Photohigher skyline rsgs

Macsgrafs

Active Member
Ross that looks good!

Much more stable than mine is right now! Can't see anything wrong with that :tennis:

watch the horizon as I turn the octo, it slants over badly, check just after the first flyover of the stones on the cliff.

How about using this forum for helping each other. This forum needs to stay on topic a bit better. I don;t have the time to read the endless rants. Sure, make your complaint and end it there. I have had better luck with RSGS than picloc so far. But, I sure would like some advice getting this thing set up better. It is close to being what I need.

Thats what I thought we were all here for..to help each other. But it seems certain individuals would rather have a go at people who have a legitimate complaint! The bottom line is, I bought a peace of equipment in good faith, which has turned out to be not as advertised..in this country (UK) that is false advertisement, but if you are happy to be conned & lose money, please feel free to take a back seat & let people like me fight on to get it sorted.

Thanks you.

Ross
 

vip

Member
How about using this forum for helping each other. This forum needs to stay on topic a bit better. I don;t have the time to read the endless rants. Sure, make your complaint and end it there. I have had better luck with RSGS than picloc so far. But, I sure would like some advice getting this thing set up better. It is close to being what I need.

I agree, we really do not care about personal information.. that is better suited to PM's..
mine is close also but everything need to be stabalizes still with after effects.

My gains are 275 on roll and 2oo and velocity, what is others doing? I still get twitches on both.
I have tried default settings but reaction time is to slow.
 

MikeH

Member
Guys, i am meeting with Kim and Brendan in Wellington on friday in order to rewrite the instructions over the next couple of weeks.
The tuning will be the area i focus on.
You have seen my vid which proves that the skyline does work.

If you have any common recurring issues or thoughts on the how you would like to see the manual please either pm me or email at skylinemanual@gmail.com and i will try and include it in a troubleshooting guide (time dependant)
 


vinnygogo

Member
Can some one please explain clearly to me what the different gains are and do? I have my skylines working pretty well, but would like to now exactly what for example velocity gain does and how you see when to high/low.

Thanks in advance!
 

Efliernz

Pete
Can some one please explain clearly to me what the different gains are and do? I have my skylines working pretty well, but would like to now exactly what for example velocity gain does and how you see when to high/low.

Thanks in advance!


Guessing the tuning due to non-existant guidelines / instructions is what many of us want to see sorted... and in comes MikeH :D
There are obviously lots that can be fiddled with on the RSGS but PH ability to sell and not support has seriously damaged their reputation. Not all of us Kiwis work this way! I look forward to the new instructions...

I am getting reasonable video but the 10-deg roll while turning/yawing is a problem. I have got my RSGS aligned perfectly inline with the tilt axis - easy to check in my case (diy gimbal) by tilting the mount straight down from level. If it rolls at all while tilting, it is not aligned fully.

@MikeH - I will sent you a link to a video I have of the BSOD issue (V1.7). If I am on the monitor tab, the pc crashes and corrupts the program - where it needs to be manually removed and reinstalled.

Pete
 

Sl4uGh7eR

Member
Here is where we first tested RSGS with 1.2.2 firmware on av200
Almost no post stabilization, camera used was sony nex vg-20 with 18-200 lens with optical stabilization.

Here is another video with 1.2.2 firmware, but camera used was sony hdr-cx730 on av130 and no post stabilization at all

And another one with some chrismtas shoots, also with av130 cx730 combo, still on 1.2.2

We have like 3 materials we working on now, made with 1.7b1. We hope to finish materials for clients till end of week, and i can post what we got.
If i get time in work tomorrow i will try show how i autotune our av200 or av130 with 1.7b1 on board.

Robert aka BoB
 
Last edited by a moderator:

toscatara

Member
Here is where we first tested RSGS with 1.2.2 firmware on av200
Almost no post stabilization, camera used was sony nex vg-20 with 18-200 lens with optical stabilization.

Here is another video with 1.2.2 firmware, but camera used was sony hdr-cx730 on av130 and no post stabilization at all

And another one with some chrismtas shoots, also with av130 cx730 combo, still on 1.2.2

We have like 3 materials we working on now, made with 1.7b1. We hope to finish materials for clients till end of week, and i can post what we got.
If i get time in work tomorrow i will try show how i autotune our av200 or av130 with 1.7b1 on board.

Robert aka BoB
thank
 
Last edited by a moderator:

MikeH

Member
Can some one please explain clearly to me what the different gains are and do? I have my skylines working pretty well, but would like to now exactly what for example velocity gain does and how you see when to high/low.

Thanks in advance!

Hey vinnygogo

This is definitely one section that will be covered. I want to know the same thing :)
 

rcmike

Member
Can some one please explain clearly to me what the different gains are and do? I have my skylines working pretty well, but would like to now exactly what for example velocity gain does and how you see when to high/low.

Thanks in advance!
Yep, same here. I understand the basic gains but not exactly sure where to set the others.
 

Tuukkuli

Member
Hi BoB

That´s some nice footage and atmosphere there. Congrats! What camera settings did you use in the Sony CX-730?

Tuukka
 

Tuukkuli

Member
Update on my Skyline issue: The one I got seemed to be faulty, so Photohigher agreed me to ship it back to them and they shipped me a new one. It arrived from New Zealand to Finland in four days so that was pretty fast indeed. I will pick it up tomorrow.

I also had some problems with my AV130 - it was one with the missing countersink screws - and they dealt with it also and replied to my mails instantly.

After a lot of frustration the Photohigher support has turned out to be quite a good one after all. At least in my case. If the new Skyline works well, I´m starting to regain trust in their products. Support wise I am already quite happy. But I also would like to see a well written, thorough and comprehensive user manual for the calibration, use and all the parameters of the Skyline. Hope they will come up with it soon.

Will keep you updated.

Tuukka
Finland
 

Helifish

Member
Active electronic stabilisationexplained. ( Roughly)
We all want cineflex results and in reality with the standard of gear we are using it just isn’t going to happen.....but it’s getting bloody close.
Electronic gyro stabilisation uses some tricky little components. Usually it is a combination of a Mems type gyro chip along with an Accelerometer chip and a bunch of other little bits.
Why the skyline rolls or tilts during turns.
The heart of the skyline and most electronic stabilisation systems is a Mems gyro. By its self it measures movement about 1 or more planes. Due to various quality and cost factors they develop a drift for various reasons and they need a reference as to a specific place in space to know where they should be. Enter the accelerometer. It pretty much knows where strait down is due to gravity so tells the Mems gyro this so if there is any drift it puts it back or corrects it to its perceived level position.
However....when still, in a hover. Or in strait and level and UN accelerating flight the accelerometer works just like a plum bob hanging on a string under your aircraft. Level is 90 degrees to the string. So fly along with no turning forces and we get a nice level result.
When we turn or accelerate or slow down just picture where the plum bob string would be.
In a turn it’s going to hang out the side due to centrifugal forces. When you accelerate forward it will hang back behind and you got it ....when you slow down the string would be out the front.
So when you do your turn the accelerometer tells the Mems gyro strait down is out the side and the Mems correction is to keep your level 90 degrees to the string line ...so the horizon rolls into the direction of the turn.
Same as Tilt.....accelerate forward and the tilt goes down.......if you pan 90 degrees and accelerate forward .....It rolls into the direction of acceleration.
So to fix this there is a heap of maths and corrections going on which limit the amount of input or effect the accelerometer has.....this is one of the reasons why you are getting different results with different firmware....basically different settings in the firmware.
Most of the videos we see that are nice and level you will notice the flight is pretty much strait, steady and no changes in velocity or acceleration.
Try flying around in a circle while slowing down and the horizon is going to drift....eventually.
I will almost guarantee you the nice flash level videos will have one person flying and another person driving the camera and the camera person will be keeping the horizon level with gimble inputs.....I know I do and it’s how well I can smoothly control the roll and tit that is the end result.
While all this is still going on the picture is still very stable....all the wobbles and twitches from the aircraft are being taken out.....just not the changes in acceleration.
In High end gyro gear like the cineflex there is often 2 operators on the camera in the aircraft. Plus the gyros in those cameras are military spec Technologies worth squillions of coin. Some also utilise gps position from satellites to keep it all level so as not to rely on the accelerometer input. And there is often 5 axis of stabilisation.

The gains in most gyro software like the RSGS simply allow you to adjust how sensitive the servos are in responding to a movement and how quickly they they try to restore it.

With the constant improvements in technology and such it wont be long before we can get the cinema results we would all love to have.....just if you want it now then you need a very large wallet.
 
Last edited by a moderator:

Droider

Drone Enthusiast
This thread is now MASSIVE. Amongst all the critics on here I do have to say I was very disappointed PH released it before the BETA testers could give them their full opinion. That is in the past and what is done is done. I aint going to post my footage wit it as it may jsut pee loads of people off!

I dont have any problems with it. Hardware version 1.2 firmware 1.2.2

Dave
 

nicwilke

Active Member
Dave, I dont have huge issues either. I get some anomalies from time to time, but 1.7b seems very good. I do agree with Hellfish that we all expect Cineflex quality for hobby price, and although its expensive to us its really a small investment considering we are flying camera equipment. I think as the Sony BOSS technology matures, and others develop similar tech, we will get stabilized cameras on stabilized gimbals.

The brushless gimbal market is about to explode, so we will see how well these cheaper open source driven cameras go. I'm waiting patiently for the Halo1000, and although there's the Alexmos based gopro gimbal around the corner for sweet 'f' all, I'm going to bet the Halo will be a leap in front of it.
 

Macsgrafs

Active Member
Active electronic stabilisationexplained. ( Roughly)
We all want cineflex results and in reality with the standard of gear we are using it just isn’t going to happen......

How about the results that PH published, that would be a great start for me! Your explantion of how it works is impressive.

This thread is now MASSIVE. Amongst all the critics on here I do have to say I was very disappointed PH released it before the BETA testers could give them their full opinion. That is in the past and what is done is done. I aint going to post my footage wit it as it may jsut pee loads of people off!

I dont have any problems with it. Hardware version 1.2 firmware 1.2.2

Dave

Glad you are finding some success Dave.

The brushless gimbal market is about to explode, so we will see how well these cheaper open source driven cameras go. I'm waiting patiently for the Halo1000, and although there's the Alexmos based gopro gimbal around the corner for sweet 'f' all, I'm going to bet the Halo will be a leap in front of it.

Looking forward to that happening, super stable at real prices..total winner!
 

MikeH

Member
Active electronic stabilisationexplained. ( Roughly)
We all want cineflex results and in reality with the standard of gear we are using it just isn’t going to happen.....but it’s getting bloody close.
Electronic gyro stabilisation uses some tricky little components. Usually it is a combination of a Mems type gyro chip along with an Accelerometer chip and a bunch of other little bits.
Why the skyline rolls or tilts during turns.
The heart of the skyline and most electronic stabilisation systems is a Mems gyro. By its self it measures movement about 1 or more planes. Due to various quality and cost factors they develop a drift for various reasons and they need a reference as to a specific place in space to know where they should be. Enter the accelerometer. It pretty much knows where strait down is due to gravity so tells the Mems gyro this so if there is any drift it puts it back or corrects it to its perceived level position.
However....when still, in a hover. Or in strait and level and UN accelerating flight the accelerometer works just like a plum bob hanging on a string under your aircraft. Level is 90 degrees to the string. So fly along with no turning forces and we get a nice level result.
When we turn or accelerate or slow down just picture where the plum bob string would be.
In a turn it’s going to hang out the side due to centrifugal forces. When you accelerate forward it will hang back behind and you got it ....when you slow down the string would be out the front.
So when you do your turn the accelerometer tells the Mems gyro strait down is out the side and the Mems correction is to keep your level 90 degrees to the string line ...so the horizon rolls into the direction of the turn.
Same as Tilt.....accelerate forward and the tilt goes down.......if you pan 90 degrees and accelerate forward .....It rolls into the direction of acceleration.
So to fix this there is a heap of maths and corrections going on which limit the amount of input or effect the accelerometer has.....this is one of the reasons why you are getting different results with different firmware....basically different settings in the firmware.
Most of the videos we see that are nice and level you will notice the flight is pretty much strait, steady and no changes in velocity or acceleration.
Try flying around in a circle while slowing down and the horizon is going to drift....eventually.
I will almost guarantee you the nice flash level videos will have one person flying and another person driving the camera and the camera person will be keeping the horizon level with gimble inputs.....I know I do and it’s how well I can smoothly control the roll and tit that is the end result.
While all this is still going on the picture is still very stable....all the wobbles and twitches from the aircraft are being taken out.....just not the changes in acceleration.
In High end gyro gear like the cineflex there is often 2 operators on the camera in the aircraft. Plus the gyros in those cameras are military spec Technologies worth squillions of coin. Some also utilise gps position from satellites to keep it all level so as not to rely on the accelerometer input. And there is often 5 axis of stabilisation.

The gains in most gyro software like the RSGS simply allow you to adjust how sensitive the servos are in responding to a movement and how quickly they they try to restore it.

With the constant improvements in technology and such it wont be long before we can get the cinema results we would all love to have.....just if you want it now then you need a very large wallet.

That is the best explanation I have heard, simple and concise so everyone can understand it. Very well done, I am truly impressed!
Would you mind if i borrowed that for the manual :) It would go along way to helping people understand whats occuring, thus aiding there tuning

This is exactly how we fly, the smoother my flightpath the easier job my brother has on the camera,it truly is a team effort - aircraft flying perfectly, pilot flying smoothly, camera operator using smooth motions and compensating if any drift does occur and the skyline taking out 90% of the unwanted movement. Take out any one of those ingredients and the shoot will fail.
 

nicwilke

Active Member
That is the best explanation I have heard, simple and concise so everyone can understand it. Very well done, I am truly impressed!
Would you mind if i borrowed that for the manual :) It would go along way to helping people understand whats occuring, thus aiding there tuning

This is exactly how we fly, the smoother my flightpath the easier job my brother has on the camera,it truly is a team effort - aircraft flying perfectly, pilot flying smoothly, camera operator using smooth motions and compensating if any drift does occur and the skyline taking out 90% of the unwanted movement. Take out any one of those ingredients and the shoot will fail.


View attachment 9302 Hence my title here. HEHEH. Yes, Theres drift, but slowly pan and you'll avoid most of it. I'd like to see skyline2 with integrated GPS and compass for horizon hold.
 

Attachments

  • Screen shot 2013-01-30 at 4.04.21 PM.jpg
    Screen shot 2013-01-30 at 4.04.21 PM.jpg
    21 KB · Views: 340

mounirz

Member
hi folks,
I have been testing the new software 1.7beta1 and i find it better than 1.2.2 i am satisfied with the results. i did some flights on my skyjib 8, av200 360.
i have created a simple tool i call it my tuning chair :nevreness: that aids me tune my gimbal, it cuts my tuning time a lot and is much easier and simple. you can see the picture below, the black aluminum arms are 90deg to each other at the top, in the middle i have fabricated a mounting system for the av200 to mount on and off easily for testing,,,,,,,and if you look at the chair arms you can see the swivel point of the device. when i move the aluminum arms forward and back i can easily monitor the reaction of the tilt axis and if i rotate the gimbal 90 deg i can monitor the roll axis. and you will be surprised to how much this is precise since the tilt axis will not shift forward or back its just there and you will tune as you go and experiment etc..... i have been testing and flying for quite some time now and i can say i am not 100% happy. though i am very very very close to having the perfect setup who everyone of us dream of.
my problem is simple...
when the gimbal is moving it is holding both axis very well but when there are very smooth inputs for ex lets say i move 5 deg forward and back, very gentle slow movements, you can clearly see the servo almost takes a bit of time before it starts to correct, even if my gains up to almost 300. i tried everything to speed up the initial reaction of the servo and nothing cuts it. this translates in an up down movement in my video when hovering. i am talking here only about small fractions depending on how the OCTO is correcting.
anyone noticed this problem ? or at least has it and solved it ? for me its the same issue on both axis on roll its more :upset:
i would be very happy if i can fix this small issue !!


View attachment 9306
 

Attachments

  • 2013-01-30 13.39.57.jpg
    2013-01-30 13.39.57.jpg
    131.4 KB · Views: 337

iceman

Member
hi folks,
I have been testing the new software 1.7beta1 and i find it better than 1.2.2 i am satisfied with the results. i did some flights on my skyjib 8, av200 360.
i have created a simple tool i call it my tuning chair :nevreness: that aids me tune my gimbal, it cuts my tuning time a lot and is much easier and simple. you can see the picture below, the black aluminum arms are 90deg to each other at the top, in the middle i have fabricated a mounting system for the av200 to mount on and off easily for testing,,,,,,,and if you look at the chair arms you can see the swivel point of the device. when i move the aluminum arms forward and back i can easily monitor the reaction of the tilt axis and if i rotate the gimbal 90 deg i can monitor the roll axis. and you will be surprised to how much this is precise since the tilt axis will not shift forward or back its just there and you will tune as you go and experiment etc..... i have been testing and flying for quite some time now and i can say i am not 100% happy. though i am very very very close to having the perfect setup who everyone of us dream of.
my problem is simple...
when the gimbal is moving it is holding both axis very well but when there are very smooth inputs for ex lets say i move 5 deg forward and back, very gentle slow movements, you can clearly see the servo almost takes a bit of time before it starts to correct, even if my gains up to almost 300. i tried everything to speed up the initial reaction of the servo and nothing cuts it. this translates in an up down movement in my video when hovering. i am talking here only about small fractions depending on how the OCTO is correcting.
anyone noticed this problem ? or at least has it and solved it ? for me its the same issue on both axis on roll its more :upset:
i would be very happy if i can fix this small issue !!


View attachment 12009


I had a similar problem on my tilt axis and after careful re-balancing of the camera and making sure the tilt belt was under good tension it's now smooth. The problem I'm having is with the pan axis, I just can't seem to stop it shaking when making small left/right yaw (on the bench, not flow with pan 360 yet). If I make large sweeping rotations left to right it looks OK but very twitchy/shaking when either side of straight. I have tried all the pan gain settings from default, up and down but it’s like the servo isn’t responding strong enough, anybody got any ideas. AV200, stock servos, 1.7 beta
 
Last edited by a moderator:

Top