A bit more than pan tilt head for BMMSC

Thanks Casper.

The gimbal is currently working well for this application and we have accurate encoders for reading the Zoom and Focus/Iris position as needed.

IMG_2467.jpg (92.83 KiB) Viewed 976 times

However we are still working with lens control and could use some advice. Any suggestions for lens control system that could be controlled externally would be most welcome. We would rather integrate them to our system with serial etc. low level interface and of-course would need documentation for the interface.

Currently we are using just a cheap servo. It works but is noisy, and this one is not quite powerful enough, so it does not provide smooth low speed zooming. We have also some better servos that overcome most of the problems, but still somewhat noisy.

IMG_2468.jpg (84.24 KiB) Viewed 976 times

The reason we use separate encoders is that after calibration they provide very accurate readings and fast update rates. So even now that the control is not that good, the virtual environment follows very well the Zoom movements. Also some lenses have inbuilt Zoom motors.

There is still many things to overcome, but already at this time I am sure the end result will be very good, not existing at the price point this system will be.

Ps. As obvious, the above pictures are from 3D printed prototypes.

Take a look into the motors of Tilta.. The Nucleus motors are quite "cheap" ($300) Are able to control on a bus structure and also wireless..

https://tilta.com/shop/nucleus-m-brushless-wireless-follow-focus-motor/


I do have a redrock follow focus motor with control.. Was about to rebuild the control interface of that one.. And buy some more motors.. But with $695 they are quite expensive..
https://shop.redrockmicro.com/product/torque-motor-for-movi-pro/

Thanks Daniel, They both look good and Tilta is also affordable. I contacted both to request if they would be interested to provide them for this kind of solution, and provide the interface definition.

There will be a control SW that allows the user to control Pitch, Roll, Yaw, Zoom, Aperture, Focus and the can be also preprogrammed to repeat accurately the movements and lens parameters.

We have been lately working with lens control. It is really not something we would like to do at this time, but have not found suitable solution, that would provide smooth control at slow speeds, accurate control and real time feedback of the position. Tomorrow visiting IBC so mybe this widens the perspective and there is a solution.

This is what we have for now, it does the job, only problem realy is the noise, ok for custom builds. (these are just demonstrating the servo, it will be computer controlled.)



Slow speed demo

Looking good Kim.
Cheers

So is there a timeline on when the Ethernet control will be available? Also can this be mounted on a pole hanging upside-down from a truss?

We received today the B4Macig adapter to fit B4 lenses to MFT. As I understand this is not great at wide open, but should be ok stopped down to F5.6.

B4Macig.jpg (243.02 KiB) Viewed 791 times

Waiting the lenses to arrive to see what kind of image quality these inexpensive B4 lenses can provide and how good lens control they have, and further more how to interface with them.

The Fujinon A19x8.7BMD-D24 1:1.8/8.7-165mm arrived today.

IMG_2768.jpg (66.08 KiB) Viewed 731 times

As expected it is not really usable at F1.8, not at the wide end anyway, but cleans up nicely at f5.6 as recomended by the B4 Magic.



It has 3 lens servos that are the main reason we got this lens, for testing and development on that area.

IMG_2764.jpg (55.64 KiB) Viewed 731 times

The mechanical parts look good, but the all analog electronics will be fully replaced. Planning to do first tests still during this week.

The lens got all new electronics. Now it is microprosessor controlled, connection via serial port or USB. It will use the LeViteZer protocol https://github.com/LeViteZer-Oy/LeViteZer-Protocol but we could make it work maybe also with the B4 lens serial protocol that this lens did not originally have.

IMG_2786.jpg (546.98 KiB) Viewed 621 times

The lens control motors are good, but the potentiometers that provide feedback of the lens position are badly worn out. This can be clearly seen on this video, when it sudenly makes a jump at about 40 mm. I guess it has been mostly at that setting and does not make good contact any more.



Need to find similar potentiometers and replace them. After it should be possible to get this kind of slow Zoom even smoother. Faster Zooms are ok already (except when passing 40 mm).

This is a prototype for testing, when we have it fully functioning we will post more videos the hole system installed on the gimbal.

Ps. a bit more technical view of the potentiometers. The blue line is focus, red is zoom. Both have bad jumbs and noise. This is unfiltered output just to troubleshoot.

I guess this is a problem for many such old lenses.

Here some more details about this modification viewtopic.php?f=4&t=70896&start=50#p443100

We will have soon the lens control ready and demonstrate the lens, camera and gimbal control working in perfect harmony, but meanwhile the LeViteZer 360 Pro gimbal on its originally intented use.

Heli-Ski.jpg (108.38 KiB) Viewed 482 times

‘Behind the Scenes: Creating Heli-Skiing in Virtual Reality’ https://stories.cmhheli.com/heli-ski/he ... l-reality/

VR landing page + tour dates http://www.cmhheli.com/vr

Thanks for Sherpas Cinema for putting the gimbal to great use and providing this.

We are still thinking a bit the functionality the lens control should have and asking about opinions.

This is how it works now, zooming from 9 mm to 160 mm at speed 500 raw units per second. The full scale is about 16 000 raw units. This is still the development user interface.



We will keep the engineering approach available, using the raw units etc, but askin how would you like to control the Zoom speed

- the time it goes from position A to B
- the time it would take for the full zoom
- zoom speed in mm per second
- something else?

There is also control how the zooming starts and stops, how fast it accelerates to given speed. currently we use S curve for this.

It will be possible also to control directly the Zoom in mm and control it with direct speed control.

There is so many options, but must start somewhere and keep the user interface understandable. Asking for your help defining the first version of it. what would be important for you?

We could have similar controls for Iris and focus, but thinking keeping them simple to start with, i.e. the control will be

- Speed controlled with the time it takes for the full range
- the wanted f-stop of focus in m (absolute control)
- option to control directly with speed control (relative control)

for all parameters we will get feedback, currently 120 Hz.

Just tested the Geo Point functionality, and thought to share this, though we are still working to get it totally smooth.



We have full camera control on this system, but had our hands full with the testing and forgot to adjust the colours. It has also software controlled Zoom, Focus and Iris. The Zoom was tested couple of times on the video.

The jitters you can see are due we used raw GPS data and some parameters we still need to get right.