HD Video Quality, gymbals, and tuning

[Michail_Belov]

I am trying to solve the shakiness/vibrations of the HD Video. Right now I have an old Runcam mounted on a simple sponge, and I have several types of vibrations which are apparent in the video. Since I do not have much experience with copters and video stabilization I wanted to ask here what is the best way of solving it.

I have uploaded a complete video, but for instance 8:10…9:00 has all the different problems with video:

Lake2

I also have these very low amplitude vibrations seen in log:

Some of the shakiness defintively comes from turbulance, I had about 25 km wind with gusts of up to 40 and since I was on the the lee side, there were also constant downdrafts, and updrafts, but I am really interested in eliminating high frequency oscillations only.

The question is this:

• I do not want to use gymbal due to weight, but… would a gymbal solve issues seen in the video? Would gymbal be sufficently fast to react to high frequency oscillations of the copter?
• I have now a simple sponge, but would a different type of vibration mount eliminate the shakiness? I feel that it needs a much softer dampening, like cutting off 80 % of the sponge material, but I maybe wrong.
• I have these high frequency, low amplitude (0.2 degree) oscillations which I can not really tune out (to get rid of them, I need to lower PIDs way below the safe values which are not compatible with flights in mountain terrain), but are they the real culprit for the video quality?

[amilcarlucas]

Yes most issues on your video, are from vibrations.
Add weight to the camera, that will make a quite efficient “poor-man” image stabalizer.

[Michail_Belov]

Thanks, will try doing so. More weight is the same as making the sponge softer…Would a gymbal be able to eliminate these vibrations?

[amilcarlucas]

No its not the same as making the sponge softer. Its better than that.
Goggle “mass dampen system”

No, for high frequency the gimbal itself needs to be dampen.

[rmackay9]

Hi @Michail_Belov,

A gimbal with vibration isolation could definitely help reduce vibration. Perhaps a CADDX gimbal (which can support one of the RunCam cameras although not the one you have in the video) or an all-in-one camera gimbal like the Siyi A8.

Copter’s list of supported camera gimbals is here.

[Michail_Belov]

Thanks for the suggestions. If Siyi could really solve all the problems, I definitively will think about it (I am looking for something like DJI quality), but I must make sure that the copter is rock solid in terms of flight safety, before putting a more expensive camera.

I had only 10 real world flights (although pretty complicated ones, with wind of more than 30 knots), and I still have not flown the copter at high altitudes (above 3500 meters, objective being 7000 m), so I have to wait out for a few more flights.

What I have is not really vibrations, but oscillations, generated by motors, with a frequency of about 5…7 Hz and amplitude of 0.2 max 0.3 degrees. They seem difficult to eliminate, i.e. the PIDs I have come from autotune with agressivenes of 0.1, but even a much softer autotune still resulted in the same oscillations. I tried manually to eliminate them, but that leads to a less stable aircraft. So I prefer to keep it on the overtuned side, but which is much safer to fly.

It seems that the camera comes into resonance, and I am tempted to put a Speedybee FC on the camera to verify the oscillation amplitude (this FC would be used for logging only), because it seems that at some points the camera oscillates more than the copter itself.

I also have vibrations from high speed, but in the video there were almost no examples of that. The vibration come from the airframe and occur at speeds above 80 km/h. I am building another airframe, where I want to separate the arms from the body, with bolts and maybe even with some thin layer of rubber between the arms and the body.

I will also try to play with weight/sponge supports as suggested by Amilcarlucas, I feel that if I increase weight or reduce the supporting crossection of the sponge, I could get a better dampening.

Also, if all else fails, I will simply use two sets of PIDs, activated by a switch, so that when things get hot, I would sacrifice video, but save the copter.

[flat-earther]

have you tried balancing propellers?

[Michail_Belov]

I am 99.99 % sure the oscillations are not due to prop disbalance. I have run and felt by hand the potential vibrations, and there are none. With airplanes, I know pretty well how a disbalanced prop feels like. These are EOLO props, and they are pretty good.

[Michail_Belov]

I tried putting something like this:

Weight 24 grams, that cut transversal (roll) oscillations by more than 70 %, and reduced putch oscillations by about half. The video now is not perfect but quite usable.

I am thinking of making a lightweight cross with weights at the tips and I feel that could eliminate the video vibrations almost completely!

[Michail_Belov]

I think I got it right:

This setup is a 3d printed stabilizer, 19 grams, with 3 x 9 gram weights, for a total of 50 grams. That eliminated about 80 % of the visible oscillations. Only on high speed descent and above 65 km/h horizontal flight slight oscillations can be seen, but it is a far cry from what I had in the lake video…

If I wanted to improve it, I would replace the frame of the stabilizer with carbon fiber to get more rigidity and less weight, and probably would extend it 1…2 cm, out (right now it has a span of 150 mm)

[Tristan_Redish]

Can we get a before and after?

Amazing that simple ideas can work so well sometimes.

What foam did you use?

[Michail_Belov]

OK. I will upload shortly. The foam is a simple carpenter foam used for stools, very similar to the sponges used for dishwashing.

The way it is installed now is below, to get rid of the props, but I still got them in the view…

I do not trust 100% the foam/instant glue combination, so I put extra green ties just in case…

[Adam_Borowski]

That looks like something Bear Grylls would make on desert island for looking some proteins and calories from the sky (sorry, had to ) What is the size of it?

[Michail_Belov]

2.5 kg. Could go as high as 3.3 kg.

Here is the video, it has some passes up to 90 km/h:

- YouTube

[Michail_Belov]

My motto is: it is not how it looks, it is how it flies… so far, I am satisfied, it has a range of 50 km, endurance 70 minutes (under ideal conditions), I am still pushing up the altitude envelope, but I think it should be flyable up to 7000 m.
With the smaller 5S 12000 mah battery it uses only 1000 mah per 1000 meter climb…

[Michail_Belov]

And you know why I want it to be 100% DIY? I have DJI mavic air 2, and the mother f…er, when I got up hiking for 4 hours to a mountain ridge said that I need to log in, without phone signal… I really wanted to smash it! And another story is, I wanted to take video of a ridge (different one), I fly there and the motherf…er say 500 meters, and I need some 40 meters more to arrive at the ridge.

My copter does not have limits…

[Tristan_Redish]

Impressive!

What are you using to get a 50km range?

[Michail_Belov]

I am not sure, what you mean?

I am using this copter, shown in the photo.

With 1 kg battery (5S x 4P), I have a max flight time of 47 minutes and 40 km.
With 1.5 kg batteries (5S x 4P, 5S x 2P), I have max 69 minutes and 50 km.

For cells, it is Samsung Liions, 35E. I bought a cheap batch which has about 10 % less capacity, nominally now 3050 mah as measured (datasheet 3350), at 4 USD/piece (I was warned about it).

Now to get that range you must fly very precisely, and the altitude change must be quite limited, I am using efficiency indicator to set the most economic speed.

In real world, I would say that I would venture to 15 km with smaller battery and 18 km with the bigger one if the wind is below 10 knots.

[Tristan_Redish]

Sorry I meant for RC signal. 50km is alot.

[Michail_Belov]

Oh, 50 km is maximum total flight range, out and back. So I would not go further than 20 km out. In this copter I have a simple LRSExpress 2.4 GHz (Radiomaster), it seems to be good to at least 15 km., with some tweaking I feel its absolute range would be 25 km.

For fixed wing long range flights I use Dragon Link, with correct settings (low update rate of 6 Hz), I got to about 95 km with standard whip antenna. Dragon Link has also a modem, but for long range it must be set to prehistoric baud rates of about 1800…2000 (not 2400). With a simple Yagi, the range can be increased to about 150…180 km. With yagi, I had rock solid signal to 105 km.