If you’re into astrophotography, a drone typically isn’t your camera of choice. That’s why I was surprised to see Haida’s release of a light-pollution reducing filter for the new Mavic 3, which they’re calling the NanoPro Clear-Night Filter. Does this filter change the experience of shooting with a drone at night?
Drones and low-light photography have never been a perfect match, particularly when shooting with smaller, consumer-grade drones. Between the smaller sensor sizes required by the platform, the inherent stabilization challenges making long exposures difficult, and a messy regulatory environment around any night flying, night photos via drone were less common. Each of those points has changed significantly over the last year or two, particularly with the release of the Mavic 3, in my opinion. I've started to shoot into the night, and have been extensively testing Haida's Clear-Night filter.
A quick note on regulatory matters: rules, restrictions, and laws will vary depending on where you’re flying. Always make sure you’re in compliance with any relevant restrictions and always fly safely. Night flights in particular present challenge to visibility and require the highest standards of care.
Back to the fun stuff: the Mavic 3’s camera is the first drone that I’ve used that is a competent night-photography tool. Towards sunset and even after the blue hour, the Micro Four Thirds sensor is capable of producing clean images with good detail. However, as you push the camera to its limits, the files are a bit more “brittle.” That means it’s tougher to make big swings in exposure, and rebalancing colors isn’t as easy.
This is where the Clear-Night Filter comes in. Fstoppers has previously reviewed the Clear-Night filter in a format meant for use with regular DSLR and mirrorless lenses, whereas Nando concluded that the filter was more useful when photographing predominantly transitional skies, namely those outside the city, but not at a completely light-pollution free site. Additionally, he found that he could replicate the effect in the city with some careful Lightroom tweaks.
For the Mavic 3, I thought those characteristics would mean this filter would be useful. Many of my intended compositions for drone night photography will inevitably feature a lot of sodium-vapor lights, the exact color cast this filter can reduce. Also, as previously mentioned, drone images don’t hold up to heavy tweaking in post, unlike files from my Z 7, for instance. That brittleness makes the idea of “getting it right” at capture more important, and filters can be a big part of that concept. Lastly, I wasn't expecting to shoot under any true dark-sky conditions with the drone, as the tech unfortunately just isn't at the point where you can get truly long exposures yet.
Getting an exact side-by-side test of filters on a drone is a bit of a challenge. There’s inevitably a large difference in both time and position between filtered and unfiltered shots, as you have to land, swap filters, relaunch, and line back up. As a result, I’m going to focus on the experience shooting with the filter in place, instead of trying to present exact comparisons.
I did run into one issue before I even got into the air, but it isn't Haida’s fault. Instead, it’s DJI’s. The mounting system, at least on my copy of the Mavic 3, is awful. Attaching and removing filters is very awkward, requiring what feels like a ton of pressure on the camera and gimbal, which is a delicate mechanism. Haida has fully replicated the DJI mounting system, even down to the magnets, and there’s not an issue with their filter; it’s the entire mounting system that I don’t like. With that in mind, I’d love to see a future clip-on or magnetically attached filter that just goes over the DJI glass filter.
Once I got the filter on, I used it on multiple flights for shots at both blue hour and sunset, in both areas with a significant amount of light pollution and some areas with a more suburban/rural feel. In both places, I felt like the filter did make a difference in reducing the intensity of the yellow sodium vapor lights, with only a moderate reduction in exposure. It did lead to a significant color shift in those lights, as well. It seemed to shift sodium vapor lights from that stereotypical yellow to a more orange-ish color, with almost a salmon undertone.
This exposure reduction that the filter causes come from the fact that the filter is blocking some amount of light coming into the lens, and while this reduction can vary depending on what light makes up your scene, I’d guess it’s about a half-stop. It's not massive, but could impact shots in windy conditions.
This color change isn't unpleasant and almost makes the resulting image closer to what you'd see in-person compared to the unfiltered drone capture. I wouldn't characterize the effect as light-reducing, however. Instead, it seems to be almost entirely a color shift.
Another major consideration, as shown in the above before and after, is that the filter doesn't have an impact on light sources that aren't in that 589 nm band occupied by sodium vapor lights. Here, the metal halide stadium lights are virtually unaffected, as would be LED street lights or fluorescent lights in offices. The "before" is without the filter, and these shots are processed through Lightroom, with cropping to make them overlap better.
I didn’t notice a significant difference in the sky, however. I’m not sure if the camera wasn’t sensitive enough to the change, the exposure wasn’t configured to fully reveal any differences, or if the location I was in just didn’t create as much atmospheric light pollution. I did some additional testing with sky-centric photos, but couldn't find a meaningful difference in the images. Throughout testing, the major difference in shots with and without the filter was confined to the area directly around sodium vapor lights; the drone's camera may just not be sensitive enough to pick up light pollution.
This reduction and change in light make the use of this filter a deliberate choice. You’re trading some exposure latitude for a slightly different color palette. As both the degree of color change that you can get from the camera, as well as the degree to which you can push the exposure are limited, I think this filter is a useful, albeit specific tool. It's also a particularly useful option if you're shooting drone video but aren't comfortable color grading footage. Here, it can provide a different and more pleasant look without requiring video editing expertise.
If you’re expecting to be photographing in an area with a lot of that yellowish light and it’s not particularly windy, you can trade a slightly longer exposure for more pleasant colors. On the other hand, if you’re looking to maximize the light available to your drone, leaving the stock protective glass on may be a better choice. Overall, the filter is an inexpensive and interesting option for drone photographers who shoot after sunset. The Clear-Night filter is currently available for the Mavic 3, and is also available for the Mavic Air 2, although that model wasn't tested in this review.
What I Liked
- The effect is pleasant, if minor
- The price point is very reasonable, particularly for a drone accessory
- Build quality matches that of first-party filters
- Useful for both photo and video
What Could Be Improved
- Marketing may be unclear about what impact this filter has on the photos
- Mounting filters of any kind to the Mavic 3 is unpleasant — a "click-on" or magnetic version would be even better
Why is the word "astrophotography" in the title of the article? I only see pictures of the ground.??
Was wondering the same. Clickbait at its finest (chef’s kiss)!
The filter is marketed as an astro photography filter, but as I stated in the review, the differences aren’t visible in the sky. That’s the whole point of the review - questioning the suitability of this filter for astro.
Alex, that's the point of the complaint, you didn't take any ASTRO (sky) photos. If you are questioning the suitability of the filter for ASTRO, shouldn't you be proving your point by taking ASTRO photos?
Not a flyer but as far as the Clear Night filter and others like it just a few years ago city lights/street lights were the amber halogen, still used along coastal areas to keep young hatched turtles from coming toward land and more toward the water like the moon light does. Today you have what I call Moonlight temp LED lights, which the same if you take a daylight temp light and shine on a grey card. A problem with the white LED lights is a camera image of the many lights come out blue, unless your camera has the AWB/white selection it gets rid of the blue of shadows on white boats say at sunset/rise. The Clear Night filter came along kinda late because of the white LED lights came out fast. The clear night filter gets rid of the orange/yellow sky glow BUT I must add the blue of the of the new white led light sky glow. Where the the clear night filter helps is with the very many different colors of lights when doing a cityscape and keeps the white blue to good clean white as well as all other colors the way your eyes see them. I use the Clear Night filter at the back of my lens when doing Milky Ways and this maybe a little of subject but the colors of gases sky glow (green/red/etc) pop more along with the yin and yang colors of pegasus above the Galactic Center that comes out whiter also. I can see where a drone flying say in and over the Grand Canyon where hotel lights are still Amber maybe for the wildlife would be great.
Your compare images show the green of the baseball field brighter and greener but the led lights still have a blue tint. I live in a city where each of 7 bridges are a different color lit as well as every building just about and night shots are so hard to get right in post, this will be a add to any drone. It will add to sunset colors, I forgot to take mine off my lens once! Also if on the back of a lens it will give a baby blue sky with no effect of direction (90 degree rule) even with a ultra wide 12mm vs the front of the lens. 1st A night with many lights 2nd using the filter on my Voigtlander 10mm f/5.6 in a multicolor city. Last image no color correction, AWB in camera A7Siii and FE 1224 mm f/2.8 at f/8 facing west toward a 2pm sun, I was scouting and forgot to remove the Clear Night Filter that was mounted at the rear of the lens and you can see no distortion in the blue of the sky. All my lenses have the Haida rear filter mount. Using both a rear and front filter.
Just another thing the governments are going to have to make illegal because "some people"...
I'm done with fstoppers! Either their writers, editors, or both suck!
I don't understand your beef with Fstoppers and this article...?
The Mavic 3 is pretty decent at night. Haida has offered the filter for other DJI drones for a while. I was never sure what the purpose was. Plus, Haida gives no examples on their web site. One question, did you try the filter during the day? If so...what changes did you see?
Yeah, their marketing around it seems a bit confusing. They've got a picture of the Milky Way on their promotional materials, which is a bit unrealistic for a drone IMO.
In the daytime, you won't really see any effect - some orange tones may be a bit different, but daylight is such a broad spectrum that this filter's specific cut won't make much difference.
Hi, Alex, it has been a long time since I worked with these filters, but I recall that there were perhaps three night sky filters that were marketed for amateur astrophotography. One was for stripping out wavelengths of light emitted by low-pressure sodium vapor bulbs. Another was for stripping out wavelengths of light emitted by mercury lamps. And the third, and one that was not as effective as the other two, was a filter for stripping out the light emitted by high-pressure sodium lamps. The former two are easier to strip out because those emit light at specific bands of wavelengths of light. But the latter, the high-pressure sodium lamps, are messier (no pun intended - messier - like Messier objects) (ba dmp sssss) (bad joke). Anyway, these filters all are not so effective unless you take long exposure shots.
Now that most public street lighting is LED, it makes the mercury lamp and low-pressure sodium lamp filters fairly useless and renders the one that you are testing even less effective, due to the many and different wavelengths emitted by white LED street lighting.
Yup. I briefly mentioned that as something to note - these filters won't have any impact on LED, metal halide, or fluorescent lighting, for better or worse.
Yes, it's complicated now, since white LED lighting is wildly inconsistent as far as the spectrum goes, as compared to dealing with the characteristic glow of an excited gas or heated filament. Even then it was tricky and those filters had limited efficacy. But yet they were still usually worth the purchase. If we could standardize on one single-color LED street lamp such as amber, it might be easier. Thanks for the article!