Freeflight 6 mod app

JANGER · May 7, 2019

6.5.0 created own folder Android/FreeFlight 6/ and push there subfolders with Flight Plans, 1 json file, 1 jpeg of preview.
I already confirmed (before I saw 6.5.1 available) that the manually edited json allows us to set altitude over 150 meters and such Flight Plan works. I tested carefully, upt to 300 meters, but I am not sure if the speed of flying is calculated correctly and safely in this case.

d95gas · May 7, 2019

@ds2atc They have just released 6.5.1 There was a couple of dodgy bits in 6.5.0

FREEFLIGHT 6 V6.5.1, ANAFI V1.5.4, SKYCONTROLLER V1.5.0
Fixes

Fixed an issue when the GPSlapse progress bar would not behave as expected.
Once Anafi is restarted, it will now boot to the photo mode last used instead of going back to the default mode.
Fixed an issue where the format of Panorama and Burst shots could be inconsistent between the application and Anafi

KGRAT · May 7, 2019

ds2atc first off thank you for a fast new mod FF6 6.5.0 FCC MODE
Just installed it and all updates got installed,payed appz are there also.
Then did 2 test flights one a flight plan 7 nodes about 1.6 miles in total.
The other flight was just me flying about and then run out about a mile,all testing done in a heavy urbun area all worked fine so far?

Signal seems better to me on the new parrot update tbh anyways thx again.

orybarfly · May 7, 2019

KGRAT said:
ds2atc first off thank you for a fast new mod FF6 6.5.0 FCC MODE
Just installed it and all updates got installed,payed appz are there also.
Then did 2 test flights one a flight plan 7 nodes about 1.6 miles in total.
The other flight was just me flying about and then run out about a mile,all testing done in a heavy urbun area all worked fine so far?

Signal seems better to me on the new parrot update tbh anyways thx again.

I'd also like to thank ds2atc for his efforts. When you ran the flight plan, does it still have the small "hesitation" delays when it moves the camera to point to a new POI?

JANGER · May 7, 2019

ds2atc said:
ff6_6.5.0_TESTONLY2.apk

drive.google.com

Could you try this file again?

It could be easier, if you publish what exactly you did changed in xml. To compare with my / our other ideas for modifications.

KGRAT · May 7, 2019

orybarfly said:
I'd also like to thank ds2atc for his efforts. When you ran the flight plan, does it still have the small "hesitation" delays when it moves the camera to point to a new POI?

Everything was fine tbh m8

DeWe · May 7, 2019

It is possible to allowed the installation also on a tablet like a nexus 2013 2nd Edt. the latest FW is something like 6.x (32Bit CPU)?
All newer version of the OS (from XDA) have problems with the live(black) screen, so I must go back to the latest official version.
Uncomfortable the version up from 6.5.x are only available for OS 7.x and higher

Drone Capt · May 7, 2019

orybarfly said:
I'd also like to thank ds2atc for his efforts. When you ran the flight plan, does it still have the small "hesitation" delays when it moves the camera to point to a new POI?

Unfortunately, when using Flight Plan the drone will stop at any waypoint that has a camera angle adjustment associated with it. This has been an on going point of contention with Parrot as I reported this issue nearly 4 months ago, they claim it has been submitted to the developers but there has yet to be a fix published for it.

I submitted the below to Parrot back in mid-January:

I just made a significant discovery about the pausing at waypoints in FlightPlan. The drone seems to only pause at waypoints that have a camera angle adjustment associated with them or waypoints that are sharp turns (90° or greater, which is understandable). On waypoints without sharp turns or camera adjustments, the drone flies right through the waypoint as it should.

Instead of pausing at waypoints with camera adjustments, the drone SHOULD continue to smoothly flyby the waypoint, starting the camera adjustment as it crosses the waypoint and continuing the camera adjustment as it flies to the next waypoint (the angle change and degrees per second change in the camera adjustment would then determine how long the adjustment actually takes). As it is now, the drone stops, begins the camera adjustment and then continues flying (even if the camera adjustment isn't complete). There should be no stop just because a camera adjustment is requested.

I did several test flight plans and have CONFIRMED this is the issue. You can view the flight data for these flights from my data that has been submitted (my drone serial number is: XXXXXXXXXXX) for flights on Jan 20, 2019 at 11:28am, 11:46am & 11:56am (PST), the 11:46am flight had no camera adjustments (and no pauses (except at the 90° turns, which is expected), the 11:56am and 11:28am flights had 3 camera adjustments (and 3 pauses).

I'll be happy to provide any further data needed to help correct this problem.

ds2atc · May 7, 2019

JANGER said:
It could be easier, if you publish what exactly you did changed in xml. To compare with my / our other ideas for modifications.

There's a class called "AnafiGeofecnce" and it handles geofence settings. I modified that class using smali codes.

However it was failed because in somewhat reason it recovered to 150m when the drone is connected or when the other settings were changed.

JANGER · May 8, 2019

Drone Capt said:
Unfortunately, when using Flight Plan the drone will stop at any waypoint that has a camera angle adjustment associated with it. This has been an on going point of contention with Parrot as I reported this issue nearly 4 months ago, they claim it has been submitted to the developers but there has yet to be a fix published for it.

I submitted the below to Parrot back in mid-January:

I do not observe. When I set different camera angle in neightbour waypoits, the camera adjusts smoothly while flying between points. With no stop at the waypoints.
Could you show sample of your Fiightplan settings where you get suchproblem?

JANGER · May 8, 2019

ds2atc said:
There's a class called "AnafiGeofecnce" and it handles geofence settings. I modified that class using smali codes.

However it was failed because in somewhat reason it recovered to 150m when the drone is connected or when the other settings were changed.

That's we know from months

.

I changed google.map.api keycode com.google.android.maps.v2.API_KEY
I changed default country US gsdk_reverse_geocoder_default_country_code
A couple of changes in string names, just for fun.

Anything more, beside trying geofence limits?

orybarfly · May 8, 2019

JANGER said:
I do not observe. When I set different camera angle in neightbour waypoits, the camera adjusts smoothly while flying between points. With no stop at the waypoints.
Could you show sample of your Fiightplan settings where you get suchproblem?

Example: you have a flight plan with total of 5 waypoints and 2 POIs. The first 3 waypoints point to the first POI and then next 2 waypoints point to the second POI. The transition from one waypoint to another for the 3 first waypoints is normal (without delay), but when it reaches the 4th waypoint that the camera must point to the second POI, there is a small delay of 0.5 - 1 sec. During this delay the drone and camera stay still and is very noticeable in the output video.

JANGER · May 8, 2019

orybarfly said:
Example: you have a flight plan with total of 5 waypoints and 2 POIs. The first 3 waypoints point to the first POI and then next 2 waypoints point to the second POI. The transition from one waypoint to another for the 3 first waypoints is normal (without delay), but when it reaches the 4th waypoint that the camera must point to the second POI, there is a small delay of 0.5 - 1 sec. During this delay the drone and camera stay still and is very noticeable in the output video.

interested point of view.

I will test too

ds2atc · May 8, 2019

JANGER said:
That's we know from months .

I changed google.map.api keycode com.google.android.maps.v2.API_KEY
I changed default country US gsdk_reverse_geocoder_default_country_code
A couple of changes in string names, just for fun.

Anything more, beside trying geofence limits?

From version 6.4.1, I did not changed xml files except for android.manifest.
I figured out how to modify smali codes without errors, so I'm focusing on modifying actual codes.

I'm doing these things nowadays.
- Figure out data flow of values(like country code, altitude settings)
- Find the point to pass a fake value to the app
- Add smali codes

For your information, there's AnafiGeofence source code on Parrot Github.
There are 3 kinds of storages used in the geofence code.
- StorageEntry
- Preset Dictionary
- Device Dictionary

My thought was 'altitude range' is stored in the drone because when the drone is connected all values were reset. and I added lines of smali code that intercepts StorageEntry MAX_ALTITUDE_RANGE_SETTING value to 0.5d, 500d.

This one was TESTONLY version.

After some feedbacks, I added more codes to intercept the values from mpresetdict and mdevicedict.

This one was TESTONLY2,3

After some research, what I think now is,
mpresetdict and mdevicedict are stored in the same xml file in shared prefs. It has a preset section and device section. preset section is created by each drone IDs, and device section is a kind of model specific defaults.

so I will try modify the xml file.
I modified it before, but never tested with the code modification.

JANGER · May 8, 2019

ds2atc said:
From version 6.4.1, I did not changed xml files except for android.manifest.
I figured out how to modify smali codes without errors, so I'm focusing on modifying actual codes.

I'm doing these things nowadays.
- Figure out data flow of values(like country code, altitude settings)
- Find the point to pass a fake value to the app
- Add smali codes

For your information, there's AnafiGeofence source code on Parrot Github.
There are 3 kinds of storages used in the geofence code.
- StorageEntry
- Preset Dictionary
- Device Dictionary

My thought was 'altitude range' is stored in the drone because when the drone is connected all values were reset. and I added lines of smali code that intercepts StorageEntry MAX_ALTITUDE_RANGE_SETTING value to 0.5d, 500d.

This one was TESTONLY version.

After some feedbacks, I added more codes to intercept the values from mpresetdict and mdevicedict.

This one was TESTONLY2,3

After some research, what I think now is,
mpresetdict and mdevicedict are stored in the same xml file in shared prefs. It has a preset section and device section. preset section is created by each drone IDs, and device section is a kind of model specific defaults.

so I will try modify the xml file.
I modified it before, but never tested with the code modification.

It looks like the fcc mod freezes on startup already. I tried to use today. FF6 FCCMODE started and frozen. First time it unfozen after a long 5 minutes or tens of screen touchings. A few Android alerts "unresponse" displayed. After that it staryed to work and fcc was used. BUT after battery changed it did not unfrozen next time at all. I waited over 10 minutes and gave up.

Then installed 6.5.1.
It works without freezing but of course it does not show fcc additional channels. But wait, maybe it looks like they release limits, because I flight over 1000 meters without connection broken. Strange. I supppose it was not happend with 6.4...

JANGER · May 8, 2019

ds2atc said:
From version 6.4.1, I did not changed xml files except for android.manifest.
I figured out how to modify smali codes without errors, so I'm focusing on modifying actual codes.

I'm doing these things nowadays.
- Figure out data flow of values(like country code, altitude settings)
- Find the point to pass a fake value to the app
- Add smali codes

For your information, there's AnafiGeofence source code on Parrot Github.
There are 3 kinds of storages used in the geofence code.
- StorageEntry
- Preset Dictionary
- Device Dictionary

My thought was 'altitude range' is stored in the drone because when the drone is connected all values were reset. and I added lines of smali code that intercepts StorageEntry MAX_ALTITUDE_RANGE_SETTING value to 0.5d, 500d.

This one was TESTONLY version.

After some feedbacks, I added more codes to intercept the values from mpresetdict and mdevicedict.

This one was TESTONLY2,3

After some research, what I think now is,
mpresetdict and mdevicedict are stored in the same xml file in shared prefs. It has a preset section and device section. preset section is created by each drone IDs, and device section is a kind of model specific defaults.

so I will try modify the xml file.
I modified it before, but never tested with the code modification.

TESTONLY did not installed in my hardware. FCCMODE did but please read about freezing problems today

ds2atc · May 8, 2019

JANGER said:
It looks like the fcc mod freezes on startup already. I tried to use today. FF6 FCCMODE started and frozen. First time it unfozen after a long 5 minutes or tens of screen touchings. A few Android alerts "unresponse" displayed. After that it staryed to work and fcc was used. BUT after battery changed it did not unfrozen next time at all. I waited over 10 minutes and gave up.

Then installed 6.5.1.
It works without freezing but of course it does not show fcc additional channels. But wait, maybe it looks like they release limits, because I flight over 1000 meters without connection broken. Strange. I supppose it was not happend with 6.4...

Can you check the logcat?