This new drone from DJI brings an truly staggering amount of innovation and integration to the FPV space:
- Low res analog video -> high res digital video
- Modest transmission range -> long range
- Integrated simulation to learn to fly
- "Bail out" button that almost instantly nulls drone movement and puts it in a stationary hover
- ~5 minutes fly time -> ~20 minutes fly time
- Adjustable camera tilt on the fly to maintain level horizon at a variety of drone speeds (drone tilts more with speed)
Extremely impressive first offering in this category.
The 20 min flight time is a lie, Joshua Bardwell measured 10 minutes hovering. That's extremely easy to surpass on typical light weight 5 inch freestyle drone with a 1300mah battery that cost a fifth of the price of the DJI ones.
It is an impressive drone, but for experienced FPV pilots it's very limiting.
For better or worse, because there's a lot of irresponsible drone flyers as a result too, but this product makes fpv flying much readily available for those who want to try it without investing the time/effort to learn about how to do it better with other setups
The industry is doing a horrible job of educating people and, as a result, they are ruining it for everyone else. I have been flying all manner of radio control craft for over thirty years (planes, helicopters, drones and "things"). As a simple example, I would never --ever-- fly anything over people, homes, roads or high fire risk brush areas. I have crashed enough time to understand that that these things are toys that are not designed to the standards of real aircraft, often suffering from multiple single-point failure risks that could bring them down and either hurt someone or cause serious damage to property.
Our neighborhood FB group is, at times (usually around Xmas and people's birthdays) filled with videos from people flying their new drones over roads (as in, directly over cars), over homes, over bone-dry brush surrounding our neighborhoods, over people (crowds, kids, at parks, without anyone knowing), etc.
The other kind of post is "I flew my drone for the first time in front of my house and I lost it". Which, of course, makes me crying at multiple levels. The crash could easily cause a brush fire that could take out an entire neighborhood [0], and, of course, hurt or kill someone [1].
At first I tried to educate people as to the realities of these devices. I own RC helicopters that cost well in excess of $3,000 and I still consider them unsafe toys. Yet people get offended and verbally abusive when you explain that their $300 drone, despite the hype, is a toy that could fall out of the sky at any time and potentially cause great harm. I mean, I have worked on the design of space capsules that are docked at the International Space Station and people are arguing with me about their toys actually being "reliable". Jeez.
It's only a matter of time until a tragedy occurs. Thankfully the large fires caused by drone crashes have, so far, been limited to open areas. One day someone is going to crash a drone into the bushes in a neighborhood, the LiPo is going to rupture and half the neighborhood will burn down. Sadly people seem to value their fun far more than the safety of others.
[0] https://arstechnica.com/tech-policy/2018/03/drone-crashes-in...
[1] https://www.theverge.com/2017/2/27/14755116/jail-sentence-dr...
I do not believe that this is a good thing at all.
To be fair, hovering time is often lower than flight time for copters.
The main thing anyone starting needs to realize is that it takes quite a bit of time to get used to how Quads fly in Acro/Rate mode. I started on a simulator (Liftoff) with an XBox pad and that was super punishing as gentle and precise control is needed. Switching to a proper radio was a night and day difference. I'd guess it took 5-10 hours of flying time to get to the point where I could comfortably go where I wanted 99% of the time. I'm up to 50 hours of sim time now and flinging the things around with wild abandon. Looking forward to summer and getting out to try the real thing.
My recommendation for starting out would be to grab a cheap radio that is easy to connect to a PC and a copy of a simulator.
to have your mind blown by experienced fpv pilots, check out Nurk FPV, Johnny FPV, Mr. Steele, and Rotor Riot.
FPV is a disproportionately expensive hobby ... every hard crash can cost you between $20 ( broken motor bell) to $200 (cost of entire drone when dropped in water and electronics go poof). Still great hobby, highly recommend!
> - Modest transmission range -> long range
Those are not really new. DJI has been selling their standalone digital FPV system for a while now. And there are lot's of different long range radio systems available.
But I agree that this is impressive. I wonder if Betaflight, etc. will also implement some kind of bail out feature.
Major benefit of analog video is a progressive degradation of quality as a signal gets weaker and almost instant reacquisition after a complete dropout. I'm curious to see how DJI competes with that on a digital video.
> ~5 minutes fly time -> ~20 minutes fly time
For FPV racing that's rather a downgrade. Ideally you'd like to land with almost empty battery right after finishing the course. Any leftover charge is a dead weight that only slows you down. But then again - I don't think this drone will be used in competitive FPV racing.
The good news is that you can find that out right now. DJI HD FPV has been available for over a year, this is just their first drone offering in the category.
The short answer is that it degrades very gracefully. They have what they call 'focus mode', where the edges of the video lose quality before the middle (where the data matters the most). Once you've used up that signal degradation threshold the rest of the frame slowly degrades to a lower bitrate, and then after that latency starts rising.
After flying analog for 5 years it only took me about 1 day to strongly prefer the DJI FPV video solution. The range is far better, and the video is mind-blowingly clear when compared to analog. On anything but my racing specific drones I have been installing the DJI HD FPV system instead of normal analog gear.
Pretty convincing comparison by Joshua Bardwell:
Shark Byte: https://youtu.be/9iKc2v05blw?t=926
vs DJI: https://youtu.be/9iKc2v05blw?t=1046
vs analog: https://youtu.be/9iKc2v05blw?t=1227
If the latency is low enough it would be nice if they could incorporate eye tracking into that in the future, and move around the high quality region with your eyes. And maybe feed it from a gimballess 360 camera so you have full freedom to look around. Stereo cameras would also be nice, maybe with some differential compression between eyes to keep bandwidth down (would need a gimbal for that most likely, to keep line between cameras horizontal).
At this point though it's probably something I'd look at closer once I have a bit more spare time and can think about situations where it might be applied.
Waiting for a degradation tolerant codec.
Old MPEGs were, as they had broadcast use in mind, but even 264 blow them out of the water on compression nowadays.
Loss tolerant codec + modern channel coding will likelly outrange analog for quality picture
Lots of I-frames, I guess...
Or perhaps the signal strength or SNR.
These parameters are available but not usually displayed.
Uh? Can you explain this?
I'll probably get one, but continue to complain loudly online that every DJI product, including the offline ones, requires an account and email address and registration and the transmission of your serial number for it to work - including their little handheld, no-internet, record-to-sd-card cameras. :/
I don't want to have an offline DJI-only phone.
Most people who fly drones are sending their IP, gps data, and serial number to DJI on every flight.
I'm hoping they keep their hobbyist/DIY FPV system going though, as it blows everything else out of the water. Analog video is well, bad, open source alternatives are either abandoned or mediocre, and the Divimath/Fat Shark system is not even comparable in terms of quality.
For Hobbyists, the high res digital video is definitely a great development and makes for a better flying experience overall. Real racers still complain about the latency, but they are not the target of the digital FPV system and the new DJI FPV drone always.
But the high res is nothing unique to the DJI FPV drone. DJI started offering the Digital FPV system a few years ago for integration into diy drones.
> - Modest transmission range -> long range
There are plenty of long range transmission standards in the DIY space such as TBS Crossfire etc. For video transmission, it heavily depends on the transmission power which quickly becomes the listing factor. In Germany (and most EU states iirc) the transmission power for 5GHz signals limited to 25mW, which heavily reduces the range . 2.4GHz is not supported with the DJI AirUnit or Caddx Vista iirc.
For lots of countries long range isn't even that much of an important feature since its legally only allowed to fly line of sight anyways. However, for countries with less strict drone laws this seems like an awesome addition.
> ~5 minutes fly time -> ~20 minutes fly time
That is the most misleading part imho. The 20 minutes flight time are only realistic when flying in the fully sensor supported "Mavic like" mode. If you're actually going for the powerful freestyle FPV experience, the flight time is still around 3-5 minutes with the major difference that a replacement battery for the DJI FPV is about 150€ while a LiPo for a DIY racing quad costs around 30-40€ a piece. That the flight time is heavily decreased when flying in the actual FPV freestyle (acro) mode is not surprising (its only a 2000mAh battery in a heavy drone), but I find the claim highly misleading.
> - Integrated simulation to learn to fly
In the FPV space, simulators have been a thing for quite a while now. There are multiple simulators such as Freerider, Velocidrone or Liftoff, all with different focuses and a broad support for controllers. You can even use a Xbox controller for the beginning. So integrating a simulation to learn to fly is no unique to DJI and nothing new in the FPV space at all.
> - "Bail out" button that almost instantly nulls drone movement and puts it in a stationary hover
The bail out button, even though I see a lot of risks with it, seems like a great feature for beginners. I really hope that beta flight will add a similar feature to upcoming firmware versions.
The DJI FPV drone seems like a relatively fragile drone. The most annoying part is, that there is apparently no real way to fix anything yourself besides the props. The current FPV drones really can take a beating and you probably won't break that much. If you do, you can simply order a new arm for your frame or a new motor. You need some technical knowledge and skills such as soldering, but you do not depend on the manufacturer with an Apple like repair policy.
The DJI FPV seems like a great entry option for cinematographers who just want to get some FPV style footage (which might be possible, but seems limited since the props often show up in the frame) with as little technical confrontation as possible. For people who just fly for fun and want some high speed stunts, diy drones still seem like the better option.
For anyone wanting to enter the hobby, there are a lot of bind and fly drones out there where little technical knowledge is required. For the technical knowledge that is required, there are lots of really supportive communities.
That seems to be untrue - according to the verge:
"If you’re like me and wreck this drone, you can replace the following items yourself: the top housing shell, propellers, the gimbal / camera module, and the arms. "
[1]https://www.theverge.com/22308345/dji-fpv-review-first-perso...
Most FPV channels that were reporting on the DJI FPV drone and actually tested the completely manual flight mode (which is the only one that really compares to the FPV flying you see in many videos) got around 3-5 minutes out of the battery.
And again, it makes sense when looking at freestyle maneuvers and how much power they require.
Also cool that they have been embracing virtual practicing it seems. The original DJI controllers were impossible to pair with simulators, so I had to hack it: https://github.com/Matsemann/mDjiController/
I sold (or basically gave away) all my home-made drone setup a year or so ago, after it being in a box for 3+ years. It was a hassle to set up, and very hard to fly and get good pictures and actually be in control (original DJI Phantom included). Cool tech and fun to hack on, but not very good UX heh. However I would still very much like to fly and film more, but then with some out-of-the-box-ready solutions. Modern ones seem to have obstacle avoidance, better hovering, tracking or point-to-point etc, so I can actually focus on the (for me, now) fun parts of filming the sports action.
What are some good drones for that use-case? DJI alone has a big lineup now, hard to know which one is more fitting. And are there other contenders selling ready to use stuff? Thought GoPro had one but seems like they've stopped.
With DJI, all of this is closed. I'd be surprised if you can connect/have access to any of these things.
I wonder if any consumer VR FPV drones would consider transmitting a 360 video signal (or just one that includes more FOV than can be displayed on the headset) to allow some head movements to be handled in software.
Yes, we thought of that as a possible improvement ("future work"). It may be harder to get the stereo cameras to match up, is my guess. But for a single camera setup it's probably not too hard and already exists for 360 videoes. So I guess if one's capable of streaming the video it shouldn't be too hard.
And stereo cameras arent really that useful. We explored it, and found that when looking at stuff more than a few meters away it made no difference. For inspections (which was what our drone was made for) of close stuff it has some merit, but could still probably be handled by some IR sensor or other computational stuff adding an overlay to the video.
As for vr sickness, the most jarring experience was flying behind yourself like a 3rd person shooter haha.
I also like that there finally seems to be a decent alternative to DJI in the market-- it was a bit worrisome how much better DJI has been than everyone else.
what's really important is image quality, when the mavic 2 came out it really blew me a way with the one inch sensor. It was incredible to see that kind of image produced by such a compact body.
I really like DJI and how innovative they have been in making awesome consumer products that just work and are super easy to use. I have owned a few.
DJI had been supplying drone technology to Chinese security forces in Xinjiang https://www.pcmag.com/news/us-blacklists-drone-maker-dji-ove...
Uighur camp detainees allege systematic rape https://www.bbc.com/news/world-asia-china-55794071
Gang raped, shackled and broken students: Inside China’s ‘horrific’ Uighur detention camps https://www.independent.co.uk/news/world/asia/china-uighur-d...
But the same could be said of big western tech (Amazon, Microsoft, etc) companies providing tech to western armies.
Not looking forward to random people flying these around me at 60-90mph ... but I really want one!
Imagine flying the drone, and moving your head would move the camera gimbals, and you'd experience it like you were in VR in the drone?
Human binocular vision depth perception is actually extremely short range. It varies from person to person, but some measurements put it as short as 10 feet [1] - beyond that, the mind starts depending on other cues [2] like motion and relative size.
So the costs in bandwidth and hardware might not be worth it - getting your drone that close to things sounds like an expensive skill to learn :)
[1] http://visionlab.psych.sites.olt.ubc.ca/files/2016/03/131_Wa... [2] https://en.wikipedia.org/wiki/Depth_perception
Funny, I was thinking the same when I watched the video. The optimist in me hopes there's a pair of lower-res stereo cameras located somewhere near the main camera (the one we can see) being specifically for shooting video? The way you described it is exactly how I thought it operated until I got further into the video. It's a great step forward in drones with added range, batter, etc, but the whole "first-person VR" shtick feels like it was only half-thought out and never changed during the design/implementation
They have bitten a lot of Apple’s packaging, presentation, and UX — which isn’t really a bad thing. (Imitation and flattery, and if it ain’t broke don’t fix it, etc.)
I’m intrigued by FPV, but I have no burning need for it right now — I’m mainly interested in cinematography and building maps of my property for farming, gardening, and architecture.
One thing I would say: having a big property to learn on has been handy. It does concern me how flippant some are in this thread about barriers to entry and the risks. Until you get yourself into a hairy situation, you may not fully appreciate the risks.
I’ve been waiting patiently for VR+robotic applications - this project but controlling some Boston Dynamics is a terrifying military, or an amazing football team depending on how you look at it. Exciting!
I’d happily pay a reasonable sum for it. I wonder if it could be done using a Stadia controller and chrome cast?
Likely would have to build some autonomy into the system to handle scenarios where the communication gets patchy or the user is about to crash into something.
Now you only need to install 5G on the Grand Canyon.
And ruin it for everyone who actually spent the effort to go there in person.
Basically the gave you a big "box" of air space that had 0 obstacles that you could fly around in. It's impossible to crash the drone and it takes off / lands automatically.
Also, does anyone have a solid guess as to what sort of restrictions this thing has on it? I like DJI for their mechanical engineering but their drones are way too locked down. Are we going to have to reverse engineer this one too for fun mode to be playable?
I have a DJI Mavic Mini and the lockouts are very, very strict. I can't fly it at ALL at my house outside.
I wish there was a little more compromise.. in some cases it would be a heck of a lot useful if you could fly even with some draconian restriction like a 25ft max altitude. My house is in a restricted zone due to being on the flight path for an airport. But the airplanes come over my house around 150ft AGL. There are 75ft trees all the way around my property. I would have plenty of fun never going above 50ft in my yard, but DJI won't let me fly at all.
Given it's speed this FPV model would have the largest capacity for mayhem of almost any DJI model, and since it's FPV you know there will be lots of people flying it without a spotter in violation of FAA regulations.. so it makes sense for it to be at least as locked down as the other models.
I flew my drone over salt marshes near another airport that didn't have a tower over the weekend. Because it didn't have a tower it didn't have a "Red" zone near it and I could fly after acknowledging the risks. I was able to fly and keep my altitude < 20ft and it was still a ton of fun and I got some great pictures. I only flew for about 10 minutes, I saw one airplane, but it was ~100ft AGL and I was ~15ft AGL and had some horizontal separation from the plane, no harm done.
This model looks insanely fun and I'd love to have one but realistically I bet I'd be in that situation a lot where I didn't have a spotter and couldn't fly it without knowingly breaking the rules. I didn't see anything in the material yet in terms of whether or not you can fly it line of sight without the goggles.
Check out the faq/start. I haven't looked into minis, but the larger mavics are well supported.
If you want to really have some fun with your current drone, check out a program called Litchi. It costs lunch to get a copy but it gives you a full ground station with mission planning capabilities, a Google Cardboard compatible FPV function (a bit laggy for precision flying but if you pick you lines through obstacles you will be fine), limited head tracking that moves the camera gimbal as you look around and a heads up display / audio feedback.
All in all, modified dji + litchi + google cardboard headset + burner android phone is a surprisingly capable package that can be repaired, tweaked and has a very low price point if you can navigate the used goods market.
Where does this drone sit with regard to the current set of regulations? It will certainly at least need operator ID markings, yes?
How soon can I scratch together $1300?
As a new Mavic Mini owner, the one feature I'm craving for is FPV.
As opposed to replicating the package and capability of what DJI sells.
With such fast movements in close quarters, I would love to see a global shutter.
We were watching a neighbor's house over the holiday, which meant mostly checking for delivery boxes on their porch and putting them inside. We could drive the Mini2 out the window of our house and 2 blocks away to check the porch without getting out of our seats. And this is a suburban area full of houses, trees, and a shit ton of interference at the wifi/4G/LTE ranges. The video never lost a frame.
If you tear down the hardware, it's using an Imagination IE1000 802.11ac frontend, Skyworks amps, and a custom DJI image processor chip (marked P1). So the air protocol probably looks somewhat like 802.11ac (OFDM etc.) but without the framing overhead and presumably without TCP/IP.
6 mile range test: https://www.youtube.com/watch?v=4X8ExuJWt6E
I own Quest and DJI FPV system (not the drone from the link, but custom built with DJI FPV googles)
- I can get dizzy in VR, but I won't in FPV unless I do a lot of crazy movement like spinning for a couple of seconds. Even when crashing which involves a lot of spin, it's fine - If I watch FPV while someone else is flying, I feel dizzy. It's not as fast as in VR, but still can happen
Friends who tried had the same feeling.
There's something in the connection between visual movement and your control of that movement via stick that helps alleviate dizzines.
Hope that helps in case you were wondering!
It’s a shame because this looks fantastic but I probably won’t be able to use it without getting motion sickness almost immediately.
