Video quality (jello)

Hi, I’m new to drones. I found the videos taken from Viper2450 GPS are extremely “jello”. Do you guys see the similar results. There is a youtube clip (https://www.youtube.com/watch?v=_AMisCTAzN4&lc=z23by5apqtzxjngpm04t1aokg2cxu15xnqmrweif3xakbk0h00410.1509109280821051), the author stated “It seems that the v2450 (no GPS) has little to no jello…so maybe the GPS hold or compensation is causing the frame to vibrate, thus causing the jello…”

What do you think? What would be the main cause(s) of this “jello” effects?

alandal, If you connect to the drone and do a Motor Calibration with your finger holding the drone down on a rubber mat, you can feel the relative vibration as each motor assembly as the calibration process continues to completion. The left front motor assembly of mine vibrated much more than the other three.

Thanks John. I did a short test running the in-door mode. The footage was much smother that the GPS mode … And how is 2450FPV model, does it have the same flying assembly?

What you described may be somewhat indicative to a fact I observed. My 2450GPS has a tendency to drift towards forward-left (after repeated calibrations). But I don’t know if your finding is consistent for this model. If we assume all flying components are randomly selected on the assembly line, may it be a indication that the overall frame design is not well balanced?

alandl, I’m only interested in using this product in GPS Mode, and my expectations are very low for quality video from anything with this kind of [motor/gears/shaft/prop] on the end of the body, and at such a low price for ArduCopter functionality. While there can be variation in items within a manufacturing batch, I suspect that having balanced props remains the most effective way to reduce vibration on this product. To test that assumption, I swapped the left-front prop (the noisiest by far during my Motor Cal test) with the right-rear prop. Most of the vibration shifted to the right-rear. You might want to swap out props from the spare set to see if your video quality improves.

There’s not much to isolate the camera from the drone body, so it’s possible that a physical hack with softer foam could reduce vibration transfer to the camera.

Rather than speculate further, I’ll leave it up to a SkyRocket rep to respond.

1 Like

Thanks again John. Point taken. My concerns are relating more to the user experiences than to ArduCopter/SkyViper development. We should end this conversation here.

I’ll try to swap props as you suggested … Thanks for that!

I just ordered 5 sets of props, so I can swap out for least vibration. If you want to dive deeper and learn how to connect to a SkyViper network to use the Motor Cal function through the web server, a place to start is here.

Thanks John. I’ll give a try… Saturday, 28 October 2017, 11:54AM -04:00 from John Githens noreply@ardupilot.org :

To anyone trying to minimize vibration from unbalanced props on this GPS SkyViper: I bought 5 sets of props (offered for the non-GPS version of a ‘2450’ SkyViper). Using a Win10 laptop, I waited to get a connection to the SkyViper network, opened Edge, entered 192.168.99.1, selected the Calibration page, then the MotorCal tab. I placed the SkyViper on a thin rubber mat and held it down with a finger while running either of the tests. By swapping a prop and running a test, I could roughly rank the level of vibration across a row of props, then pick the one that seemed to be the best. (This process would produce faster, better results if there was also a test only for each motor (rather than all four in sequence).)

If your results are not good enough, and you don’t mind different prop colors in non-stock positions, you can swap a green “A” prop for a black “A” prop. (Note the A or B symbol on the underside. There is also a number that may represent a manufacturing batch.)

Overall, my results were good enough for me, and noticeably better than out-of-the-box.

Other things that might contribute to “jello” in video clips: loose fit of metal motor shaft in plastic column; variation in closeness of plastic gear mesh; tight clamping of the video camera in the SV body; you did not fasten the prop screw securely; you or a flying insect harmed the prop in a previous flight.

2 Likes

Now that I’ve explored more of the SkyViper’s extensive web app… I see that any individual motor can be tested by going to the Motors page from the home page. :slight_smile: