Welcome to our Community
Wanting to join the rest of our members? Feel free to sign up today.
Sign up

Anafi in extremely high winds

Did its own foot enter the frame at 0:42? If so it looks incredibly steadily relatively speaking! I wonder how much more it could have taken before you lost it to be honest.
 
  • Wow
Reactions: Krusty Geeza
Krusty Geeza said:
Thanks for mentioning the deshaker tool, Peter, I'll look into it for possible usage on more stable video. This footage, where the poor Anafi was being moved around more than a metre at a time by the gusts is beyond that sort of help because of the resulting perspective changes.

I've done the walking experiment, using a hand-held gimbal for my mobile phone & then without. Stabilizing software is useless when you bounce as you walk! A gimbal helps to lessen that at capture, but even a gimbal can't overcome someone walking with a bounce as they are moving up & down with each step, & therefore the camera may also be moving up & down with each step too. It's not something that any software can correct, as your point of view is repeatedly going up & down slightly, so therefore the perceived line of sight & perspective of everything is changing too with each step taken. That is impossible to correct after capture! Only walking intentionally to hold the camera at a certain smooth height will achieve smooth footage or minimise this. I like a little bounce to show that you are in fact walking, as you mentioned; many don't, preferring the floating look.

Depth perception requires two eyes to work effectively, it's a whole different thing to image stabilisation, but still very interesting in how different animals perceive depth due to their differently located eyes that look in different directions. Most birds, having eyes that can only slightly look forwards, have to look through each eye repeatedly to achieve any real sense of depth as their eyes look more sideways rather than straight ahead like ours, or dogs or cats.
Click to expand...
 
Zamzara said:
Did its own foot enter the frame at 0:42? If so it looks incredibly steadily relatively speaking! I wonder how much more it could have taken before you lost it to be honest.
Click to expand...
It did appear at 0:28 - 0:30 & again at 0:42 in the footage I shared, but I have a bunch more times it happened, in the cutting room floor film! The foot appears, not due to the arms closing, but rather the severe angle she was on, fighting the winds.

I was keeping an eye on her the whole time, when she statrted not being able to move into the wind while on Film mode, I kicked her over into Sports mode. I was well above everything & watching her intently. Had she have started to go backwards at any point I would have landed her immediately. While there were gusts, it wasn't squally, if that makes sense?
 
Krusty Geeza said:
... Only walking intentionally to hold the camera at a certain smooth height will achieve smooth footage or minimise this. I like a little bounce to show that you are in fact walking, as you mentioned; many don't, preferring the floating look.

Depth perception requires two eyes to work effectively, it's a whole different thing to image stabilisation, but still very interesting in how different animals perceive depth due to their differently located eyes that look in different directions. Most birds, having eyes that can only slightly look forwards, have to look through each eye repeatedly to achieve any real sense of depth as their eyes look more sideways rather than straight ahead like ours, or dogs or cats...
Click to expand...

Though even we can see depth out of movement.
Depth can be precievd out of movement, geometric futures and stereoscopic view, most of the time we make up a environment out of these 3 futures

A friend of my who has two eyes, but his brains process depth out of movement only, he doesnt have stereo scopic view like most people.
He doesnt like 3d movies, as he doesnt see depth in it, one eye is dominant
 
  • Like
Reactions: Krusty Geeza
PGTART said:
Though even we can see depth out of movement.
Depth can be precievd out of movement, geometric futures and stereoscopic view, most of the time we make up a environment out of these 3 futures

A friend of my who has two eyes, but his brains process depth out of movement only, he doesnt have stereo scopic view like most people.
He doesnt like 3d movies, as he doesnt see depth in it, one eye is dominant
Click to expand...
There was a documentary I watched not too long ago, but can't remember the name of now, I'm sorry. In it they mentioned what you have spoken about here & they also mentioned that there is evidence that the brain makes up certain aspects of what we see! This applies to the form, dimension & colour of things amongst others!

The ability for the plasticity of the brain to fill in the blanks is evident in your friend developing or increasing his ability to see depth based solely on movement. This is a fascinating time to live in as more & more is being proven about how we work!
 
I'd like to thank @h-elsner for the assistance offered me on the Show Anafi Log thread. (y):)

I would like to have an own thread for problems during flight. This one is dedicated to the tool itself.

The JSON file looks like there is only one flight, maybe the last one (2019-04-05 10:25:29 to 10:34:36, Duration 00:09:03).
Screen_Chart altitude_1.pngkml_1.jpg
You have an older firmware running. The 'True Air Speed' is empty.

Roll angle is between -40° and +40°, same for pitch angle. Seems to be the maximum. Good flying. Speed up to 12m/s.

Screen_Charts attitude_1.png

br HE
Click to expand...

That's what I discovered too! I did finally manage to get it to load & work on the Linux laptop!

The firmware's for the Anafi, Skycontroller & batteries are all as up to date as they can be on FF6 iOS. This flight was performed Friday, I did the FF6 iOS update this morning, but there was no update for the Anafi or Skycontroller in that.

This was actually the first of three flights, working through the three batteries rather quickly. I have always had problems with FF6 not remembering all of my flights, maybe the latest update will change that, as they were all there, from this mornings flights as I just checked them then.

It looks like poor young Ms Anafi the 2nd was working hard in these winds on Friday! :eek:

Pitch Angles:
1784

Roll Angles:
1785
 
I think it's most noticeable in the portion where you're following the chimney straight up. Less noticeable and not really distracting (IMHO) when you're circling around objects. That said not too bad for such a light drone.
 
  • Like
Reactions: Krusty Geeza
Went to one of my own movies, because i was curious about deshaking anafi with vritualdub and deshaker.
Oh and there is some good news there is port virtualdub2 (hyperlink) that can read the MP4 file the anafi creates.
In the past i had to convert my canon movies to avi first.
The free deshaker plugin for virtualdub2 can be found here

I was just curious processing an Anafi mp4 with it,
Well i noticed that the correction for rotation (2d tilt of video) is not needed (the anafi is allready realy good at it)
I did notice some improvement when turning (z axis anafi rotations), and left right movements.
Under windy condition, slight random movement is reduced as well (but can depend on your configuration of deshaker focus area as as well).

Conclussion overall some errors can be corrected,
But the anafi doesnt make as much errors as my human hands when filming.
So unless those errors are disturbing you, most footage will not improve a lot by it.

Another thing I noticed is for a recompile to a new video (ea assign a codec), that my quality went down a lot in my first attempt. The reason for that was that 4K movies require more bits/sec then my non 4K movies.
And with that i also wondered if the quality of 4K hits some limit with 100mbit/sec (maybe the anafi takes better quality at less detailed settings...hmm might be maybe something to test a next time.


Another thing i wondered, but pure technically programmable math.
Is that the kind of movement shakes might be used in a layered deshake function.
In which one could record previous frames (and due to parallax movement) a history for pixels could be build.
In such a way that short shakings can have their pixels resolved from previous pixel locations.
Not seen any deshaker working like that though (it would be kindof an instant depth reconstruction).
Those are times i wished i had more knowledge about writing filters for virtualdub.
 
  • Love
Reactions: Krusty Geeza
Zamzara said:
I wonder how far it can tilt and still lift its own weight?
Click to expand...
I'm afraid we're going to have to wait for one of the crazy YouTube reviewers that keep pushing the Anafi as far as possible to find that out. I won't be trying it again under these sort of conditions. :oops:

pmh1nic said:
I think it's most noticeable in the portion where you're following the chimney straight up. Less noticeable and not really distracting (IMHO) when you're circling around objects. That said not too bad for such a light drone.
Click to expand...
I agree. I am also most impressed with her abilities for such a tiny, light thing! :love:

PGTART said:
Another thing i wondered, but pure technically programmable math.
Is that the kind of movement shakes might be used in a layered deshake function.
In which one could record previous frames (and due to parallax movement) a history for pixels could be build.
In such a way that short shakings can have their pixels resolved from previous pixel locations.
Not seen any deshaker working like that though (it would be kindof an instant depth reconstruction).
Those are times i wished i had more knowledge about writing filters for virtualdub.
Click to expand...
Now this is gold! You're on to something there! :love::cool: It's way beyond me also, but hopefully somebody who can explore it sees this, or it can be shared on a forum for others to explore further.
 
That has been my experience too. Impossible to get usable footage with strong wind. I think that it is mostly the digital 2 axis stabilization that ruins it but might be wrong. I could fly with my mavic air with the same winds and perfect footage. Anafi has a lot of limitations but the weight is great but it comes at a price, flying with wind and taking photos is ok, filming NoNo......
 
mastix said:
That has been my experience too. Impossible to get usable footage with strong wind. I think that it is mostly the digital 2 axis stabilization that ruins it but might be wrong. I could fly with my mavic air with the same winds and perfect footage. Anafi has a lot of limitations but the weight is great but it comes at a price, flying with wind and taking photos is ok, filming NoNo......
Click to expand...
Anafi has a digital 3 axis stabilization plus a 2 axis mechanical stabilization.
 
  • Like
Reactions: mastix
For those of you flying camera drones in strong wind you must investigate windspeed and windshear speeds, speeds might be doable at ground level however go 100 meter higher can easily surpas the speeds of your drone be it DJI or Parot. Just warning you, most beginners don't think about it. But do your research find those tables and fly safe for others around you just saying.
 
PGTART said:
For those of you flying camera drones in strong wind you must investigate windspeed and windshear speeds, speeds might be doable at ground level however go 100 meter higher can easily surpas the speeds of your drone be it DJI or Parot. Just warning you, most beginners don't think about it. But do your research find those tables and fly safe for others around you just saying.
Click to expand...
Note if you use a racing drone you might be more safe if the drone can do 80 mph wind of 30 mph then a camera drone that can go max 40
 
Wow i admire your balls(so to speak ??)! I dont fly in anything above 15mph/25kmph.
 
I was wondering, looking at your video, if the feet were in the frame due to extreme rotations of the Anafi fighting against the wind (so the gimbal ‘looks at the feet’ to maintain an horizontal view) or if it was the shadow of the feet that are captured (you had long casted shadows if I refer to the chimney views this day...) ?
 

New Posts

Members online

No members online now.
Total: 170 (members: 0, guests: 170)

Forum statistics

Threads
5,313
Messages
45,116
Members
8,013
Latest member
Ritchh