- Joined
- Jan 21, 2019
- Messages
- 369
- Reaction score
- 147
Hey all I did a test to confirm @raulvolador ’s claim on “separation between phone and repeater = better vid”. It was pretty windy and my tello drifted many times but I have the results in.
I flew diagonally across my local football field which is 120m one way. Then return. Height maintained at 5m (it was windy so at times I had to adjust it). I recorded the trips with 3 setups.
1, plane mode bt off just control by phone.
2, plane mode with t1d connected via bt, fly with repeater bounded to the bt controller.
3, same as 2 except I had the repeater hanging on the fence 3m to my right.
I used iphone x running 1.3.0 offcial app, 1.11 fw for t1d and 01.04.78.10 fw for tello. Tello was always rebooted after all other devices before each test.
The first setup had nice vid until 50m then it all breaks down, at 60m I wasn't getting any video feed and I had to turn around. So first lesson, use a repeater haha. I deleted the vid, useless. Due to time constraints I missed the repeater + phone control tests. In pursuit of the best performance and practicality I will try carry that and more tests out tomorrow. Day 1 tests are conclusive enough that having the repeater away is better. Therefore day 1 tests will not be repeated on day 2 and the day one results cannot be directly compared to day 2 results. PS: I repeated day 1 tests and got less differentiating results than day 1.
The for the vids I upload. I counted frame skip for them, 1st vid was on setup 2. 2nd vid was on setup 3 suggested by Raul. 3rd vid was a repeat of setup 2. Notice how sometimes the vertical angle shifts, it is a result of EIS shifting the frame in high wind, not a frame skip due to interference so this does not count.
Day 1
1st vid 6 skips (Setup 2, repeater bounded to controller)
link
2nd vid 1 small skip while turning (Setup 3, away from repeater)
link
3rd vid 3 skips (repeat of setup 2)
link
So needless to say. I think Raul's suggestion worked nicely! I had the least frame skip when the repeater is away from me. But would I be bothered to carry a stand with me or seek a place for hanging repeater every flight for the sake of avoiding a few frame skips? Probably not... I would try hanging the repeater when it is practical though. What you do is entirely up to you!
Day 2 [With bt controller: Head mount, chest mount, gound placement 3m away and at my feet. Without bt controller: repeater 3m away hanging on a fence and then very close to the phone.]
*all other conditions are the same as day 1
**45 deg back looks like ( /me), 45 deg back 3m away looks like ( /____________me)
It turned out that 3m away from the repeater on the ground was pretty horrible and repeater on the ground in general isn't too good. So I pushed the repeater about 3m forward to get to a slightly higher ground spot that is more level to the rest of the field. 4 more tests below for that.
Hey being on the ground isn't too shabby when you stand next to the repeater!
I decided to run the same tests I did on day 1 since I had the time. Plus the phone + repeater results. Very close means repeater as close to the phone as setup 2, far means 3m apart. Both close and far are at chest height.
Final verdict: Best result is a achieved by running the setup at chest height, having the repeater 3m away maybe marginally better (a few extra skips over a 120m long return flight in slow mode). Ground placement is usable but I noticed noticeably weaker signal compared to chest height at the 120 mark although the number of frame skips aren't so much different (1 skip difference). 45 deg angle could help with high altitude signal (need to test this) not so much more helpful at 5m height than vertical. At least on iPhone X having the bt controller connected doesn't make video noticeably worse than using the phone alone. Given the convenience, tactile feeback and comparable (to setup 3) video quality of setup 2 shown by results above, I would prefer to use setup 2 all the time. I consider the improvement by repeater separation marginal.
Disclaimer: This experiment was by no means scientific despite my effort to approach that. Ideally one needs many more repeats to obtain a convincing result. The large discrepancy between day 1 and day 2 in terms of the setup 2 & 3 comparison indicates there is still a chance where there is no statistical difference between the 2 top setups.
Update: I had a chat with a Rhode & Schwarz technical officer who is very knowledgeable about RF antennae. He explained that when the receiver is too close (half lambda) to the antennae it is in what is called near field regime where it cannot "see" the transmission wave very well. The separation is smaller than half of the RF wavelength which is about 6.1cm. In my setup 2 the rough antennae separation is about 13cm from xiaomi logo on the repeater to the centre of my iPhone X's wifi antenna. So I think the separation is large enough in my setup.
I flew diagonally across my local football field which is 120m one way. Then return. Height maintained at 5m (it was windy so at times I had to adjust it). I recorded the trips with 3 setups.
1, plane mode bt off just control by phone.
2, plane mode with t1d connected via bt, fly with repeater bounded to the bt controller.
3, same as 2 except I had the repeater hanging on the fence 3m to my right.
I used iphone x running 1.3.0 offcial app, 1.11 fw for t1d and 01.04.78.10 fw for tello. Tello was always rebooted after all other devices before each test.
The first setup had nice vid until 50m then it all breaks down, at 60m I wasn't getting any video feed and I had to turn around. So first lesson, use a repeater haha. I deleted the vid, useless. Due to time constraints I missed the repeater + phone control tests. In pursuit of the best performance and practicality I will try carry that and more tests out tomorrow. Day 1 tests are conclusive enough that having the repeater away is better. Therefore day 1 tests will not be repeated on day 2 and the day one results cannot be directly compared to day 2 results. PS: I repeated day 1 tests and got less differentiating results than day 1.
The for the vids I upload. I counted frame skip for them, 1st vid was on setup 2. 2nd vid was on setup 3 suggested by Raul. 3rd vid was a repeat of setup 2. Notice how sometimes the vertical angle shifts, it is a result of EIS shifting the frame in high wind, not a frame skip due to interference so this does not count.
Day 1
1st vid 6 skips (Setup 2, repeater bounded to controller)
link
2nd vid 1 small skip while turning (Setup 3, away from repeater)
link
3rd vid 3 skips (repeat of setup 2)
link
So needless to say. I think Raul's suggestion worked nicely! I had the least frame skip when the repeater is away from me. But would I be bothered to carry a stand with me or seek a place for hanging repeater every flight for the sake of avoiding a few frame skips? Probably not... I would try hanging the repeater when it is practical though. What you do is entirely up to you!
Day 2 [With bt controller: Head mount, chest mount, gound placement 3m away and at my feet. Without bt controller: repeater 3m away hanging on a fence and then very close to the phone.]
*all other conditions are the same as day 1
**45 deg back looks like ( /me), 45 deg back 3m away looks like ( /____________me)
Setup | Head mount | Chest mount | Ground 3m vertical away | Ground vertical at feet | Ground 45 deg back 3m away | Ground 45 deg back feet | Ground 45 deg back 3m away replicate |
Frame skip count | 7 | 4 | 10 | 4 | 10 | 5 | 12 |
Video link | link | link | link | link | link | link | link |
It turned out that 3m away from the repeater on the ground was pretty horrible and repeater on the ground in general isn't too good. So I pushed the repeater about 3m forward to get to a slightly higher ground spot that is more level to the rest of the field. 4 more tests below for that.
Setup | Ground 3m vertical away | Ground vertical at feet | Ground 45 deg back feet | Ground 45 deg back 3m away |
Frame skip count | 4 | 2 | 2 | 6 |
Video link | link | link | link | link |
Hey being on the ground isn't too shabby when you stand next to the repeater!
I decided to run the same tests I did on day 1 since I had the time. Plus the phone + repeater results. Very close means repeater as close to the phone as setup 2, far means 3m apart. Both close and far are at chest height.
Setup | Setup 2 | Setup 3 | Phone & repeater very close | Phone & repeater very close replicate | Phone & repeater 3m apart |
Frame skip count | 3 | 2 | 3 | 2 | 4 |
Video link | link | link | link | link | link |
Final verdict: Best result is a achieved by running the setup at chest height, having the repeater 3m away maybe marginally better (a few extra skips over a 120m long return flight in slow mode). Ground placement is usable but I noticed noticeably weaker signal compared to chest height at the 120 mark although the number of frame skips aren't so much different (1 skip difference). 45 deg angle could help with high altitude signal (need to test this) not so much more helpful at 5m height than vertical. At least on iPhone X having the bt controller connected doesn't make video noticeably worse than using the phone alone. Given the convenience, tactile feeback and comparable (to setup 3) video quality of setup 2 shown by results above, I would prefer to use setup 2 all the time. I consider the improvement by repeater separation marginal.
Disclaimer: This experiment was by no means scientific despite my effort to approach that. Ideally one needs many more repeats to obtain a convincing result. The large discrepancy between day 1 and day 2 in terms of the setup 2 & 3 comparison indicates there is still a chance where there is no statistical difference between the 2 top setups.
Update: I had a chat with a Rhode & Schwarz technical officer who is very knowledgeable about RF antennae. He explained that when the receiver is too close (half lambda) to the antennae it is in what is called near field regime where it cannot "see" the transmission wave very well. The separation is smaller than half of the RF wavelength which is about 6.1cm. In my setup 2 the rough antennae separation is about 13cm from xiaomi logo on the repeater to the centre of my iPhone X's wifi antenna. So I think the separation is large enough in my setup.
Last edited: