InfinityMR Universal high performance gimbal for GH4/A7S size cameras
- Thread starter jakuban
- Start date
SamaraMedia
Active Member
Looks good Tom, wind didn't seem to bother your gimbal tracking at all. Sent you a PM early this week, wondered if you got it?
Hmmm...I don't have any PMs in my box from you. Try e-mail - tombrown1@yahoo.com
There is post stab in the above video - but like the corn maze vid we didn't need too much. Gimbal and CP give great stabilization. We also did some horizon adjusting in about half the shots.
There is post stab in the above video - but like the corn maze vid we didn't need too much. Gimbal and CP give great stabilization. We also did some horizon adjusting in about half the shots.
econfly
Member
Here is a couple of minutes from my first test flight today. I have a little bit of wobble to tune out. Wind was about 5 mph with gusts to 10-15 (based on my weather station).
Nothing fascinating here, just a test. The camera is the Sony a7R ii with a Zeiss Loxia 35mm lens. No post stabilization added here at all -- just footage straight from the camera with a little color adjustment.
The gimbal is very nicely constructed. I have it mounted on a DJI S1000+ which has dampening on the 12mm rails. I may need to stiffen that up given the dampening on the gimbal itself.
Nothing fascinating here, just a test. The camera is the Sony a7R ii with a Zeiss Loxia 35mm lens. No post stabilization added here at all -- just footage straight from the camera with a little color adjustment.
The gimbal is very nicely constructed. I have it mounted on a DJI S1000+ which has dampening on the 12mm rails. I may need to stiffen that up given the dampening on the gimbal itself.
SamaraMedia
Active Member
Looking good econfly for first test flight. Little horizon drift when hawing the the left it seems fine going other direction. This is with AM and encoders correct?
Thanks for sharing, look forward to more as you tune it.
Thanks for sharing, look forward to more as you tune it.
jakuban
Member
That looks very good Rob! What you can do right now is to increase a bit values of P and D parameter. You can also try to autotune each axis separately (in following order: pitch, roll, yaw).
One important thing: this is how the footage looks when you receive the gimbal and just install your camera, without any changes (InfinityMR right now is delivered ready to fly). But of course after some small tweaks you can get it working even better. Most frequently final effect depends of multi-rotor IMHO.
One important thing: this is how the footage looks when you receive the gimbal and just install your camera, without any changes (InfinityMR right now is delivered ready to fly). But of course after some small tweaks you can get it working even better. Most frequently final effect depends of multi-rotor IMHO.
violetwolf
Member
I'd love to see some high speed sideways flying, car chase style. I've found that to be the best test of horizon holding.
But I also think perfect horizon is elusive due to physics. The imu simply cannot differentiate between G forces and gravity. And encoders only know where the frame is, not the level.
It literally only takes seconds to rotate a clip 1-2 degrees in Premiere.
I never sweat horizon unless its off more than 2 degrees (which looks like a lot on screen.)
But I also think perfect horizon is elusive due to physics. The imu simply cannot differentiate between G forces and gravity. And encoders only know where the frame is, not the level.
It literally only takes seconds to rotate a clip 1-2 degrees in Premiere.
I never sweat horizon unless its off more than 2 degrees (which looks like a lot on screen.)
violetwolf
Member
PS: love the look of that camera!
I might be able to arrange that
jakuban
Member
This week I will send info about InfinityMR-S testing to those two person I will choose. I get lots of emails and I think I'll just make a draw so everyone has the same chances Thanks again to all of you!
I am working now on multirotor frame with lots of nice features Light, right, vibrations free, for professional use together with my gimbals. More info in about 2months.
Thanks again to all of you!I am working now on multirotor frame with lots of nice features
Light, right, vibrations free, for professional use together with my gimbals. More info in about 2months.Hi Jakub,
I really want to integrate yours new small gimbal in our new frame we are developing right now.
Please can you tell how much weight it's right now the InfinityMR-S the dampening system?
Please it's really important to us because we are developing the frame and we are making all the calculations and decision for motors, ecc...
Thanks
I really want to integrate yours new small gimbal in our new frame we are developing right now.
Please can you tell how much weight it's right now the InfinityMR-S the dampening system?
Please it's really important to us because we are developing the frame and we are making all the calculations and decision for motors, ecc...
Thanks
jakuban
Member
InfinityMR is going to be maximum 800g with everything including damping and multirotor mounting system. Prototype version that I use now is 720g, but there are few things that will change and might affect final weight result.
What you can see on this photo is not everything (for example multirotor mouting system missing, and some others small details...), but mostly this is final shape.
What you can see on this photo is not everything (for example multirotor mouting system missing, and some others small details...), but mostly this is final shape.
Mactadpole
Member
Will the InfinityMR-S have the orbit dampening system? It looks like the standard system in the pic.
Looks great and I like the weight!
Looks great and I like the weight!
SamaraMedia
Active Member
Looking real good Jakub!
Yesterday I've spoken with Jakub and he asked me to share my thoughts and raw movies taken by using InfinityMR (version without orbit dampening system, AM + encoders)
Test platform is dji s900, InfinityMR (AM + encoders), sony a6000 with SEL20F28 + VCL-ECU1
First movie was recorded using PID's provided in the documentation, although I've lowered power values bcs there were to high. I think that they were provided for GH4. On the ground PID's were working correctly.
As you can notice there is a lot of shake in first aerial video, after stabilization picture was extended by 3%, so it is a lot from my perspective.
So there was need to improve it!
After consultation with Jabkub about autotune and how it will be working with encoders I decided to try it.
I want to add that movie editor which asked for above video didn't complained at it Even 3% was acceptable for him
Anyway..
Second video is a test shot with pid's set by autotune.
Motor power the same as in first video, only PID's changed.
Autotune set them to higher values, but as you can notice shaking is reduced.
After stabilization video was extended by 0,7%, which is for me good value which I can accept.
As all of you know s900 is problematic set if we are speaking about vibration.
I will not dig more in this subject
Right now I'm counting to be selected to check smaller version of infinity on gaui 950Q coaxial frame, with t-motors u5 and wooden props.
I've been using t-motors with wood props on other frame also in coaxial configuration, and with simple AM gimbal without encoders I got almost stable vid (0,3% after stabilization). So I'm counting that using gimbal with encoders on gaui 950Q frame I can achieve a lot better video
Maybe without any shakes It would be something right!
If I would be lucky to have infinity on test, I will make a review of mentioned set if there will be demand for it.
Test platform is dji s900, InfinityMR (AM + encoders), sony a6000 with SEL20F28 + VCL-ECU1
First movie was recorded using PID's provided in the documentation, although I've lowered power values bcs there were to high. I think that they were provided for GH4. On the ground PID's were working correctly.
As you can notice there is a lot of shake in first aerial video, after stabilization picture was extended by 3%, so it is a lot from my perspective.
After consultation with Jabkub about autotune and how it will be working with encoders I decided to try it.
I want to add that movie editor which asked for above video didn't complained at it
Even 3% was acceptable for him Anyway..
Second video is a test shot with pid's set by autotune.
Motor power the same as in first video, only PID's changed.
Autotune set them to higher values, but as you can notice shaking is reduced.
After stabilization video was extended by 0,7%, which is for me good value which I can accept.
As all of you know s900 is problematic set if we are speaking about vibration.
I will not dig more in this subject
Right now I'm counting to be selected to check smaller version of infinity on gaui 950Q coaxial frame, with t-motors u5 and wooden props.
I've been using t-motors with wood props on other frame also in coaxial configuration, and with simple AM gimbal without encoders I got almost stable vid (0,3% after stabilization). So I'm counting that using gimbal with encoders on gaui 950Q frame I can achieve a lot better video
Maybe without any shakes
It would be something right!If I would be lucky to have infinity on test, I will make a review of mentioned set if there will be demand for it.
Last edited by a moderator:
Seems coming very good!InfinityMR is going to be maximum 800g with everything including damping and multirotor mounting system. Prototype version that I use now is 720g, but there are few things that will change and might affect final weight result.
What you can see on this photo is not everything (for example multirotor mouting system missing, and some others small details...), but mostly this is final shape.
But don't you think it will be overkill 800g for a camera + lens that will weight 355 ?
Also the new Blackmagic Micro Cinema Camera will be very small, how InfinityMR-S could be perfect for it?
Do you see room for make it lighter?