Hi all. I'm still very much a noob to FPV (had my first flight with my self-built Martian a couple of months ago), but I have 10+ years experience with sim racing (iRacing specifically), and one of the things I'm very aware of is minimising input lag.

For my first ever quad, I was given a Flysky FS-i6 radio from a friend, and paired it up with an X6B reciever. I have since held an X9D and tried a QX7, which both felt OK, but for me I'd actually gotten to like the smaller sized Flysky, and for me the key thing was I always felt really in control with it. I looked at some latency testing, and in turns out that the Flysky protocol seems to be the least laggy one going, with around 15 ms vs 30-40 ms of the FrSky one. Even crossfire seems unable to compete, although it's faster than FrSky.

The main problem I had with the Flysky was range problems. Range is something I never really wanted to worry about, and so I started considering getting a Taranis radio with a crossfire module, which seems to be the best of both worlds. After some research I eventually settled on the X-Lite for it's compact size, and the R9M module, which seems to be similar to crossfire in terms of latency and range. I built a 2nd quad with identical parts and configuration, and paired it up to the X-Lite.

Here's my issue. To me, it feels like the X-Lite has something like twice the latency of the Flysky, even with the R9M module. I actually did some testing on betaflight, and while I realise betaflight definitely adds extra lag, the Flysky is clearly much, much snappier (100 vs 200 ms): 

This mirrors what I was feeling. Even hovering close to the ground is much easier for me with the Flysky. With the X-Lite I'm fighting to avoid over-correcting and bouncing up and down.

I read in Oscar's article here https://oscarliang.com/x-lite-latency-testing/ that the version of OpenTX used introduces some delay from filtering, and I wasn't sure if it could be entirely down to that? Oscar says Andrey made a custom version of OpenTX without the filtering; any chance of this being made available? I feel like there's more to it than just this though...

One of the things I haven't done yet is to update the firmware on the receiver I'm using, which I should have said is the R9 Slim Plus, with dual antenna's. This will be my next step.

Any help would be greatly appreciated, thanks!

I did some more flying today, and I'm getting used to the X-Lite with the R9M, but it still feels more laggy to me than the Flysky. According to most sources, the R9M should be just as fast as the crossfire protocol, which has had plenty of latency testing carried out. According to this video, crossfire with the Horus X10 (which should be the same as the X-Lite with R9M right?) has an average of 19.5 ms latency, while the Flysky has just 13.65 ms.

https://youtu.be/0RH_TSwT1oM?t=268

Is there any way I can verify the latency I'm getting on my quad without expensive equipment? To me it feels like it must be worse than 19 ms, I can't imagine I'd notice much difference between 13 and 19.

Also, is there any reason why racers don't use the Flysky protocol? It's ridiculously fast compared to anything else out there...

are you using RC_Smoothing in Betaflight? Set it to "Filter", also try turning it off to see if it makes a different.
And make sure you setup channel range to 8 on your Taranis.
https://oscarliang.com/ctt/uploads/2018/...-setup.jpg

There is no quick way of testing latency in RC system without a logic analyser.

The Flysky protocols are very nicely designed. Their drawbacks are mostly in the hardware (gimbal, build quality etc), range is also not ideal, and their telemetry can't compare to Frsky's, and lacking LUA script support.