FPV Drones

I'm flying a DJI Mini 3 Pro, which I know isn't the most beastly camera drone out there but for where I'm at it's plenty to work with. The 1/1.3-inch sensor is decent and it shoots in RAW which turned out to be really important for this whole experiment.

The Location and Timing

I drove out to a reservoir about 45 minutes from my house. Nothing Instagram-famous or anything, just a spot I'd seen from the road a few times and thought looked interesting. There's a dam on one end, some rolling hills, and a little patch of pine trees that kind of juts out into the water. I checked PhotoPills the night before to figure out exactly when golden hour would hit and where the sun would be. That app is genuinely useful if you haven't tried it — it shows you the sun position overlaid on a map so you can plan your shot angles ahead of time instead of just guessing when you get there.

I got to the spot about 40 minutes before sunset. That gave me time to walk around a bit, find where I wanted to launch from, and just get a feel for the area. Wind was light, maybe 8 mph from the southwest, which was totally manageable.

Camera Settings — This is Where I Struggled

Okay so I'll be honest, my camera settings have been kind of a disaster in the past. I was mostly shooting in Auto and telling myself I'd fix it in post. That works okay sometimes but you lose a lot of control and the files aren't as good as they could be.

This time I switched everything to manual. Here's roughly what I ended up with after some trial and error in the field:

• ISO: Started at 100, bumped to 200 as the light dropped
• Shutter speed: 1/500 at first, then down to 1/250 as it got darker
• Aperture: f/2.8 (the Mini 3 Pro has a variable aperture and I find f/2.8 gives me the most usable light without too much diffraction)
• White balance: Set manually to around 5600K and adjusted it once when the sky started going more orange
• Shot format: RAW+JPEG so I had a backup to compare against

The thing I kept messing up was the exposure. I was so worried about blowing out the sky that I was underexposing the foreground. Some of my shots had this really dark water that just looked muddy. I started using the histogram more instead of just eyeballing the screen, which helped a lot. The screen on the Mini 3 Pro is pretty good but it's still hard to judge exposure accurately outside.

Filters — Did I Even Need Them?

I bought a set of ND filters a couple months ago (a third-party brand, nothing fancy, like 35 bucks on Amazon) and I've barely used them. I brought them out this time because I'd read that ND filters help during golden hour to slow your shutter down and get a more cinematic look, even for stills.

I tried the ND16 for a few shots. It let me drop to around 1/60 shutter which was slower than I expected to use for stills. Some shots came out slightly motion-blurred on the ripples in the water, which I actually kind of liked. Others looked just soft in a bad way. I ended up going back to no filter for most of the session because the light was dropping fast and I didn't want to be fumbling with filters when the best colors were hitting.

Honestly I think filters are more useful for video or for shooting in really bright midday sun. For golden hour stills where you're working fast and the light is changing every two minutes, I found it easier to just adjust settings manually and skip the filter.

Composition Stuff

This is the part I'm still weakest at. I tend to default to just hovering and pointing the gimbal straight down, which can look cool sometimes but gets repetitive fast. I made myself try different angles this time. The Mini 3 Pro has that upward-tilting camera which I've almost never used — I tried pointing it forward at a low altitude, like 30 feet off the water, and flying slowly toward the tree peninsula. That gave me this really dramatic low perspective that I haven't gotten before. The reflection in the water came out really well in a couple of those frames.

I also tried the classic rule of thirds stuff, keeping the horizon on the upper third to emphasize the reflection. That worked better than centering it, at least for this spot. And I flew a bit to the side to get the dam in the background of one shot instead of always having it dead center. Small thing but it made the photo more interesting to look at.

Editing — Lightroom on the Laptop

Got home and pulled the RAW files into Lightroom. This is where shooting RAW really paid off. A few shots that looked kind of flat on the controller screen had way more color information in the RAW file than I expected. I was able to bring up the shadows on the foreground without the sky falling apart, which would have been impossible with JPEG.

My basic edit on the best shots:

• Pulled highlights down quite a bit to recover the sky
• Lifted shadows on the water and trees
• Added some warmth to the white balance — the camera setting was close but the RAW file let me nudge it further toward that golden tone
• Texture and clarity bumped slightly to bring out the tree detail
• Dropped the dehaze slider a little bit — I sometimes see people crank dehaze and it makes shots look crunchy and weird

I did not go crazy with saturation. I see a lot of drone shots online where the colors look like a cartoon and I'm trying to avoid that. I want it to look like a really good version of what was actually there, not like I applied some wild preset.

What I'd Do Differently

Get there even earlier. The 40 minutes of prep time was fine but I would have liked to scout flight paths more carefully before the good light started. I wasted probably 10 minutes of golden hour just repositioning because I hadn't planned my shots well enough. Also I want to try a CPL filter next time to reduce the glare on the water — I've heard those make a real difference for reflections and I'm curious to try it.

Overall though I'm genuinely happy with how a few of these shots came out. Like, I'd actually put one of them on my wall. That hasn't happened before. Still a long way to go but it felt like real progress and I'm already thinking about where to go next weekend.

If you're also learning aerial photography and want to swap notes or whatever, feel free to drop a comment. I don't have all the answers obviously but it's nice to talk through this stuff with people who are actually doing it.

Last Saturday I drove out to a reservoir about 45 minutes from my house. I'd scouted it on Google Maps and it looked incredible from above — there's a long curved dam, some dead trees sticking up out of the water on the east side, and open fields behind it. Golden hour was supposed to hit around 7:15 pm and I got there at 6:45 thinking I had plenty of time. I did not have plenty of time.

Getting There Early Actually Matters

By the time I walked out to a decent launch spot, checked the airspace on the app, ran through my pre-flight, and got the drone in the air, I had maybe 12 minutes of real golden light before the sun dropped behind the tree line. I knew intellectually that golden hour is short but I really underestimated how fast everything changes. The quality of that light literally shifted like every two or three minutes. Next time I'm getting there an hour early minimum, getting everything ready, and just waiting. Don't be like me and rush your setup.

Camera Settings I Was Messing With

I shoot in manual on the camera when I can. For this session I was running ISO 100, f/2.8, and I was adjusting shutter speed as the light changed. The rule of thumb I've been using is shutter speed at double your frame rate — so if I'm shooting at 24fps on video that's 1/50th, but for stills I've been going a bit faster just to keep things sharp when there's any wind. Out at the reservoir there was a pretty steady breeze coming off the water so I was keeping it around 1/200 for photos.

One thing I want to flag for anyone else shooting with the Mini 4 Pro — the ND filters make a huge difference and I feel dumb that I didn't use them for the first two months I owned this thing. I picked up a set of Freewell filters (the all-day set) and I had the ND16 on for most of the golden hour shoot. It let me pull the shutter speed down and get that slight motion blur in some of the water that looks really nice in photos. Without the filter in that light I was having to jack the shutter speed way up and everything looked a little clinical and flat to me.

Composition Stuff I'm Still Figuring Out

Honestly composition from the air is something I'm actively struggling with and I don't want to pretend I have it figured out. What I'm trying to practice is the same stuff you'd do with any camera — rule of thirds, leading lines, looking for natural framing — but it's weirdly harder to see from above at first. Your brain isn't used to that perspective.

At the reservoir I was trying to use the curve of the dam as a leading line and I think some of those shots actually came out decent. I pulled the drone back and up to about 80 meters so I could get the full curve in frame with the water on one side and the fields on the other. The long shadow the dam was casting across the water in that light was something I wasn't even planning for and it ended up being the best shot of the session. So I guess part of the lesson there is just fly around and look at what's actually happening in front of you, not just what you planned in your head before you got there.

I also tried some shots directly overhead looking straight down — the nadir shots — at the dead trees coming up through the water. That texture looked really cool in the golden light but I still feel like I'm not getting the framing right on those. The trees are spread out kind of randomly and I couldn't figure out the right altitude to capture them in a way that felt intentional rather than just random. Work in progress.

Editing When You Get Home

I shoot RAW (DNG files) pretty much always now. When I first got the drone I was shooting JPEG because I didn't really edit photos and I thought it was fine. It's not fine. The RAW files have so much more latitude when you're recovering highlights or pulling up shadows and at golden hour you have both happening at the same time — the sky is really bright and the ground is getting dark. In Lightroom I was pulling highlights way down, lifting shadows, and bumping clarity just slightly. I also dropped the whites and brought the blacks up a bit to keep it from feeling too contrasty.

Color-wise I'm keeping it pretty natural. I see a lot of drone photos online that are way over-saturated, especially the oranges and yellows in golden hour shots, and I'm trying not to do that. I boosted saturation like 8-10 points at most and then used the HSL panel to pull the orange hue slightly warmer and nudge the luminance on the yellows up a bit. That felt more like what it actually looked like standing there than the hyper-orange stuff you see everywhere.

One thing I'm not sure about yet is sharpening. I've been applying some sharpening with the masking slider pulled way up so it's mostly hitting edges and not adding noise to the sky. But I honestly don't know if I'm overdoing it or underdoing it. If anyone has opinions on that I'm genuinely curious what your workflow looks like in the comments.

What I'm Taking Into the Next Shoot

Get there way earlier. Have the ND filter already on when I leave the car. Spend the first ten minutes just flying around and actually looking before I start trying to capture anything. And honestly — check the weather more carefully. The sky was pretty flat on Saturday, just a thin haze on the horizon, and I think the shots would've been significantly better with some actual cloud texture up there to catch the color.

I've got another location I want to try that has an old railroad bridge over a river valley. I'm going to try to time it for a morning with some fog in the valley. No idea if I'll pull it off but that's kind of the whole point — figuring it out as I go. I'll write about it when I do.

Thanks for reading. If you're newer to aerial photography like me and you've got tips or things you've learned the hard way, drop them below. I'm definitely not the expert here, just sharing what I'm working through.

Anyway I've been meaning to document this rebuild because I made some component changes I'd been putting off for months and this felt like the universe forcing my hand. If you're also flying a budget 5" freestyle build and you're on the fence about some upgrades, maybe this is useful. Or maybe it's just me processing my trauma publicly. Either way.

The Frame Situation

I was running an IFlight Nazgul frame, which honestly held up way better than it had any right to for the number of times I've crashed it. But the arm that snapped was already had a hairline crack I'd been ignoring. Classic pilot behavior. I'd been eyeing the TBS Source One v5 for a while partly because replacement arms are cheap and easy to find, and partly because I wanted a slightly more open build for better airflow around the ESC stack. Ordered the Source One. It shipped fast and the build quality is exactly what you'd expect for the price — functional, no frills, the standoffs are fine but nothing special.

Frame swap on a 5" isn't a huge deal but it does mean basically pulling everything apart, which I hate and also secretly enjoy. It took me about three hours spread across two evenings mostly because I had to rerun all my motor wires and I cut them a little short the first time around. Measure twice, cut once. I know this. I still don't do it.

Motors — Finally Made the Switch

The motors I'd been running were some older BrotherHobby Returner R5s I bought secondhand. They were fine. Not exciting, just fine. Two of them had slightly noisy bearings and I'd been living with that hum for too long. With the rebuild I went ahead and ordered four Emax ECO II 2306 1900kv motors. I wanted something in the 1900kv range for 4S since I mostly fly parks and open fields rather than racing gates, and I didn't want to spend a lot. The ECO IIs have a solid reputation for the price and I've seen people put a lot of hours on them without complaints.

Soldering the motors to the new ESC was the most satisfying part of the whole rebuild honestly. I've gotten a lot more comfortable with the iron over the past year. I'm using a Hakko FX-888D and if you're still on a cheap Amazon special, please just get a decent station. It makes such a difference. Clean joints, quick, no cold solder messes. I tinned the pads on the ESC first, then tinned the wire ends, then brought them together. Simple stuff but worth saying for anyone newer to this.

The VTX Antenna Drama

The original antenna mount on my old frame was garbage and I'd jury-rigged something with some heat shrink and prayers. When the crash happened the coax pulled out of the connector on the VTX itself — which, honestly, it was only a matter of time. I replaced the whole antenna with an Axii 2 stubby that I had sitting in my parts bin. If you don't have a couple of spare antennas in your parts bin you're going to have a bad time eventually. I also added a small piece of TPU printed mount I grabbed from Printables that actually holds the antenna at the right angle and protects that coax connection point a lot better than what I had before. Should've done this from the start.

Flight Controller and ESC — Kept These

My FC and ESC stack survived the crash without issue. I'm running a Mamba F405 MK2 with the matching 45A 4-in-1 ESC underneath it. I've had zero issues with this stack and I wasn't about to replace something that wasn't broken. I did take the opportunity to clean up my wiring though because it had gotten pretty messy over various repairs. Used some thin silicone wire for the power leads to the FC and zip-tied everything a little more carefully. The stack went back in clean and I'm genuinely proud of how tidy it looks now. For maybe six weeks until the next crash anyway.

One thing I did change was the capacitor. The old one was a small 35V 1000uF cap I'd slapped on there originally and I swapped it for a slightly higher rated 50V version just because I had one and it fits fine. Whether that actually does anything meaningful I honestly can't tell you but the voltage spikes in Betaflight blackbox logs do look cleaner, or at least I think they do. Could be placebo.

Tuning After the Rebuild

New motors mean new tune, or at least checking your tune. I ran RPM filtering since I already had bidirectional DSHOT set up and that's been solid. What I did have to adjust was master multiplier — the new motors have slightly different characteristics and it felt a bit mushy on quick direction changes in my first test hover in the backyard. Bumped P and D a bit on roll, left pitch alone since it felt fine. Did a couple slow test flights in a big open area before I trusted it fully. I know some people just rip it immediately after a rebuild but I've learned my lesson there.

First real flight was yesterday evening and it felt genuinely good. Snappy, responsive, and no vibrations that I can feel or see in the DVR. The ESC temps after a few packs were reasonable — nothing that worried me. Motors were warm but not hot.

Takeaways If You're Doing Something Similar

Honestly the biggest thing I'd say is don't wait for a crash to do maintenance you've been putting off. I knew those motor bearings were rough. I knew that antenna mount was sketchy. I just kept flying anyway. Sometimes that's fine and sometimes a tree catches you and now you're doing a full rebuild on a Tuesday night when you should be sleeping.

The Source One frame is genuinely a solid choice if you want cheap arms and a community that's already solved most of the fitment questions. Motor-wise the ECO IIs feel great so far but I'll report back after I've put more time on them. And seriously, get a proper soldering iron if you're building anything more complex than a toy. It will change your life a little bit.

Next up I want to actually try GPS rescue on this build since I've never set it up properly and that's a whole other post. Assuming I keep the quad out of trees long enough to get there.

I'm running a Mavic 3 Classic, which I picked up back in the spring. The Hasselblad sensor on that thing is no joke, and I knew going in that I was probably leaving a lot on the table by not knowing how to use it properly. My plan was to get there about an hour before sunset and shoot through golden hour. Simple enough, right?

The Camera Settings Disaster

Here's where I embarrassed myself. I showed up and just kind of panicked. The light was already doing this gorgeous thing where it was hitting the water at a low angle and everything looked orange and soft. I was rushing to get the drone up and I completely forgot I had left the camera in auto mode from my last flight. So my first like fifteen shots are all overexposed, blown out highlights, and the histogram is just a flat wall on the right side. Completely unusable.

Once I calmed down and remembered what I had actually studied, I switched to manual. For golden hour I ended up settling on ISO 100, f/2.8, and I was adjusting shutter speed as the light changed. I've read that a lot of people use the 2x focal length rule for shutter speed when shooting stills to avoid motion blur, but honestly with aerial photography it's a bit different because the drone itself introduces vibration. I found that around 1/200s gave me sharp images without the highlights getting too crazy.

I had also packed my ND filters and almost didn't use them because I was flustered. I ended up putting on the ND16 for a little while which helped me get the shutter speed down to something that gave the water surface a slightly smoother look without going full silky blur. For stills I'm still experimenting with how much motion I actually want in water. Video is obviously different and that's a whole other thing I haven't really dug into yet.

Composition Stuff I Didn't Expect to Matter

Okay so here's the thing nobody really warned me about. When you're looking at a landscape from ground level, you have these natural leading lines everywhere. Roads, shorelines, tree lines. From above, everything flattens out and those lines become way more abstract. I was flying out over the reservoir and the shots just looked kind of... boring? Flat water, flat sky, flat treeline. Pretty, but not interesting.

What actually saved a few of my shots was noticing a small dock that stuck out into the water. I repositioned and used that dock as a leading line pointing toward the sun on the horizon. Immediately the composition felt more intentional. I also tried shooting at a few different altitudes. My instinct was to go high, like 200 feet or so, but I actually got more interesting shots between 60 and 100 feet where the trees along the bank were still visible as a foreground element and created some depth.

I've heard people talk about the rule of thirds for aerial photography and I had the grid overlay on my screen, but I kept ignoring it because I was distracted managing the flight itself. That's something I need to practice more — trusting the drone to hover and actually thinking about the frame instead of just watching the aircraft.

The Light Was Incredible and I Still Almost Blew It

Golden hour from the air is genuinely something else. The shadows get really long and they reveal texture in the landscape that you completely miss at midday. The water was picking up this deep amber color and there was just enough cloud cover to get some drama in the upper part of the frame. For about 15 minutes everything was perfect.

And then I almost had to land because I forgot to check my battery before I started shooting. I had used a partial charge earlier in the week messing around and I didn't top it off fully. So right in the middle of the best light I got a low battery warning and had to bring it in. I had a second battery with me luckily, but by the time I swapped and relaunched, the light had shifted. Lesson learned, charge everything the night before.

Editing When I Got Home

I shot in D-Log M format which the Mavic 3 Classic supports, and honestly the raw files look terrible straight out of camera. Really flat and desaturated. If you haven't shot in a log profile before, don't be alarmed when you first see the files. You have to apply a LUT to get it back to something that looks normal, and then you grade from there.

I brought everything into Lightroom. I applied a free Mavic 3 D-Log M LUT I found online as a starting point and then did my own adjustments. For the golden hour shots I pushed the oranges and yellows a bit warmer, brought the highlights down, and lifted the shadows just slightly to bring out some detail in the darker parts of the trees. I also did some selective masking on the sky to keep it from looking overcooked while I was adjusting the foreground exposure.

Out of probably 80 shots I took that evening, I'd say maybe 6 or 7 are actually worth showing anyone. Two of them I'm genuinely really happy with. One is the dock leading line shot I mentioned, and the other is a high angle looking almost straight down at the water where you can see the reflection of the clouds and this one kayaker who happened to be paddling through. That one was pure luck honestly.

What I'm Doing Differently Next Time

I've already made a little pre-flight checklist for photography sessions specifically. Batteries fully charged the night before, camera set to manual with D-Log M before I even leave the house, ND filters in the bag and sorted by strength, and I want to spend the first few minutes of a shoot just hovering and actually looking at the composition options before I start hammering the shutter button.

I'm also going to try arriving even earlier. Like two hours before golden hour so I have time to scout the location from the air without the pressure of fading light. The reservoir has some interesting little coves I could see from above that I never would have known about from the ground.

If you're just getting into aerial photography like me, honestly just go shoot. I spent way too long watching other people's footage and not enough time making my own mistakes. The mistakes are where you actually figure out what works for your style and your gear. I'll post the dock photo in the gallery section if I can figure out how to do that on this site.

Pretty hard, it turns out. But also really fun, and I feel like I learned more in those two sessions than I had in the previous six months combined. So I wanted to write this up while it's still fresh because maybe it'll save somebody else some of the same headaches.

The Location and Planning Part

I drove out to a reservoir about 45 minutes from my house. There's a spot where you can park and walk down to a flat area near the water with some hills behind it. I'd actually scouted it on foot a few weeks before, which I'm really glad I did because it meant I knew roughly where I wanted to position myself and what direction the light would be coming from. I used the app PhotoPills to check where the sun was going to set and when golden hour would actually start. If you haven't used that app, seriously look into it. It's not free but it's worth it.

I got there about an hour and a half before sunset, which gave me time to set up and also just fly around and get a feel for the area. Wind was maybe 8 mph out of the northwest, pretty manageable. No other people around, which meant I wasn't stressing about flying over anyone.

Camera Settings - Where I Went Wrong First

Okay so this is where I made my first dumb mistake. I had the Mini 3 Pro set to auto exposure because that's just how I'd been flying. I took maybe 20 shots and when I pulled them up later on my laptop they were all over the place - some blown out, some weirdly dark, and none of them had that warm punchy look I was going for. Auto is fine when you're just flying around but for serious photography you really need to take control.

I went back out four days later and this time I switched to manual mode and shot in RAW. For the settings I ended up at ISO 100, f/2.8 (the Mini 3 Pro doesn't give you a ton of aperture options but f/2.8 is the widest), and then I adjusted shutter speed to get the exposure right - ended up around 1/200 for most shots during golden hour. I also turned on the histogram overlay so I could actually see if I was clipping the highlights. That made a huge difference.

One thing I read before going out was about the 180-degree shutter rule for video, but for stills I was more focused on just keeping ISO low and not blowing out that sky. The sky during golden hour is doing all kinds of crazy stuff and if you let it blow out you've basically lost the shot.

I Also Tried an ND Filter for the First Time

I picked up a cheap set of ND filters off Amazon - I think it was like a 4-pack with ND4, ND8, ND16, and ND64. I used the ND16 for most of the shoot. Honestly I'm not totally sure I needed it for stills in the way you need them for video, but I do think it helped me slow down the shutter a little and get a slightly smoother look on the water. The reservoir had some gentle ripples and with a slightly slower shutter you could see just a little bit of silkiness to it. Nothing dramatic but I liked it.

Putting on the filter was kind of nerve-wracking the first time because I was standing there trying to snap this little piece of glass onto my drone and I was terrified of dropping it or scratching the lens. It wasn't hard though, just a little fiddly.

Composition Stuff I Was Thinking About

I'm not a trained photographer or anything so I've just been reading and watching a lot of stuff online about composition. The main things I tried to do were use the rule of thirds (put the horizon on the upper third so you're showing more water and foreground), look for leading lines, and try different altitudes. I think the altitude thing is something people don't experiment with enough. I see a lot of drone shots that are all taken from like 200-300 feet and they all start to look the same. I tried some shots from maybe 80 feet and some from up around 350 feet and they feel totally different.

The lower shots felt more intimate I guess - you could see the texture of the water surface and the rocks along the edge. The higher shots gave you more of that sweeping landscape feel with the hills in the background. Both were useful, just for different things.

I also spent some time just flying laterally and tilting the gimbal at different angles instead of just pointing it straight down or straight forward. Some of my favorite shots from the session were with the gimbal at maybe 45 degrees tilted down while I was slowly moving sideways. Kind of an awkward angle to think about but it gave an interesting perspective on the shoreline.

Editing in Lightroom

I use Lightroom Classic and this is where the RAW files really paid off. The JPEGs I had from my first session were pretty locked in - not much you could do with them. The RAW files from session two I was able to really work with. I pulled the highlights down hard on the sky, brought up the shadows in the foreground, added some warmth and a little bit of dehaze. Nothing crazy, I'm not trying to make it look fake.

The colors during golden hour were already doing most of the work honestly. I just tried to make sure the edit matched what I remembered seeing when I was standing there.

What I'd Do Differently

A few things. I'd bring a jacket - I was out there until after sunset and it got cold fast. I'd also bring a second battery because I used almost all of battery one just messing around during my first practice pass and then had to rush through the actual golden hour window. Also I'd get there even earlier and do a full test flight first so I'm not figuring out my camera settings when the light is already going.

But overall I'm really happy with how session two went. I got maybe five or six shots I'm genuinely proud of, which doesn't sound like a lot but for me at this point it felt like a big deal. I'm planning to head back out to the same spot in a few weeks and try it in different conditions - maybe some cloud cover which I've heard can actually look amazing for this kind of shot.

If you're also kind of new to aerial photography and haven't tried golden hour yet, I'd really recommend just going for it. Even if your first attempt goes like mine did - kind of sideways - you'll learn a ton.