Digital (paired) video tx/rx needed

MattFL

Junior Member
Can anyone direct me to a digital video TX/RX setup? In short, my goal is eliminating interference between multiple cameras, so I figured the best way is to digitize the video and use paired TX/RX to transmit it. Similar to how our control TX/RX work, but for video. Or maybe using BluTooth, etc.. So basically what I'm looking for is a black-box system with analog video input, analog video output and with TX/RX that can be paired so they won't interfere with other nearby systems. Are there any products that can be used to do this?
 

joshuabardwell

Senior Member
Mentor
Nearly all RC-specific video TX/RX technology is analog. The only digital system that I'm aware of is DJI's LightBridge. If interference is a concern, you should be able to get several channels out of the 5 GHz band, but I don't know how many for sure. Some transmitters support 7 or 8 channels; other transmitters support 32 channels, but I suspect that is 32 distinct center frequencies, and the channels themselves overlap so that there are fewer simultaneously-usable channels.

There are two main reasons why RC-related vTX/RX technology is analog. The first is latency. Flight requires minimum latency. It is possible to do a very-low-latency digital system, but it is expensive. Analog systems are relatively low-latency by their nature. The second reason is that an analog system degrades gracefully. You may have a screen that is 90% snow, but you can still make out the horizon and a major obstacle in front of you so you can save yourself from crashing. A digital system goes straight from 100% to nothing. This would be really bad for flying.

Digital wireless AV transmitter and receiver devices exist in abundance, but they are not geared towards RC flight, so they will tend to be large, heavy, and may require some hacking to run off of a plane's battery pack.
 

Craftydan

Hostage Taker of Quads
Staff member
Moderator
Mentor
Some transmitters support 7 or 8 channels; other transmitters support 32 channels, but I suspect that is 32 distinct center frequencies, and the channels themselves overlap so that there are fewer simultaneously-usable channels.

There's power overlap between the 8 channel VTX's, enough so that flying neighboring channels need some filtering to keep from stepping on each other, like reversing antenna polarity.

16 of the 32 channel units are to line up with other OEMs center frequencies (like one band is fatshark/immersion, another is boscam/skyzone, but the freqencies overlap each other). The other 16 . . . well, just more center frequencies in the same region.


Otherwise, I completly agreed about the digital -- Until the can get the range and latency issues sorted out, it's only good for convienice store securtiy cam DVRs . . . but this *is* where we got all this analog gear from . . . maybe someday, but from what I've seen, not yet.
 

bitogre

Member
To my knowledge (which is admittedly limited), there are not any digital video transmitters that are suitable for FPV. There are some used in video production but they are usually too big/heavy for use in FPV. Your broadcast digital TV signal is 19 megabits per second which uses more than 19MHz in RF Bandwidth (compared to 6MHz of bandwidth of SD analog used for FPV) and is compress. The hardware to do that compression has a significant delay (aka lag in the order of 0.5 to 2 seconds). The production video transmitters I was able to find quickly are only good for a few hundred feet and cost well over $1000 for the pair (up to $8000 for the one that is good to 2000' with less than 1 frame of delay).

BlueTooth is not designed for long range (100m max) or high bandwidth application (1Mbits/s max). WiFi would be better suited but even it is not long range without highly directional high gain antennas (which has its own problems). But, even with these technologies, you need to compress the video signal in some way that is low latency (not currently practical on RC planes with off the shelf hardware).

I think the problems of digital video transmissions for FPV can be solved now with current technologies and done in a way to eliminate the need for HAM Radio license to do FPV. But, I think it will take a company specifically designing hardware for FPV use to solve the problems. I do not see any other industry solving the problems in a way that FPV can take advantage of for several years. Also, I suspect that the first time these issues are addressed, it will be an integrated camera and transmitter solution so that appropriate compression and signal processing can be done as the image senor is being read to avoid serializing/serializing and ADC conversion delays).

Finally, why do you want to take an analog input and transmit it digitally? Analog to Digital conversions add to the latency problem. And what are these "other nearby systems" you are trying to avoid interfering with? Maybe if I understood the problem you are trying to solve better, we might be able to work out a better solution without going to a digital video transmitter.

Now, again, I am not an expert. There may be something that is out there that I am unaware of. So, take my information with a grain of salt.
 

Craftydan

Hostage Taker of Quads
Staff member
Moderator
Mentor
Finally, why do you want to take an analog input and transmit it digitally?

I think this is the crux of the issue.

we'd do the conversion to compress the signal so you can use spectrum sharing from a frequency-hopping burst method . . . but that assumes the data can be compressed (not easy, quick or that fruitful as Bitogre mentioned earlier)

Users see the control signal go out as 6-18 analog signals and the digitizing worked sooooo well, but they miss ho little bandwidth they were using -- control signals are REALLY simple. 6-18 10-12 bit messages sent every 22 to 11ms . . . that's *nothing*. oh, and what happens if I miss one? there's another one coming in *at least* 22ms . . . the pilot will *never* know the radio dropped a packet or two or twenty if it doesn't tell him (ok, maybe 20 in a row you'd notice).

With video, it's simply not that easy. each pixel is at least 24b (8b x 3 colors), with even small resolutions demanding *Hundreds of thousands* of pixels, and a refresh rate of that every 30-40ms . . . That's a *LOT* of bandwidth to compress!

It can be done . . . but not easy, quick and small . . . yet.
 

joshuabardwell

Senior Member
Mentor
I think this is the crux of the issue.

Users see the control signal go out as 6-18 analog signals and the digitizing worked sooooo well, but they miss ho little bandwidth they were using -- control signals are REALLY simple.

Bingo! You hit the nail on the head here. You can use spread spectrum to squeeze a LOT of low-bandwidth signals into a frequency space, but as the bandwidth of the signals goes up, the number of channels you can support must go down. Ultimately, Shannon always asserts itself.
 

MattFL

Junior Member
Thanks everyone for the replies, excellent information and points all around. I completely overlooked the latency, that would definitely be a deal breaker for FPV. As would the on/off behavior of digital when the signal gets weak. In this particular case, these won't be an issue, and a range of 30-50' should be plenty. Also quality doesn't have to be top shelf, 640x480 and anything over 20fps should be fine.

The sole problem I'm trying to solve (and the purpose for considering digital) is to have paired TX/RX, so that if another camera system operating on the same frequency turns on nearby, the video from one system won't show on the others screen, as can happen with analog. It's OK if the signal gets bad and the video cuts out a bit, as long as video from one system doesn't show on the other system's screen. I have ordered some 8 channel analog TX/RX to experiment with, but for the final solution 8 channels won't be enough, there's too much chance of having 2 systems on the same channel in this particular setup.

I would love to hear of any digital suggestions if there are any, but assuming digital just doesn't pan out then I wonder if I can accomplish the same goal with a combination of low power multi-channel analog TX/RX combined with directional antennas? The cameras won't be moving quickly so keeping antennas pointed their way shouldn't be difficult.
 

joshuabardwell

Senior Member
Mentor
Directional antennas will have minimal effect over the ranges you are talking about. Directional antennas typically have significant side and back lobes. At longer distances, these lobes become negligible, but at closer distances, signal from the lobes can be plenty strong enough.

Maybe if you gave us a more complete description of your use case, we could help more.
 

MattFL

Junior Member
OK I'll come clean; these cameras won't be flying, they will be on slow moving land based vehicles and there is a high likely hood that multiple vehicles will be operating within 50 yards of each other at random times, sometimes as close as a few feet. The wireless link will be to get video from one end of the vehicle to the other so the driver can see where they're going. The camera will be mounted in an attachment to the vehicle, and for several reasons a wired camera link just won't work. Land based FPV, only the pilot is riding along. ;) They will be operated by people who may be cognitively impaired, so if video from the wrong camera shows up on the screen then the driver may not realize that they're watching the wrong video stream and might continue to drive, and this is why it's important to keep the streams from interfering with each other. And after thinking about it some more, I think latency will in fact be a concern in the final version, but it's not a concern for the first shot at it. The most important challenge at this point is to keep the video streams separate. My thought was that paired TX/RX should do the trick, if I can overcome the challenge of encoding/transmitting/decoding the video digitally. I'm open to any and all suggestions. :)
 

Craftydan

Hostage Taker of Quads
Staff member
Moderator
Mentor
Cognitively impaired? while operating the vehicle they're in?

I'm not sure whether I'm more curious or frightened . . .

Are these "drivers" cronologically challenged as well?
 

MattFL

Junior Member
Cognitively impaired? while operating the vehicle they're in?

I'm not sure whether I'm more curious or frightened . . .

Are these "drivers" cronologically challenged as well?


Indeed they are, lol. I'm not the one deciding who drives, I'm just the guy trying to make it safer. ;)