Experimenting with in-brakevan video

What is your latest project?
Post Reply
User avatar
RylstonLight
Trainee Fireman
Trainee Fireman
Posts: 209
Joined: Fri Oct 26, 2012 3:10 pm
Location: Pontefract West Yorkshire

Experimenting with in-brakevan video

Post by RylstonLight » Tue Jul 14, 2020 2:47 pm

This is a re-hash of an old abandoned project. However having been greatly inspired by Greg’s brilliant thread on in-cab video I decided to hunt out the remains of my experiments and have a play. My title is a respectful nod to Greg’s thread whose approach is to use a RC controlled servo to pan the camera.

My project is similar but, I feel sufficiently different to Greg’s, should have its own thread to avoid high-jacking Greg’s.


I’m afraid that it will not be a detailed account as some details are esoteric (unless you enjoy coding Picaxes), and much details is lost in time. I too abandoned the projects due to at the time insurmountable problems, but Greg’s success with newer developments prompted me to look again.

My task was to have a tail-end video that would look as though it was taken from a brake van veranda. Not new I know and I disliked the very fixed view that resulted. My idea was to mount a small camera, like Greg, to a servo, and then use a Picaxe microprocessor to control the servo. I hoped to induce a sense of panning onto passing scenes. Also I hoped to track the front of the train around curves. I always feel that a fixed viewpoint loses realism when it doesn’t track the engine as almost all photographers like the three-quarters side shot of the loco as it rounds the bends.

The only photograph (poor) of the original rig.
EBA02FD1-5943-46E5-8656-3F751D1E413C.jpeg
EBA02FD1-5943-46E5-8656-3F751D1E413C.jpeg (96.61 KiB) Viewed 3364 times
I used a micro processor, rather than a RC unit, as I wanted to automate the panning so I could concentrate on driving the loco.

The Picaxe was programmed to randomly pan back and to in a limited arc. There was a Reed switch slung underneath the chassis so that strategically placed magnets on the sleepers (moveable, fixed with blutack) could trigger a hard pan, followed by a return to the limited arc pattern. Strategically placed magnets could cause the camera to pan to interesting vistas. This all worked well, although I never solved the tracking of the front of the train.

So why did I abandon the project? The main reason was that the small cameras then were poor resolution. Jokingly so compared with today’s standards. The second problem was that the servo chatter (even when panning slowly) was so loud that the video footage was useless. I contemplated separate sound recording but it was all getting too complicated for a side project.

So, having been inspired by Greg, back to today’s experiments.

I realised that the Picaxe still worked, although the random panning is a bit busy and needs taming down. I’m glad I included a programming port on the veroboard now. The camera also worked but was not fit for purpose. I acquired a micro 4/3 digital camera with good video capabilities since the original project. I realised that this would just sit within the loading gauge. So if I can fix a camera mount to the wagon then the video quality is sorted. I use QR plates on my camera to attach it to a tripod, so it made sense to use one here.

The other problem is the servo sound. I realised that I need to acoustic de-couple the servo from the camera. The only way I could think of this is to have the camera on a separate chassis to the servo. As long as the wagons were rigidly coupled then the servo and the camera could then be linked mechanically.

I also realised that if I placed the servo at one end of the first wagon and the camera mount near the rear of the second wagon then the connecting rod could be used to mechanically pan and track the front of the train. (If it is away from the midline of the train). The static resistance of the servo rotation was greater than the resistance of the camera rotation. When the servo is off it will act as a fixed-point for one end of the connecting rod. As the train goes around the curve the connecting rod is relatively longer (or shorter) and pushes the crank on the camera mount turning the camera. Hard to express clearly but simple in operation.

Any servo movement just superimposes on the mechanical movement on of the servo rod.

Picture from above hopefully helps.
0DC4265B-4C54-441F-99EF-A86F66894157.jpeg
0DC4265B-4C54-441F-99EF-A86F66894157.jpeg (145.59 KiB) Viewed 3364 times
I had a suitable reclaimed ball race to form the bases of the camera pivot.
7B3CDAD6-3733-42D5-8849-DA53B3B7A5D4.jpeg
7B3CDAD6-3733-42D5-8849-DA53B3B7A5D4.jpeg (144.49 KiB) Viewed 3364 times
I hope the photographs help.

Other brief details the outer ring of the ball race is epoxied to the chassis. An M5 bolt is run through the inner ring of the ball race and screws onto the QR plate. The two sandwich a home made crank cut from scrap brass-etch fret.

The connecting rod connects between the wagons with the brass insides of an old “chocolate-block” electrical connector.
35361E99-9274-4E7C-9923-447B62DC53FA.jpeg
35361E99-9274-4E7C-9923-447B62DC53FA.jpeg (24.34 KiB) Viewed 3364 times
Detail of suspended Reed switch
68B40645-7DC5-49C1-8B04-CD80600FFF66.jpeg
68B40645-7DC5-49C1-8B04-CD80600FFF66.jpeg (28.79 KiB) Viewed 3364 times
It It has a frame that sits in front of the camera lens. It gives the illusion of the veranda supports as the camera pans.
E5EB560B-FFC3-4486-A9FB-CF00CD3DF0AD.jpeg
E5EB560B-FFC3-4486-A9FB-CF00CD3DF0AD.jpeg (91.17 KiB) Viewed 3364 times
It works without binding on PECO minimum radius set track arranged as an S-curve, so should work outside.

I know you really want video but I haven’t got a YouTube account :roll:

Andy.
Andy S. at the Rylston Light Railway

User avatar
tom_tom_go
Driver
Driver
Posts: 4824
Joined: Wed Feb 23, 2011 3:08 am
Location: Kent, UK
Contact:

Re: Experimenting with in-brakevan video

Post by tom_tom_go » Tue Jul 14, 2020 5:31 pm

You can share a video via any file sharing service.

User avatar
gregh
Trainee Driver
Trainee Driver
Posts: 566
Joined: Sun Apr 01, 2018 5:44 am
Location: Sydney, Australia
Contact:

Re: Experimenting with in-brakevan video

Post by gregh » Wed Jul 15, 2020 7:19 am

Lots of very clever ideas there. Such a simple thing putting a rod between two wagons. I like the 'dummy' end platform to provide a frame for the camera.

I have found that my phone fits inside the cab between the loco and tender. ( I have a simple phone which is smaller than smart phones).
So after a few experiments I'll put some more info up on my topic.
Greg from downunder.
The Sandstone & Termite's website: https://members.optusnet.com.au/satr/satr.htm

User avatar
RylstonLight
Trainee Fireman
Trainee Fireman
Posts: 209
Joined: Fri Oct 26, 2012 3:10 pm
Location: Pontefract West Yorkshire

Re: Experimenting with in-brakevan video

Post by RylstonLight » Wed Jul 15, 2020 9:08 am

gregh wrote: Wed Jul 15, 2020 7:19 am So after a few experiments I'll put some more info up on my topic.
Look forward to seeing them.

Andy
Andy S. at the Rylston Light Railway

Post Reply

Who is online

Users browsing this forum: No registered users and 4 guests