Taking photos on a Litchi Mission

[Storyline]

I have just flown my first Litchi Mission and it was great. As I predicted I flew far too high as I wanted to make sure I did not run into any of the large trees around my house.

However none of the photos I asked for were taken. At each waypoint - Stop video, Take Photo, Start video. I have done a search and have found some info about having to build in a wait period but no exact information. Does anyone know how long the wait period needs to be after stopping the video before taking the photo ? Is there any other reason my photos did not work ?

Many thanks

 

[parkgt]

You will get better help if you share your mission; hard to guess.

 

[MaviCam]

Very good question. I encountered exactly the same problem last night. I have used nothing but Litchi for the past 3 months and I am finding similar inconsistencies with it. I do not have any plans of going back to DJI GO, but I wish I understood a lot more about what Lichi is looking for me to do in order to get the desired end results. If there are multiple poi'si in a mission, Litchi won't keep poi's in the frame when it should. Why does Litchi allow multiple poi's if it can't deal with them correctly? Maybe it is just that I don't understand how to use the program??? If there is a comprehensive user manual, I haven't found it. I believe a good user manual, that provides details on the use of each feature, would resolve 95% of the problems I have with Litchi.

 

[parkgt]

Very good question. I encountered exactly the same problem last night. I have used nothing but Litchi for the past 3 months and I am finding similar inconsistencies with it. I do not have any plans of going back to DJI GO, but I wish I understood a lot more about what Lichi is looking for me to do in order to get the desired end results. If there are multiple poi'si in a mission, Litchi won't keep poi's in the frame when it should. Why does Litchi allow multiple poi's if it can't deal with them correctly? Maybe it is just that I don't understand how to use the program??? If there is a comprehensive user manual, I haven't found it. I believe a good user manual, that provides details on the use of each feature, would resolve 95% of the problems I have with Litchi.

There are several YT videos that are useful in explaining some but not all of the features.

I do agree that their documentation is scanty and that the framing is pretty sloppy.

 

[dvarapala]

A common noob mistake with Litchi is trying to use waypoint actions with a curved path. When the path is curved the waypoints are never actually reached and the actions don't trigger. Double-check your curved path setting and make sure it's disabled and see if that helps.

 

[gerrymad]

A common noob mistake with Litchi is trying to use waypoint actions with a curved path. When the path is curved the waypoints are never actually reached and the actions don't trigger. Double-check your curved path setting and make sure it's disabled and see if that helps.

Exactly. I did not catch that the first time I used it, but it is a key point. Actions don't happen on a curved path. The drone will not pause or take pictures etc. I've found Litchi to work very well once you understand how to transition from one point of interest to the next. I've gotten some really great looking movements around the points.

 

[gerrymad]

Very good question. I encountered exactly the same problem last night. I have used nothing but Litchi for the past 3 months and I am finding similar inconsistencies with it. I do not have any plans of going back to DJI GO, but I wish I understood a lot more about what Lichi is looking for me to do in order to get the desired end results. If there are multiple poi'si in a mission, Litchi won't keep poi's in the frame when it should. Why does Litchi allow multiple poi's if it can't deal with them correctly? Maybe it is just that I don't understand how to use the program??? If there is a comprehensive user manual, I haven't found it. I believe a good user manual, that provides details on the use of each feature, would resolve 95% of the problems I have with Litchi.

You need to be sure you specify which POI you want for each movement. Also I made a mistake once with a tall narrow tower. I was using the satellite image and the POI that I tagged was where the top of the tower appeared. The issue was that due to the angle of the satellite image the top of the that I tagged was about 30 feet away from the base so in effect I was focused on a point about 30 feet from the tower. The end result was actually (accidentally) a beautiful shot as the Air went around the front of the tower then the tower went out of frame only to slide into the frame quite closely as the Air went behind the tower.

 

[MaviCam]

I use Litchi on my Spark, Phantom 4 Pro and Mavic pro with similar unexpected results.
As for the ONE point of interest limitation, maybe the response I got from litchi support will make it clearer.

SEP 20, 2018 | 09:50PM UTC
Litchi replied:
Hi,
For best POI framing, make sure you have only one single POI in the mission, in this case it will be optimized.
Kind Regards,
Litchi

 

[parkgt]

I use Litchi on my Spark, Phantom 4 Pro and Mavic pro with similar unexpected results.
As for the ONE point of interest limitation, maybe the response I got from litchi support will make it clearer.

SEP 20, 2018 | 09:50PM UTC
Litchi replied:
Hi,
For best POI framing, make sure you have only one single POI in the mission, in this case it will be optimized.
Kind Regards,
Litchi[/QUOTE


Even with a single POI (with no height to throw things off) the framing can be very sloppy.

 

[Storyline]

You will get better help if you share your mission; hard to guess.

I have seen the option to share the mission on the pc version of MissionHub but can't find it on the android app.
Also, how do I share it without disclosing my email address.? Edit : Here is the mission with the failed photos Mission Hub - Litchi

Anyway, thanks for the replies. It seems that Litchi is not quite as perfect as I thought it was ! I wonder if some of these inconsistencies are caused by one app having to control several different aircraft types. It is a bit frustrating to discover a mission might not work as expected especially if it would be impossible to repeat such as filming live action. I have not tried the program that allows you to fly the mission virtually but a question to those that do use it is whether it will always show up unexpected behaviours ie will a virtual mission always be exactly the same as a real one ? It would be great if it was as you could then iron out any bugs before going up in the air for real.

To go my problems. I had noticed the thing about straight lines so had that option selected. I also only had one poi in the mission. It is disappointing that having only one in a mission is recommended by Litchi. I will try a modified version of my mission and report back here if I discover any more about taking photos.

 

[Storyline]

Just had success taking a mixture of video and stills. I built in 5 & 10 second waits and both worked. The only problem was that although I had a single poi none of the videos or photos were aiming at it. I think this was maybe because I had all of the gimble actions set to interpolate whereas I think maybe the first one needs to be set as the focus point and then everything subsequently can be left as interpolate. Just a guess but it feels logical.

 

[MaviCam]

. I have not tried the program that allows you to fly the mission virtually but a question to those that do use it is whether it will always show up unexpected behaviours ie will a virtual mission always be exactly the same as a real one ? It would be great if it was as you could then iron out any bugs before going up in the air for real.
.

I create my missions in VLM. Then I SAVE the mission and then EXPORT it as CSV. I always run the mission in Google Earth to get a preview of the final product before I actually fly the mission.

On a typical 10 or 12 way-point mission, I might tweak the details and repeat the above process a dozen times before Google Earth displays what I actually want to video. When it shows the paths and views I want I will fly the mission. With the 15 or 20 missions I have flown, I have NEVER had the real mission turn out like the one on Google Earth. The flight path is usually perfect, but the gimbal pitch is rarely even close to what I believed it should have been. I did a mission earlier this week that, as an experiment, had NO points of interest. It turned out to be very close to the Google Earth simulation preview. As it was flown with a Spark, I posted the video on the SparkPilots Forum, here:
I Cover the Waterfront
The last half of that video was from another mission that had the range marker tower as the poi. As you can see, it did a poor job of framing the tower. Hopefully, we can keep comparing notes and figure out how to reliably get the results we are looking for.

 

[Storyline]

.....
.....
Hopefully, we can keep comparing notes and figure out how to reliably get the results we are looking for.

Yes, that sounds a very good idea although I am right down at the bottom of the learning curve. Hopefully other experienced Litchi users will see this thread and join in. From the results I have from my first two flights I can see that this is without doubt the best way to get smooth looking footage. I have my gimble tweaked with custom softened settings and the same for the aircraft controls but even so I have not been able to get any footage as good when flying manually.

Another point that occurs to me is that time spent at the planning stage in the MissionHub also helps with constructing the story you are trying to tell and reduces the amount of time that is spent editing footage afterwards.

 

[Storyline]

Just flown my amended mission and I was wrong to think that the gimble should be set to Focus poi on the first waypoint with subsequent ones being set to Interpolate. The reason I thought I had a problem was that I had the height on the poi set to Above Ground whereas in my first mission that had the poi centred I had this option unchecked in the Poi Settings box. I think all waypoints can be set to interpolate.

However, for some reason when I came to fly the mission for the second time, one of the still photos failed even though the actions at each waypoint were identical !* I can see how Litchi can become frustrating sometimes Instead of changing the waypoint actions I think I will just fix the Poi height problem and then see if I get the same actions executed a second time around. If they change then I really will start to think about pulling some hair out !

If anyone is bored - here is the Mission url. There is no video up yet - I will wait until I have got it all working before uploading it. Mission Hub - Litchi

* Identical to one another - not identical to the first time I flew the mission.

 

[Storyline]

Latest update:

I have now flown the mission several times and on every occasion I have had a problem keeping the POI centred in the frame. I knew it was a problem with altitudes as the Mavic icon would point in the right direction on the pc screen and this was mirrored on the actual flight. What was happening is that the image drifted down in the frame and instead of having our house in the centre the actual view was looking above the roof and into the distance. I had all the Gimble Pitch settings set to Interpolate which had worked perfectly on the very first mission I flew. I tried reducing the height of the POI to a negative figure but this did not seem to do anything. I also changed the global settings in the app with no improvement.

Finally the penny has dropped and I am so confident I have worked out the problem that I am writing this post before trying out the new version ! What I needed to do was to go into each waypoint and then switch from Interpolate to Focus POI. This then changes the Gimble Pitch from zero to the computed figure. Then you can change the selection back to Interpolate and the angle will stick. It will probably tell you to do this in the instructions somewhere but I missed it. The only reason it worked first time was that I did not know what Interpolate was then and had everything set to Focus POI which I subsequently changed.

I am now going out to fly the mission and will edit this post with the result - If it fails I will be coming back via the fridge where the beer is kept !

Edit: Yay - it worked It is very boring but here is the link which I have made public this time - I will upload some pics and video later.

Mission Hub - Litchi

I decided to come back via the fridge anyway

 

[Prmath]

A common noob mistake with Litchi is trying to use waypoint actions with a curved path. When the path is curved the waypoints are never actually reached and the actions don't trigger. Double-check your curved path setting and make sure it's disabled and see if that helps.

Thanks...... That sounds good to me........

 

[Storyline]

A further note about Gimble Pitch and keeping the image centred. Again this may be mentioned in the manual but if not it is worth mentioning as it is easily overlooked.

If you make any changes to an existing waypoint that has been set to Interpolate, you have to select Focus POI after making the change and then switch back to Interpolate. This will ensure the gimble pitch angle is updated to reflect the change. If you don't do this then the angle does not alter with any change. This is the case for both position and altitude changes.

 

[MaviCam]

Litchi is starting to look like it is not yet ready for prime time. If I edit a waypoint, it should be able to deal with the changes I made without requiring me to go through some arcane, undocumented process to get the intended results. If the software allows me to put in multiple points of interest, it should be able to deal with the poi's correctly. I have been using Litchi exclusively for about three months, but, as I learn more about it, I am ready to give up on it. It provides a lot of benefit for my Spark, but I have gone back to using a DJI MOD app for the Mavic. I just wish Litchi would fix their software so it would work as advertised.

 

[Prmath]

Litchi is starting to look like it is not yet ready for prime time. If I edit a waypoint, it should be able to deal with the changes I made without requiring me to go through some arcane, undocumented process to get the intended results. If the software allows me to put in multiple points of interest, it should be able to deal with the poi's correctly. I have been using Litchi exclusively for about three months, but, as I learn more about it, I am ready to give up on it. It provides a lot of benefit for my Spark, but I have gone back to using a DJI MOD app for the Mavic. I just wish Litchi would fix their software so it would work as advertised.

When the rubber meets the road.......
DJI engineers designed DJI stuff

 

[Storyline]

Litchi is starting to look like it is not yet ready for prime time. If I edit a waypoint, it should be able to deal with the changes I made without requiring me to go through some arcane, undocumented process to get the intended results. If the software allows me to put in multiple points of interest, it should be able to deal with the poi's correctly. I have been using Litchi exclusively for about three months, but, as I learn more about it, I am ready to give up on it. It provides a lot of benefit for my Spark, but I have gone back to using a DJI MOD app for the Mavic. I just wish Litchi would fix their software so it would work as advertised.

I can understand your frustration but personally I prefer your approach as per your post earlier when you suggested that users on here should compare notes so we can better understand the app's odd behaviours. There is just one caveat to that and that is I am not yet fully certain that Litchi always repeats missions exactly. If it does then it should be possible to get it to do most things with a bit of trial and error. If however it sometimes behaves randomly then I will lose all trust in it and not waste further time. As I have said above it does a better job as a cameraman than I can do consistently so I really want it to work.

Having said all that there are real grounds to be pi$$ed off with the developers. According to the Google Play Store there have been 100,000+ downloads. I can't remember how much I paid but say an average price of USD20 then that is serious money especially as those figures do not include the ios version. In typical Apple style it does not say how many sales on I-Tunes but it will probably be at least as many as the Android version. So, that could be somewhere north of USD 5 million of gross revenue before cost of sales. So yes, we deserve a professional working app.

There could be another strategy and that is a group of us combining together and having an app written for us in India where they have highly talented app developers. The app could specialise in just cinematography for DJI drones. Then we launch it on Kickstarter