FPV

[Bebopperoo]

In that case you should go full FPV quad for immersive flying & leave your photo drones to what they are best for.

It will be digital if you want some viewing quality ... not analog, & it will most probably be DJI straight through goggles, controller & video/camera unit on that quad.

And if you by "aerobatics" mean flippy flops ... that entirely up to you, scenic FPV flying is relaxing. But if you mean the mode called acro (no height hold, not horizontally stabilized & no tilt angle limitations) it's another thing, that require some training, preferably in a simulator before you try it for real... & it's acro that will give you that immersive pilots seat view of the flight.

If you're committed to go for it, I recommend you going into it in 2 steps ... first buy a controller (absolutely easiest is to go with the DJI FPV controller first version) & a simulator for your PC (Velocidrone good for a bit older slower computers or Liftoff if your computer can handle it).

Then during the winter you practice there in full acro mode ... after 30-50h you should be able to fly without any disasters, the only thing that still is unknown for you by then is aerodynamic stuff like how fast will your real quad fall when you cut the throttle, how will it slide in the turns, affected by winds... & thing's like that, but after 2 real flights with your real quad you will have figured that out & can go exploring in a cruising manor.

Once you have seen that you can learn & progress in the simulator ... it's time to spend on the full kit with a "Bind & Fly" quad, the DJI FPV goggles, a charger & batteries. The total spending will be similar as if you go with the DJI FPV drone. So going with a pre-built "Bind & Fly" quad will not be cheaper, will not be more difficult to get to work (all is DJI so ...), will not require you to tinker or solder anything ... you're just cheaper prepared for the second quad that eventually will come & can reuse most of the accessories

I started exactly like this ... bought this "long range" 4", DJI goggles, the DJI controller, charger & 3 different types of batteries that will give different characteristics regarding weight, how agile the quad feels & flight times.

View attachment 137562View attachment 137563
Now a year later I have 2 more quads ... one 5" & one 7". Use them all 3 regularly but for different purposes.

This was from a short 2 battery pack test flight of a new location & to see if unstabilized footage from my old GoPro 3+ could be used at all (as I had lent my GoPro 10 to a friend) ... no flip flops, just flowy cruising, perhaps not especially scenic but ...

Couldn't watch. Every time i tried, the music forced me to dance around the room, missing the visual.

 

[EpicFlight]

I definitely want a full up factory-built unit.

I dunno...a $1300 drone and a $2000 computer...that's starting to get a little painful! And with winter coming up fast...

But this is what I needed, that actual data to chew on.

Thanks!

TCS

Listening to you here, I suspect you won't regret the investment, and will end up relegating your other DJI drones to shelf space, pulled out when you need to take pictures or video. The rest of the time – just for flying – it'll be the FPV.

And just to spend a little more of your money, you're gonna want the Motion Controller too.

 

[Chaosrider]

I've now read both the DJI FPV manual, and watched all of the Official DJI FPV training vids. But I'm still puzzled by a few things.

Is the phone used during actual flight operations, or just during set-up and updates? It doesn't attach to the RC, and I didn't see any reference to its use in operation in either the manual or the vids.

Anyone know what's up with that?

Thx!

TCS

 

[Kilrah]

Phone's only used for activation, firmware updates and to provide a live view to spectators if desired.

 

[Chaosrider]

Phone's only used for activation, firmware updates and to provide a live view to spectators if desired.

OK, that's consistent with what I've seen.

I assume the phone will then be seeing the same transmission from the AC as the goggles, yes?

Part of what's confusing me is that I don't yet understand why a second set of goggles wouldn't be looking at that same signal, and get the same video quality, as the "primary" goggles.

If the two sets of goggles are both looking at the same signal from the drone, why would the video quality be any different?

There's something fundamental that I'm still missing here...

Thx,

TCS

 

[Tomk_]

Shouldn't be. Both sets of goggles are just receiving the same radio signal. Seeing as how the screens on each set of goggles are the same, they should get the same quality. You just need to go into audience mode and set the one that the operator is wearing to Transmission mode and the other to Spectator mode.

 

[Chaosrider]

Shouldn't be. Both sets of goggles are just receiving the same radio signal. Seeing as how the screens on each set of goggles are the same, they should get the same quality. You just need to go into audience mode and set the one that the operator is wearing to Transmission mode and the other to Spectator mode.

Well, that's what I would think too, but I've seen a number of people complain of degradation in the second set of goggles, and no one has yet said they do it, and it works just fine.

This raises a number of questions....

I'm not sure how I missed this one in the DJI FPV training vids, but I sure didn't see it after I thought I had gotten them all at the DJI site.

What's that mechanical/electrical contraption shown with the controller and the goggles near the end? Do you need an extra gizmo to make multiple goggles work correctly?

I appreciate your help, I'm really trying to get closure on this.

Thx,

TCS

 

[Kilrah]

There has been some supposition that the audience goggles had their reception disturbed by the main ones transmitting in close proximity and getting some separation helped.

There's no obvious technical reason for the audience goggles to have worse reception, but it's how it is anyway...
DJI have had such issues basically on all their products that support dual video receivers all the way to the Inspire 1. At the time they mentioned it was related to synchronization/encryption IIRC.

 

[Chaosrider]

There has been some supposition that the audience goggles had their reception disturbed by the main ones transmitting in close proximity and getting some separation helped.

There's no obvious technical reason for the audience goggles to have worse reception, but it's how it is anyway...
DJI have had such issues basically on all their products that support dual video receivers all the way to the Inspire 1. At the time they mentioned it was related to synchronization/encryption IIRC.

I've suspected this could be a linking/encryption problem for a while, but only in a vague sense.

Having the 2nd goggles get their signal from the primary goggles strikes me as a profoundly lame architecture for that process. Is that really how it works? Why wouldn't they just have the 2nd goggles pull the signal directly from the AC? If the second goggles aren't transmitting, there would be no bandwidth competition at all.

Trying to get this sorted out has been a frustrating problem. What I'm looking for is:

1) How severe is the problem, and how common is it? A proper answer will be a statistical answer, which is fine with me.
2) What specific steps, if any, can be taken to mitigate the problem?
3) What's the root system architecture cause of the problem? WHY is there a problem?
4) How hard would it be to correct/mitigate that system architecture problem? Specifically, might there be a firmware solution, or does it require different hardware?

If anyone has any knowledge on any of these, I'd greatly appreciate you sharing it!

I'll contact DJI about this directly.

Thx,

TCS

 

[Kilrah]

Having the 2nd goggles get their signal from the primary goggles strikes me as a profoundly lame architecture for that process. Is that really how it works? Why wouldn't they just have the 2nd goggles pull the signal directly from the AC?

The talk back then was that the secondary receiver DOES receive the feed straight from the AC, but also needs to receive synchronization/decryption information from the master receiver at the same time, and if that doesn't go through what's received from the AC can't be decoded.

1) How severe is the problem, and how common is it? A proper answer will be a statistical answer, which is fine with me.

The "audience" goggles will sometimes have a somewhat distorted picture and more dropouts than the main, but it does the job of being able to see what happens.

 

[Chaosrider]

The talk back then was that the secondary receiver DOES receive the feed straight from the AC, but also needs to receive synchronization/decryption information from the master receiver at the same time, and if that doesn't go through what's received from the AC can't be decoded.

If that's the case, then something interfering with the signal path could cause the problems that have been reported. Perhaps, it works great when the signal path is operating normally. If true, this could point to possible solutions/mitigations.

The "audience" goggles will sometimes have a somewhat distorted picture and more dropouts than the main, but it does the job of being able to see what happens.

I'm looking to get a bit more of a quantitative sense of how bad this is, and how often it happens.

Thx!

TCS

 

[EpicFlight]

The talk back then was that the secondary receiver DOES receive the feed straight from the AC, but also needs to receive synchronization/decryption information from the master receiver at the same time, and if that doesn't go through what's received from the AC can't be decoded.

Speculatively, that makes a lot of sense, given the security and encryption on Occusync. The video feed from the AC is encrypted with a key exchanged during binding, IIRC.

 

[Chaosrider]

Speculatively, that makes a lot of sense, given the security and encryption on Occusync. The video feed from the AC is encrypted with a key exchanged during binding, IIRC.

I've had a sense for a while that it's some kind of signal encoding problem. If that's true, then there's no reason in theory why the security and encryption couldn't be sent in its full glory to the audience goggles as well. Essentially allowing two goggles to fully link to the collective at the same time.

It strikes me that that could be a firmware fixable problem, without necessarily requiring new hardware.

But that's just a guess.

Thx,

TCS

 

[Kilrah]

If that's true, then there's no reason in theory why the security and encryption couldn't be sent in its full glory to the audience goggles as well. Essentially allowing two goggles to fully link to the collective at the same time.

While pairing you'll define a master key, but then there will likely be a temporary key based on that that's changed every X packets, each time the master goggles sends the new one to the audience goggles so it can decrypt the feed, that way the audience goggles (or any other device trying to snoop the signal) never have a way to either view the signal or take control without authorization (when audience mode is turned off the keys aren't sent out -> nothing).

There no doubt are a lot of implications to consider around the system's design. Even if it's acknowledged as a problem for audience mode it might not make sense redesigning fundamental parts of the system to improve that given all of the other changes that might cascade from it since DJI have used minor variants of the same thing for years now, sell things like Aeroscope that must be compatible with all their products, etc...

I do hope it improves one day, and being able to "permanently" link 2 sets of Goggles like on the Mavics would be useful for the use you're after. But given how the current system is designed it was obviously not devised with the "2 people watching a good stream next to each other" in mind, but with something like a drone race where one guy is flying and 100 people should be able to view it i.e. manual pairing isn't an option, even if it causes quality issues since that was all what the "hype" was about around 2-3 years ago when they likely started working on the FPV.

They might have missed the target and more people actually want the former than the latter, but whether they choose to recognise it or not, who knows. DJI don't exactly have a track record of listening to what people want, they have more of a history of Apple-like "this is what we decide you need, can/can't do, etc".

 

[EpicFlight]

What would make the most sense would be an HDMI out port on the goggles.

But that makes sense.

 

[Chaosrider]

While pairing you'll define a master key, but then there will likely be a temporary key based on that that's changed every X packets, each time the master goggles sends the new one to the audience goggles so it can decrypt the feed, that way the audience goggles (or any other device trying to snoop the signal) never have a way to either view the signal or take control without authorization (when audience mode is turned off the keys aren't sent out -> nothing).

This sounds like a signal configuration/security issue to me. I worked in IT security for a long time. It seems to me that it shouldn't be that difficult to give all the decryption keys needed to a second set of goggles, so that it could receive the signal in parallel to the primary goggles. It seems to me that this could be accomplished with a relatively minor change to the signal path architecture for the second set of goggles, and to the linking process, that would make this fixable via a firmware change.

Of course, not knowing the details of the current architecture, I can't really estimate how much work would be involved in implementing the fix.

There no doubt are a lot of implications to consider around the system's design. Even if it's acknowledged as a problem for audience mode it might not make sense redesigning fundamental parts of the system to improve that given all of the other changes that might cascade from it since DJI have used minor variants of the same thing for years now, sell things like Aeroscope that must be compatible with all their products, etc...

I just had exactly that experience with my home IT-AV infrastructure. First I had to face the fact that I needed a much faster computer, vs my 2013 laptop, for editing my ever-expanding volume of drone vids. I had an old Sony 60" TV that was somewhat high end at the time (2013), but it was only 1080p. It died recently, and it was clear that I was going to replace it with a 4K TV. So, of course, I had to get a 4K Ultra HD disk player...

It was an annoyance, and ended up costing a lot more than I wanted, but now my IT-AV infrastructure has moved to a much higher plane of existence.

I do hope it improves one day, and being able to "permanently" link 2 sets of Goggles like on the Mavics would be useful for the use you're after.

HOLD THE PHONE!

You can link two sets of goggles to the Mavics now? Directly, with no signal degradation to the second set? If that's true, I may need to look in that direction. I had no idea that was an option!

DJI don't exactly have a track record of listening to what people want, they have more of a history of Apple-like "this is what we decide you need, can/can't do, etc".

There are many reasons why I don't use any fruity computing or comm gear. I'm strictly a PC/Android guy. On a dimly related note, virtually all new TVs have either Android or iOS operating systems. Sony is my preferred TV vendor, and I'm very happy they went with Android.

My DJI FPV purchase is on indefinite hold until I understand this audience video quality issue a lot better than I do now. I may still decide it's worth doing with the as-is FPV hardware, but I don't yet have close to enough information to make that decision.

Thx!

TCS