Hello Fellow Yuneec Pilot!
Join our free Yuneec community and remove this annoying banner!
Sign up

High Rise

Joined
Jul 6, 2018
Messages
991
Reaction score
647
Location
UK
60 flights in with the TH so far, and levels of trust are high, and so I am gradually working my way over the South Downs, and ultimately up to the highest hill in Hampshire (Butser). And I'm getting close now - only 800m away from the summit this time - I have at least made it to the ridge of the hill.

The Typhoon H is an unknown quantity in proximity to that massive transmitter mast up there on the far hill, so I wanted to do a 'scout flight' where we fly in far proximity of it, and monitor for problems, with zero risk to the mast itself lest anything go horribly wrong. I was also on extra high alert by the KP index of 4, which is the maximum I will fly in, so was ready to kill GPS at a second's notice...

But there weren't any signal or video problems - not one, and indeed the H held its own, even at 400 ft and remained flawlessly nailed to the sky and fully maneuverable in some pretty furious wind once we got above the treeline and out into valleys. What was 5 mph on the ground was about 6 times that further up...

Even though we're pretty much as high as we can legally get, and over some suitably epic scenery, this vid isn't a great win, as I had driven round for an hour first trying to find somewhere else to fly, and ultimately by the time I arrived here had left it a little too late, light was difficult, and I had an ND4 on (and a rather inferior one at that !) that I probably should have removed before I started.

From this video we can learn the following:

1. Neewer filters really are rubbish in contrasty light conditions, and I am now fairly sure they are not a properly neutral density evenly across the filter. And that non-uniformity results in a blue-ish cast that gives slightly unnatural results when you try and grade it out...
2. Probably should have left an ND off altogether
3. Next time get up earlier and go in the day :cool:
4. Don't do one flight in Sunset / Sunrise WB Mode, then another one in Cloudy - it will make matching shots in the grade much more annoying. Idiot :rolleyes:

But never mind - it's good to admit these mistakes so we can all do better next time. Such as it is, enjoy :)

 
Last edited:

Steve Carr

Missed Approach
Staff member
Premium Pilot
Joined
Dec 15, 2015
Messages
8,272
Reaction score
4,754
Location
Bessemer, MI
It's a beautiful location and nice flying and editing. I usually watch these in full screen and I was surprised that detail was poor in many parts of the video. I'm wondering if you had the bit rate set too low when rendering?
 
  • Like
Reactions: AeroJ

FlushVision

Premium Pilot
Joined
Jun 19, 2016
Messages
3,095
Reaction score
1,555
Age
66
Location
Shaw, Gtr Manchester, U.K.
I have a job coming up in the next few days for a conservation group to take some pictures of some moorland so that they can plan some tree planting. But the area is pretty close to a couple of transmitter masts. How far away from that mast do you think you were?
 
Joined
Jul 6, 2018
Messages
991
Reaction score
647
Location
UK
It's a beautiful location and nice flying and editing. I usually watch these in full screen and I was surprised that detail was poor in many parts of the video. I'm wondering if you had the bit rate set too low when rendering?

Yes I thought that too ! It was rendered at 20,000 kb /sec, which Youtube reduced to 12,000 !
 
Joined
Jul 6, 2018
Messages
991
Reaction score
647
Location
UK
How far away from that mast do you think you were?

About 800 meters. but previously I have flown to within 50 meters of it with DJI gear, years ago, which at one point did an uncalled RTH, I suspect as result of interference from the tower. Fortunately I was expecting that so was careful not to ever fly so the tower was between me and my craft, hence it didn't get ploughed into when that happened, and everything was fine !
 
  • Like
Reactions: FlushVision

Steve Carr

Missed Approach
Staff member
Premium Pilot
Joined
Dec 15, 2015
Messages
8,272
Reaction score
4,754
Location
Bessemer, MI
Yes I thought that too ! It was rendered at 20,000 kb /sec, which Youtube reduced to 12,000 !
I have been using 80,000 or more and that seems to make a big difference. It would be interesting to see if that changes the quality of yours.
 
  • Like
Reactions: AeroJ
Joined
Jul 6, 2018
Messages
991
Reaction score
647
Location
UK
OK dokey - see if this one is any better - overall I think it is, but there are still some shots where tree detail seems largely turned into general low-res fuzz...


But I suspect I know, (or at least here are my educated guesses about!) what the main problem was here, and the other reasons for lack of detail....

The ND4 filter gave me 1/125, which looked correctly exposed to me when facing forward and slighty down, but on more steep camera down angles, this was way too dark, and I then had to ramp it up to 1/60 in the air to try and get it bright enough. This would have introduced some detail-removing motion blur because I was shooting at 1080P 60FPS (with the intention of slowing it all down in a 30FPS project file later). Even then it wasn't bright enough, so as the light failed I found myself ramping up the ISO, which started at 100, but quickly had to be moved to 400, and that introduced quite a bit of noise, and reduced quality further...

I can see I am going to have to go back and do this again with all that fixed, and either a better ND filter or no filter at all ! :) Might try the FreeWell's next, as PolarPro are a bit on the ridiculously priced side of things over here in the UK...
 

Steve Carr

Missed Approach
Staff member
Premium Pilot
Joined
Dec 15, 2015
Messages
8,272
Reaction score
4,754
Location
Bessemer, MI
I can't see much improvement. You may be right. Just the poor lighting conditions. How does the original footage look in comparison?

1534788849416.png
 
Joined
Jul 6, 2018
Messages
991
Reaction score
647
Location
UK
I can't see much improvement. You may be right. Just the poor lighting conditions. How does the original footage look in comparison?

Considerably darker, and less contrasty. That's another reason I think we are particularly suffering here - not only do we have all the reasons above, but everything probably got compounded and made worse and more noticeable by the fairly extreme grading that had to happen to get things even as bright as they are. Gain in particular had to be at around 2.0 (200%) in some cases, which is way higher than I'd normally want to push it...

Yep, I think if I just did this in daylight everything would be much better :cool: We know the CGO3+ struggles in low light sometimes, and with all those big dark stormy clouds obscuring the sun, and steady crosswinds subtly jarring the gimbal during sideways movement, it struggled quite a lot. Thanks everyone for looking, and for all the advice - I will do better next time !
 
Joined
Jul 6, 2018
Messages
991
Reaction score
647
Location
UK
The plot thickens ! Just found this very helpful information on another thread, by @Innerworld...

...the video streams (of the CGO3) are maxed out at:

49182 kbps at 3840x2160-30,
31020 kbps at 2560x1440-30,
14980 kbps at 1920x1080-60

So there's another piece of the puzzle - I was also getting the lowest possible bit-rate out of the camera...

Noted.
 

New Threads

Members online

Forum statistics

Threads
20,368
Messages
235,888
Members
25,358
Latest member
crosbo78