Ben, when building stuff using belts, it is always a good idea to add a tensioning mechanism - e.g. a screw or spring pushing the idler pulley away. Or you can add a separate tensioning pulley. That would save you the trouble with the loose belts and trying to fit an exact length. For the Y axis, it should be enough to have the stepper hold its position. It seemed powerful enough for that, if not, give it a smaller pulley to increase the gearing ratio. However, if you do this, you will probably want a heatsink on the stepper driver - a stalled stepper can pull significant current.
Watching this i came to the conclusion that the "Alison-replacement" (sorry dude, have to remember your name) does an awesome job. the 3d-printing timelapses caught me off-guard, but in a positive way. And the transitions were always really smooth no hard cuts in the flow. Thanks for this awesome episode!
His audio mixing game is not up to par. Super loud commercial and missing music during the time-lapse wiring and another dead spot at the end credits. Otherwise, good.
Great execution on this build! That green looks like Mr. Yuk, which means you could wear one of those stickers to have the camera follow you around and look cool at the same time!
nex to the counterweight i would also make the pully for the y a little bigger so that it makes the beld a bit more tence and i would make the none driven puly on the x axes have a spring so that it makes the belt tighten itself
Zoom control, Use a control object that is oddly shaped so that as you turn it the pixy camera will see it as a different size. The pixy can use this size information to control the zoom of the camera. Also, you could make the control button wireless and built into the control object.
Ben, you have to design a belt tightener for your projects. Something like a tensioner or a turnbuckle that you could 3D print and use it straight on the belt or fit to your other 3D printed parts. You've had that problem many a time now.
Ben, I was looking at the video and was wondering if you had considered suspending the camera from the rail instead of supporting it? It could possibly make the camera easier to counter-balance. Just a thought.
i really like that green filliment what is it called. i might get some when i get my printer finished. also would ramping the movement so that is more a smooth speed increase whiles its move would make it easier on the eyes for viewing than sharp fast movment.
I have heard of putting the parts in a inclosed container so the vapors of the acetone would soften the ridges, but not coating them as you mentioned. Could you show us this some time? Thanks
Hi, was using a digital streaming camera not an option? If you used one, then you could have streamed the video to a computer and have OpenCV do the tracking instead of that pixy cam. One camera, two uses... and less parts to print
I wonder if you could use the Z component to calculate an approximate distance of the target from the camera and use that to zoom in and out? Some cameras have an IR remote that could be used for this purpose. I wonder if using a GoPro or similar aciton camera might have been better as these a lightweight, though don't have the aforementioned zooming capacity.
Blue and yellow are not each others complimentary colors. It's orange-blue and yellow-purple. They are the other two colors in the psychological color spectrum.
Short the windings on the Y-axis stepper when it's meant to be stationary to better hold it's position? I'm sure it's ability to hold the weight of the camera would only be slightly improved because the countering force is only generated while the stepper is being spun.
if you coudl redesign the y-axis gear on the camera mount with some locking holes, you could add a mechanism, like a spring toggle switch or somethig like that, to lock and unlock the y-axis. It would unlock right before movement, then lock it into place to keep the camera from falling over.
The moment he showed us the first part of the Camera pivot mechanism I knew he'd have issues once the weight of the Camera moved far enough off center to the shaft. Should have used a design similar to a Camera Gimbal like they're used on a Multicopter where all the axes are at the center of the mass ( Camera ). Which kind of makes me wonder why they've chosen a rail design instead of having the camera rotate horizontally - Was that explained somewhere?
Cool project... I think i'd have just went with a simple manual control instead of the auto tracking. I don't think it would be that big of a deal production wise, because the person already has to place the target and toggle the activation, so its not like it is completely autonomous anyway. I don't care about how smooth the operation is since I'd probably only use to get the cam into place for a shot, then edit that out later... I like the idea of a counter weight but maybe a linear actuator, or a cork-screw drive system with gears, instead of a belt, would work better in the Y-axis application.
Here's an idea: I bet Pixie can see near IR. You can put an near IR LED on your glasses and then iluminate the surface around the object. Pixie will then track and center on the ir light reflected. You can put a filter on the camcorder so it doesn't show on the video.
I would have used a GoPro, but also used a STP-MTR-17048D from automation direct. I have used this motor with many other projects and I have found that it is very smooth.
To stop the camera from falling over why not set up small motor controlled gear stopper. It can be placed where the belt is not ridding on the gear. It only engages when the camera has stopped moving and before the motor starts rotation it disengages again? It could be just a small plastic piece that fits between the gears teeth. It slides in an out depending on the status of the motor? It could function like a piston?
Or you can just put some edges that delimits the movement to not go down more than X degrees, that should work "KISS" (Keep It Simply and S...) Sorry my english I'm ESL.
Perhaps two pixie sensors to compensate for and eliminate the dead zone. And maybe a swing so the camera hangs under the thing that supports its' weight. A lighter camera like a web cam / phone might help smooth things out as well.
I agree with the two sensors, but I think it could also be done with one sensor. It could keep track of the direction it's suppose to move in, once it reaches it the 0 point or it slightly passes it, it would stop until the tracking object moves outside of the dead zone again. In addition, it could do speed throttling, so it slows down as it gets closer to the 0 point, enabling a smaller dead zone and a little more accuracy.
The dead zone is something that ben programed into his code so that the camera wouldn't constantly move because of a small or no real change in the marker position.
Michael Dombrowski The deadzone range appears to be too large in the video, the camera ether stops a little to the left @19:36 or right @20:00 of the target, although I'm sure he can dial it in to something smaller. I'm just saying with this other method, you can get it to center right on the object and then not worry about it moving back and forth, then it could start moving again once the tracking object moves out of the deadzone range or until he pushes the button.
tehaxor69 1. that is exactly the approach he used. did you not see the first part of this pair of videos? 2. the dead zone should depend entirely on distance from the camera. (medium works well at close range, but not so well at a distance) ben set it to what he wanted, he can change it later if he wants.
Stephen Oliveau No, the approach he used was to stop as soon as it got to the dead zone (150º-160)º, not the center. With this other method, you can get it to center and not jitter. At 13:55 in the other video he explains what he did.
you should make a ring that color so u don't have to have a big object in the middle of the project. and make the switch on a foot peddle. good luck if you ever try again, this one looked pretty good. But you can always make things better.
Is Cyberdyne Systems sign up on the side of Ben's work place? xD Ben should make a Terminator now with this tech and freak people out with the head moving and stuff. xD
Ben, when building stuff using belts, it is always a good idea to add a tensioning mechanism - e.g. a screw or spring pushing the idler pulley away. Or you can add a separate tensioning pulley. That would save you the trouble with the loose belts and trying to fit an exact length.
For the Y axis, it should be enough to have the stepper hold its position. It seemed powerful enough for that, if not, give it a smaller pulley to increase the gearing ratio. However, if you do this, you will probably want a heatsink on the stepper driver - a stalled stepper can pull significant current.
The happiness of felix (19:05)
Technicus Spiriticus Ahahahah :D
Bill Cosby called. He wants his sweater back! Also asked for Alyson's phone number. Not sure why.
Let's take a minute to appreciate 3D Printers, they're amazing pieces of technology!
Nah.
dude get a life
Spoderman u r nawt teh reel spodermen!
+Sdudyoy Newton yes they are
Watching this i came to the conclusion that the "Alison-replacement" (sorry dude, have to remember your name) does an awesome job. the 3d-printing timelapses caught me off-guard, but in a positive way. And the transitions were always really smooth no hard cuts in the flow.
Thanks for this awesome episode!
His audio mixing game is not up to par. Super loud commercial and missing music during the time-lapse wiring and another dead spot at the end credits. Otherwise, good.
Great execution on this build! That green looks like Mr. Yuk, which means you could wear one of those stickers to have the camera follow you around and look cool at the same time!
A cool extension to this build would be to make the target object a wireless toggle for the tracking.
One of my favourite builds so far. Wondering how good aluminum profiles will work for this kind of build.
Neat love how simple this cam tracking system was to set up, your stepper motor project still a little above my pay grade.
i love that green for the 3d printed parts
This was an awesome project. Love it, keep the ideas coming.
nex to the counterweight i would also make the pully for the y a little bigger so that it makes the beld a bit more tence
and i would make the none driven puly on the x axes have a spring so that it makes the belt tighten itself
"For the price of a coffee, you can adopt a stepper motor" haha..
lovin the green
Yes! yet another sliders reference. Love that little known show!
Where is there a list of the electronic parts and a schematic to see how all of this worked together?
Thanks
make a raspberry pi cluster
Tinkernut did that.
gudenaurock yes tinkernut did it. Ben can do it to
gudenaurock Is that another DIY video guy?
Yes but more software
Technicus Spiriticus I'd like to see a media center/file stream setup using raspi for use with HDMI --> TV
Oh geez I thought you were talking about the TV show Sliders!!! So confused. Thanks for clearing that up.
Welcome Max!
Have you got a place where the code is written or can I get a copy please?
That turned out pretty sweet!
Where is this Screwdriver from? at 12:20
Ben, do you even use that camera in your videos?
Zoom control, Use a control object that is oddly shaped so that as you turn it the pixy camera will see it as a different size. The pixy can use this size information to control the zoom of the camera. Also, you could make the control button wireless and built into the control object.
Ben, you have to design a belt tightener for your projects. Something like a tensioner or a turnbuckle that you could 3D print and use it straight on the belt or fit to your other 3D printed parts. You've had that problem many a time now.
What is the program you use for 3d modeling The Ben Heck Show
Ben, I was looking at the video and was wondering if you had considered suspending the camera from the rail instead of supporting it? It could possibly make the camera easier to counter-balance. Just a thought.
i really like that green filliment what is it called. i might get some when i get my printer finished. also would ramping the movement so that is more a smooth speed increase whiles its move would make it easier on the eyes for viewing than sharp fast movment.
really enjoying the 3D printer time lapse, very nice
20:55 a worm gear drive or a high torque rc servo instead of a stepper.
Wow! What a Cosbytastic sweater!
Where can I buy this I need this thing
I have heard of putting the parts in a inclosed container so the vapors of the acetone would soften the ridges, but not coating them as you mentioned. Could you show us this some time?
Thanks
I like your show. Could you tell us a little bit about the software behind Auto Tracking Camera. Thanks from Europe.
Felix, the face of pure joy
RIP headphone users at that commercial...
This is the new guy editing the video.
Sfolter
Forgot the volume but didn't forget to edit Ben's sister out of the intro.
Hi, was using a digital streaming camera not an option?
If you used one, then you could have streamed the video to a computer and have OpenCV do the tracking instead of that pixy cam. One camera, two uses... and less parts to print
If someone handed you all the parts, how long would it take you to make another one? Parts meaning finished 3d prints and out of box electronics.
Ben, in your next episode, could you use a 2008 macbook screen? I have an idea for a project of my own, but I don't know how to execute it.
I wonder if you could use the Z component to calculate an approximate distance of the target from the camera and use that to zoom in and out? Some cameras have an IR remote that could be used for this purpose. I wonder if using a GoPro or similar aciton camera might have been better as these a lightweight, though don't have the aforementioned zooming capacity.
Blue and yellow are not each others complimentary colors. It's orange-blue and yellow-purple. They are the other two colors in the psychological color spectrum.
Short the windings on the Y-axis stepper when it's meant to be stationary to better hold it's position? I'm sure it's ability to hold the weight of the camera would only be slightly improved because the countering force is only generated while the stepper is being spun.
I do think that the captions come up to frequently and should scroll on left right?
What`s the model of the Canon camera, please? :)
if you coudl redesign the y-axis gear on the camera mount with some locking holes, you could add a mechanism, like a spring toggle switch or somethig like that, to lock and unlock the y-axis. It would unlock right before movement, then lock it into place to keep the camera from falling over.
I'd not have bothered with the massive rail but just made a stepper motor do Yaw. Any reason why not?
If you want to know if two wires are connected, you can just use the diode setting and it will beep if they are connected
can it recognize deck of cards or different currency notes?
also why not a three axis auto tracking camera?
The moment he showed us the first part of the Camera pivot mechanism I knew he'd have issues once the weight of the Camera moved far enough off center to the shaft.
Should have used a design similar to a Camera Gimbal like they're used on a Multicopter where all the axes are at the center of the mass ( Camera ).
Which kind of makes me wonder why they've chosen a rail design instead of having the camera rotate horizontally - Was that explained somewhere?
Ahaha, the Wiki article at the end. Nicely done.
Very cool, nice video guys.
why not make it take a QD plate you already have laying around?
how do you figure how big of a camera/lens you can move?
Everytime when I'm watching your videos, I'm thinking "Damn, I need a 3D printer!".. ;-)
dam BEN you should have worked out the code for us to see as well
Cool project... I think i'd have just went with a simple manual control instead of the auto tracking. I don't think it would be that big of a deal production wise, because the person already has to place the target and toggle the activation, so its not like it is completely autonomous anyway. I don't care about how smooth the operation is since I'd probably only use to get the cam into place for a shot, then edit that out later... I like the idea of a counter weight but maybe a linear actuator, or a cork-screw drive system with gears, instead of a belt, would work better in the Y-axis application.
how to stop the camera when the object was in center?
hey ben do you know how i would be able to transfer my pictures,movies and data from a ps3 to a ps4
Nice job!
can i know the program code??
Ben, can you pump out more videos? If you respond, thank you!
you should make a touch screen phone and publish it
I see Max's acting is just as good as Alyson's.
If I had access to all that tech and tools, I would be Ben Heck!
Notice the page up on the screen at 21:24.
en.wikipedia.org/wiki/Cybernetic_revolt
couldn't have an episode without a The Room reference huh?
Oh dear lord I got your shaft reference.
Encase a simple radio circuit in the tracker. Add a button to the tracker and you have a wireless on off for your tracking...
Make a shawarma dispenser!
I love that the laptop is open to the Wikipedia page for Cybernetic Revolt at the end.
That is quite the sweater!
was totally waiting for him to "fall asleep" in that intro sequence...
this was really great and use full.....some more focus on the program can be good
The VR stycker on the background is the same that we used to put on car of my city in italy lol
I AM GONNA NEVER WATCH THAT AGAIN IF THEY KEEP USING THAT SONG WITH THAT NOISE EVERY 5 SECONDS 5:20
Nice job on the vid max
I would have tightened the camera mount, so it doesn't shake as much when moving.
Of course it was showing an article about cybernetic revolt...
I would have added an additional Pan axis, besides the Tilt and X ones.
I have an idea for the next portable game console. A Nintendo GameCube portable. If you do use my suggestion, please say who it's from. Me.
Awesome! Inspirational. Innovative. Original - Oh, those last two were ...uh, redundant...sorry. But, suuu-weet project well done. Thanks.
18:57 felix. like a boss.
My 80 year old Bolex uses the 3/4 screw.
The beauties of having a good 3D printer.
Here's an idea: I bet Pixie can see near IR. You can put an near IR LED on your glasses and then iluminate the surface around the object. Pixie will then track and center on the ir light reflected. You can put a filter on the camcorder so it doesn't show on the video.
I would have used a GoPro, but also used a STP-MTR-17048D from automation direct. I have used this motor with many other projects and I have found that it is very smooth.
I like this Max guy...
why was it called alec?
To stop the camera from falling over why not set up small motor controlled gear stopper. It can be placed where the belt is not ridding on the gear. It only engages when the camera has stopped moving and before the motor starts rotation it disengages again? It could be just a small plastic piece that fits between the gears teeth. It slides in an out depending on the status of the motor? It could function like a piston?
Or you can just put some edges that delimits the movement to not go down more than X degrees, that should work "KISS" (Keep It Simply and S...)
Sorry my english I'm ESL.
Agustin Repetto or add a spring that's just able to lift the cameras wheigt
Perhaps two pixie sensors to compensate for and eliminate the dead zone. And maybe a swing so the camera hangs under the thing that supports its' weight. A lighter camera like a web cam / phone might help smooth things out as well.
I agree with the two sensors, but I think it could also be done with one sensor.
It could keep track of the direction it's suppose to move in, once it reaches it the 0 point or it slightly passes it, it would stop until the tracking object moves outside of the dead zone again. In addition, it could do speed throttling, so it slows down as it gets closer to the 0 point, enabling a smaller dead zone and a little more accuracy.
The dead zone is something that ben programed into his code so that the camera wouldn't constantly move because of a small or no real change in the marker position.
Michael Dombrowski The deadzone range appears to be too large in the video, the camera ether stops a little to the left @19:36 or right @20:00 of the target, although I'm sure he can dial it in to something smaller. I'm just saying with this other method, you can get it to center right on the object and then not worry about it moving back and forth, then it could start moving again once the tracking object moves out of the deadzone range or until he pushes the button.
tehaxor69
1. that is exactly the approach he used. did you not see the first part of this pair of videos?
2. the dead zone should depend entirely on distance from the camera. (medium works well at close range, but not so well at a distance) ben set it to what he wanted, he can change it later if he wants.
Stephen Oliveau No, the approach he used was to stop as soon as it got to the dead zone (150º-160)º, not the center. With this other method, you can get it to center and not jitter. At 13:55 in the other video he explains what he did.
Can we get Kari Byron as your lovely assistant for TBHS?
The slider is actually a professional camera slider. google for Konova K2 XD
189$
***** that could very well be :P
***** he said that element14 sent it to them and "use your $800 professional camera slider to make a $500 3D printer"
MEALIVEUDEAD but it's 189$ :O
***** he was exaggerating...
If you were to use a servo from a large RC airplane would provide more than enough torq for the Y axis.
it's not a rock, it's a mineral! MINERAL!
you should make a ring that color so u don't have to have a big object in the middle of the project. and make the switch on a foot peddle. good luck if you ever try again, this one looked pretty good. But you can always make things better.
i loved it :) im going to try this at home.
Forever alone camera setup.
Is Cyberdyne Systems sign up on the side of Ben's work place? xD
Ben should make a Terminator now with this tech and freak people out with the head moving and stuff. xD
god bless benheck
The y-axis really should have used a worm gear on the stepper motor.