Mini 4 Pro you can enable the histogram overlay in the camera settings and I basically never looked at it. Now I won't fly without it. Keeping my exposure consistent in camera made my LUT results about three times more predictable.

Learning to Use Nodes in DaVinci

I resisted DaVinci Resolve for a long time. I was editing in CapCut like a complete gremlin because it was easy. But once I started trying to do real color work, CapCut just couldn't do what I needed. Resolve has this node-based color system that looked terrifying at first.

The way it clicked for me: think of each node as a separate layer of color adjustment that feeds into the next one. I now have a pretty simple setup. First node is my CST — color space transform — which converts the D-Log M footage into a workable color space. Second node is my exposure and contrast corrections, where I balance clips so they match each other. Third node is where I apply my LUT at a reduced opacity, usually around 60-70% so it's not overwhelming. Fourth node is where I do any final tweaks, maybe a slight saturation pull or a vignette.

That's it for most clips. It's not fancy but it's consistent and I understand every step, which means I can fix problems instead of just undoing everything and starting over.

Matching Clips Shot at Different Times

This one is still kinda hard. When I fly a location over a couple of hours, the light changes. A shot I did at golden hour looks completely different from one I grabbed twenty minutes later when a cloud rolled in. Cutting between them without any color work looks jarring, even to people who don't think about this stuff.

What I've started doing is picking one clip as my "hero" clip — usually the one with the best light — and then manually pulling the other clips toward it. In Resolve there's a scopes panel with a waveform monitor and a vectorscope and I've slowly learned to read them instead of just trusting my eyes. Your eyes lie to you, especially if you've been staring at the screen for an hour. The scopes don't lie.

I'm not great at this yet. Sometimes I'll export a video and watch it on my TV instead of my monitor and something will look way too warm or the shadows will be crushed. But I'm getting more consistent.

The Part About Monitors Actually Mattering

Speaking of TVs — I did not have a calibrated monitor when I started this. I was grading on a cheap gaming monitor that made everything look more saturated than it actually was. So I'd think my colors looked great and then send a clip to my cousin and he'd say it looked washed out on his phone. That was frustrating until I figured out what was happening.

I'm not saying you need to drop a thousand dollars on a professional monitor. I'm saying you should at least be aware that your monitor's color output might not be accurate. I got a halfway decent IPS panel and calibrated it with a Spyder device I borrowed from a friend. Big difference in how predictable my results are now.

Where I'm At Now

Honestly my footage still doesn't look like the stuff you see from experienced aerial videographers on YouTube. But it looks like something I made intentionally, which is a huge step up from six months ago. The color grading process went from being this intimidating mystery to being a part of my workflow I actually look forward to.

My next thing to tackle is audio — as in, actually building proper soundscapes under my videos instead of just slapping a royalty free track on and calling it done. The music choices I've been making are embarrassing in hindsight. But that's probably another post.

If you're in the same place I was, shooting D-Log and wondering why your footage looks rough — just know it's a learnable thing. Took me longer than it should have because I kept assuming the problem was my gear. It was my process. Always the process.

A little background: I picked up a DJI Mini 4 Pro about four months ago. I'd messed around with a cheaper Potensic before that, mostly just flying around the backyard and crashing into hedges. The Mini 4 Pro felt like a serious step up, and I wanted to actually do something meaningful with it — not just hover over my street and call it a day. So I decided I was going to shoot a proper short cinematic video of a local reservoir about forty minutes from my house. Beautiful spot, good light in the evenings, not too many people around. Seemed like the right test.

Here's what I planned. Here's what actually happened. And here's what I'd do differently.

The Planning Stage (Where I Was Overconfident)

I watched probably fifteen hours of YouTube tutorials before I even drove out there. Watched guys talking about the rule of thirds, ND filters, color grading in DaVinci Resolve, the whole thing. I made what I thought was a solid shot list. I'm talking seven different moves: a low reveal coming up over the tree line, a slow orbit around the main dock, a pull-back from the water's surface, a top-down descent. I even sketched some rough storyboard panels in a notebook — nothing fancy, just little boxes with arrows showing direction of movement.

Here's the thing about storyboarding drone shots that nobody really told me clearly: you have to account for the fact that the drone doesn't care about your notebook. Wind, battery life, airspace restrictions, the actual geography of the location — all of that smashes into your plan the second you get there. I had a beautiful pull-back from the dock mapped out and when I got there, there were two guys fishing right at the exact spot I needed to be. So that shot was gone.

The storyboard wasn't useless — it actually helped me stay focused and not just fly around aimlessly — but I needed to treat it more like a flexible starting point instead of a script.

The Flying (Where Things Got Humbling)

I was running the Mini 4 Pro in Normal mode, not Sport, which was the right call for slow cinematic stuff. But I hadn't really dialed in my expo settings or anything like that. I just had everything at defaults. The result was that my movements felt kind of robotic and steppy when I watched the footage back. When I'd pan left, there'd be a tiny jerk at the beginning and end. Not terrible, but you'd notice it.

I also didn't use any ND filters for the first half of my session because I was waiting for the light to get lower and I thought it would be fine. It was not fine. The footage from that first hour looked too sharp in a way that felt almost like video game footage — no natural motion blur. I finally popped on an ND16 and immediately the difference was noticeable. The water looked smoother, pans looked more natural. That was a genuine lightbulb moment.

The low reveal shot — flying up from below the tree line — that one actually came out well. I was nervous about it because there were branches around but I took it super slow, went up at maybe 30% throttle, and let the landscape just unfold behind me. When I saw that in the editing timeline I actually said something out loud to my empty apartment. That was the moment I understood why people get obsessed with this stuff.

Batteries and Time Management

I brought two batteries. Should have brought three. The Mini 4 Pro gets around 30-34 minutes per battery in calm conditions, but I was dealing with light wind and the first battery drained faster than I expected — probably because I kept repositioning trying to find angles. By the time I was really getting comfortable with the location, I was on my last battery and the light was hitting that perfect golden-hour zone.

So I basically rushed my best light. I got a couple nice shots but I know I left better footage out there. Next time I'm going out there with three batteries minimum and I'm arriving a full hour before golden hour so I can scout with one battery and shoot with two.

The Editing Mess

This is where I spent most of my time after the shoot. I filmed in D-Log M, which is the flat color profile on the Mini 4 Pro, because all the tutorials told me to for maximum flexibility in post. And that's genuinely good advice. But I was not prepared for how flat and gray that footage looks right out of the camera. The first time I dragged it into Resolve I actually thought something was wrong with my settings.

I ended up using a free LUT I found from a drone videography community online, applied it, then adjusted from there. The color grade took me probably four hours across two evenings, which is embarrassing given the final video is only about two minutes long. But I learned a ton doing it. Lift, gamma, gain — I had to actually look up what each of those meant while I was using them, which felt a little backwards but it worked.

The edit itself I kept simple. No complicated transitions, no drone footage clichés like dramatic swoops set to generic cinematic music. I found a piece of ambient instrumental music through a royalty-free site, something slow with some acoustic guitar, and I just let the shots breathe. I matched some of the cuts to subtle moments in the music — not hard beat-syncing, just loose timing so it felt natural.

What I'd Tell Myself Before Going Out There

If I could go back and give myself five minutes of advice before that first shoot, here's what I'd say: bring more batteries than you think you need, put your ND filters on before the light is perfect not after, fly slower than feels natural, and don't stress if you only get three or four usable shots. That's actually a win for a first outing.

I've already been back to the reservoir twice since then. Each time I get a little more comfortable with reading the environment, planning my moves in advance, and not overcorrecting mid-flight when I get nervous about an obstacle. The footage is getting better. Slowly, but noticeably.

I'll post the actual video here once I've done another pass on the color grade — there's one shot that's still bugging me. But I wanted to write this first while everything was still fresh, because I feel like the messy middle of learning something is where the useful information actually lives.

More to come. Thanks for reading.