 | | | | | | |
| |
|
Hello all,
For an animation project I'm starting, a camera tilt and pan unit became
necessary. Here's half of it, the tilt part. (I haven't designed the pan part
yet) Forgive me if I'm misposting this; this is my first major MOC post.
Deep links: (forgive me for simply providing URL's. I haven't quite figured out
the FTX system here yet)
<http://brickshelf.com/gallery/Buachaille/camtilt/aaa_overview.jpg>
An overview of the mechanism. It's powered by three hand pumps and utilizes
four
pistons, three switches and a homebrew airtank to move the camera.
The camera sits securely on a small platform. The platform was designed to hold
the camera without allowing any camera shaking or wobbling - and it worked.
That
camera sits in there solid as a rock. The tilt mechanism may be a little shaky,
but the camera ain't goin' nowhere.
The platform is manipulated by four pistons, each connected upside-down to a
corner of the platform. One switch manipulates two pistons (a set in the front
and a set in the back), allowing me to raise, lower or tilt the platform up or
down. A fairly wide range of motion - I would guess around 30 degrees.
<http://brickshelf.com/gallery/Buachaille/camtilt/pistons.jpg>
The pistons are shown here, connected to the platform. I'm sorry if the
pictures
are unclear where the camera sits; I had to use it to take the photos. :)
<http://brickshelf.com/gallery/Buachaille/camtilt/cam.jpg>
This photo shows where the camera sits. The lense aperture fits right next to
the 1x2 white slope on the left side. The LCD viewer is located just above the
2x2 tiles on the back, allowing me to see what I'm doing while I'm manipulating
switches.
<http://brickshelf.com/gallery/Buachaille/camtilt/switches.jpg>
I kept it simple and only used three switches. The left and right switches
control a set of pistons each; the right switch controls the front and the left
switch controls the back. Moving a switch to the UP position raises the
platform, moving it down obviously lets it down. A wide variety of camera
angles
are possible by manipulating both switches together.
The middle switch is the main control valve. The line coming from the airtank
(shown below) plugs into this. One output line from this switch feeds the other
two switches and the other output line is a release valve. I've fount I can
pump
the hand pumps at least 80 times before I pop a hose. Remember that the pumps
are connected together, so that's 240 single pump compressions. In a very
scientific test, I found that the airtank, fully pressurized, could flip (from
one state to another) one set of pistons with no load around 75 times and a
single small piston around 390 times.
<http://brickshelf.com/gallery/Buachaille/camtilt/tank.jpg>
An 8oz/237ml bottle with a couple of hard tubes hotglued in, connected to three
hand pumps. It's quite a job to fully pressurize but once you do it just keeps
going...and going...and going...well, 390 times anyway. I found the hotglue to
work surprisingly well - I haven't had a single leak from the airtank. (my
hoses
tend to leak a lot, but only because the cat likes to get her claws into them)
The last three photos show the range of motion available to the camera:
<http://brickshelf.com/gallery/Buachaille/camtilt/camrange1.jpg>
<http://brickshelf.com/gallery/Buachaille/camtilt/camrange2.jpg>
<http://brickshelf.com/gallery/Buachaille/camtilt/camrange3.jpg>
As always, PLMKWYT,
~Mike "hoping-this-overlong-post-goes-through" Thorn
| | | | | | | | | | | | | | | |
In lugnet.animation, Mike Thorn wrote:
> Hello all,
>
> For an animation project I'm starting, a camera tilt and pan unit became
> necessary. Here's half of it, the tilt part. (I haven't designed the pan part
> yet) Forgive me if I'm misposting this; this is my first major MOC post.
>
> Deep links: (forgive me for simply providing URL's. I haven't quite figured out
> the FTX system here yet)
>
> <http://brickshelf.com/gallery/Buachaille/camtilt/aaa_overview.jpg>
Hi Mike,
I could not get any of your links to work..... Any hints?
Kevin
| | | | | | | | | | | | | | | | | | | | | | |
| |
|
In lugnet.animation, Kevin L. Clague wrote:
> In lugnet.animation, Mike Thorn wrote:
> > Hello all,
> >
> > For an animation project I'm starting, a camera tilt and pan unit became
> > necessary. Here's half of it, the tilt part. (I haven't designed the pan part
> > yet) Forgive me if I'm misposting this; this is my first major MOC post.
> >
> > Deep links: (forgive me for simply providing URL's. I haven't quite figured out
> > the FTX system here yet)
> >
> > <http://brickshelf.com/gallery/Buachaille/camtilt/aaa_overview.jpg>
>
> Hi Mike,
>
> I could not get any of your links to work..... Any hints?
They work like a charm to me!?
Cool use of pneumatics Mike, how is control? Pneumatics tend to "jump" from one
position to another that would give a pretty non-smooth camera pan?
The airtank (yours and the LEGO version);
Does it really need an in and out? Would not an in be enough since it would act
as out as the preassure in the rest of the system drops?
--
Best regards,
/Tobbe
<http://www.lotek.nu>
(remove SPAM when e-mailing)
| | | | | | | | | | | | | | | | | | | | | | | | | | | |
| |
|
At 05:03 AM 10/6/2003, you wrote:
> In lugnet.animation, Kevin L. Clague wrote:
> > In lugnet.animation, Mike Thorn wrote:
> > > Hello all,
> > >
> > > For an animation project I'm starting, a camera tilt and pan unit became
> > > necessary. Here's half of it, the tilt part. (I haven't designed the • pan part
> > > yet) Forgive me if I'm misposting this; this is my first major MOC post.
> > >
> > > Deep links: (forgive me for simply providing URL's. I haven't quite • figured out
> > > the FTX system here yet)
> > >
> > > <http://brickshelf.com/gallery/Buachaille/camtilt/aaa_overview.jpg>
> >
> > Hi Mike,
> >
> > I could not get any of your links to work..... Any hints?
>
> They work like a charm to me!?
>
> Cool use of pneumatics Mike, how is control? Pneumatics tend to "jump"
> from one
> position to another that would give a pretty non-smooth camera pan?
Yeah, it does somewhat. One thing that helps is not to pressurize the tank
all the way. If there isn't much pressure it doesn't "jump" as much and
helps keep increments small. Another thing that's helpful is not to flip
the switch all the way - just move it to a point where a tiny little bit of
air leaks through. I really don't want to tilt the platform more than a
degree or two, for the best results - it takes a lot of frames to make
something smooth, more than you would think.
It's really not quite the best way to do it, just because control is hard,
but I'm not sure how else I could do it that would maintain the same
steadiness and small size (which is an issue).
Once again, a regulated switch with a little more control would be nice.
And a pressure meter, since John Barnes dissolved his sensor stuff.
> The airtank (yours and the LEGO version);
> Does it really need an in and out? Would not an in be enough since it
> would act
> as out as the preassure in the rest of the system drops?
I think you're right, except for one thing: you would not be able to use
the system while you're pumping. Air would constantly be pumped IN and
there would be no way for air to get OUT. With a compressor, for instance,
this would be impossible to use.
And if one used a huge bottle (like a 2-liter soft drink bottle) and
pressurized it beforehand, there's no reason why you couldn't cap one pipe
for a one-outlet tank. It would be pretty simple, I think. The antennas
come in handy for once.
~Mike
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
| |
|
In lugnet.animation, Mike Thorn wrote:
> At 05:03 AM 10/6/2003, you wrote:
> > In lugnet.animation, Kevin L. Clague wrote:
> > > In lugnet.animation, Mike Thorn wrote:
> > > > Hello all,
> > > >
> > > > For an animation project I'm starting, a camera tilt and pan unit became
> > > > necessary. Here's half of it, the tilt part. (I haven't designed the pan part
> > > > yet) Forgive me if I'm misposting this; this is my first major MOC post.
> > > >
> > > > Deep links: (forgive me for simply providing URL's. I haven't quite figured out
> > > > the FTX system here yet)
> > > >
> > > > <http://brickshelf.com/gallery/Buachaille/camtilt/aaa_overview.jpg>
> > >
> > > Hi Mike,
> > >
> > > I could not get any of your links to work..... Any hints?
> >
> > They work like a charm to me!?
They work for me now also.
> >
> > Cool use of pneumatics Mike, how is control? Pneumatics tend to "jump"
> > from one
> > position to another that would give a pretty non-smooth camera pan?
>
> Yeah, it does somewhat. One thing that helps is not to pressurize the tank
> all the way. If there isn't much pressure it doesn't "jump" as much and
> helps keep increments small. Another thing that's helpful is not to flip
> the switch all the way - just move it to a point where a tiny little bit of
> air leaks through. I really don't want to tilt the platform more than a
> degree or two, for the best results - it takes a lot of frames to make
> something smooth, more than you would think.
I guess I was curious why you used pneumatics. I would think a
motor/gear/rotation sensor would give you very fine angular resolution and
variable speed. You also would only need to control either the front or the
back, but not both.
By using a motor/rotation sensor, you could have the RCX compute the desired
tilt angle from some kind of camera script that describes certain way points at
specific frame numbers (key frames?) or time.
You could provide a similar mechanism for pan.
>
> It's really not quite the best way to do it, just because control is hard,
> but I'm not sure how else I could do it that would maintain the same
> steadiness and small size (which is an issue).
>
> Once again, a regulated switch with a little more control would be nice.
> And a pressure meter, since John Barnes dissolved his sensor stuff.
Tecno-stuff sells pressure sensors. I just bought one.
>
> > The airtank (yours and the LEGO version);
> > Does it really need an in and out? Would not an in be enough since it
> > would act
> > as out as the preassure in the rest of the system drops?
>
> I think you're right, except for one thing: you would not be able to use
> the system while you're pumping. Air would constantly be pumped IN and
> there would be no way for air to get OUT. With a compressor, for instance,
> this would be impossible to use.
Hmmmm. I'm not sure this is true. The pressure wave that is created by you
pumping will get distributed to both the tank and the platform lift pistons.
>
> And if one used a huge bottle (like a 2-liter soft drink bottle) and
> pressurized it beforehand, there's no reason why you couldn't cap one pipe
> for a one-outlet tank. It would be pretty simple, I think. The antennas
> come in handy for once.
Light Sabres work well too.
>
> ~Mike
Kevin
| | | | | | | | | | | | | | | | | | | | | | | |
At 10:47 AM 10/5/2003, you wrote:
> In lugnet.animation, Mike Thorn wrote:
> > Hello all,
> >
> > For an animation project I'm starting, a camera tilt and pan unit became
> > necessary. Here's half of it, the tilt part. (I haven't designed the • pan part
> > yet) Forgive me if I'm misposting this; this is my first major MOC post.
> >
> > Deep links: (forgive me for simply providing URL's. I haven't quite • figured out
> > the FTX system here yet)
> >
> > <http://brickshelf.com/gallery/Buachaille/camtilt/aaa_overview.jpg>
>
> Hi Mike,
>
> I could not get any of your links to work..... Any hints?
It works for me...odd. Anyway, I made that post about half an hour after I
uploaded the pictures, so I didn't think a moderated folder link would be
much good. Now that it is moderated...
http://brickshelf.com/cgi-bin/gallery.cgi?f=58539
Thanks for looking,
~Mike
| | | | | | | | | | | | | | | | | | |
| |
|
In lugnet.animation, Mike Thorn wrote:
> Hello all,
>
> For an animation project I'm starting, a camera tilt and pan unit became
> necessary. Here's half of it, the tilt part. (I haven't designed the pan part
> yet) Forgive me if I'm misposting this; this is my first major MOC post.
>
> Deep links: (forgive me for simply providing URL's. I haven't quite figured out
> the FTX system here yet)
Hi Bauchaile,
I don't post with in-line images because not everyone has broadband.
>
> <http://brickshelf.com/gallery/Buachaille/camtilt/aaa_overview.jpg>
>
> An overview of the mechanism. It's powered by three hand pumps and utilizes
> four
> pistons, three switches and a homebrew airtank to move the camera.
>
> The camera sits securely on a small platform. The platform was designed to hold
> the camera without allowing any camera shaking or wobbling - and it worked.
> That
> camera sits in there solid as a rock. The tilt mechanism may be a little shaky,
> but the camera ain't goin' nowhere.
I love to see new pneumatic projects!
I was curious why you needed three switches. It seems like a front switch and a
back switch would be sufficient.
It seems like you have four possible positions for the camera:
Front contracted, back contracted - camera low but level
Front expanded, back contracted - camera tilted up
Front expanded, back expanded - camera high but level
Front contracted, back expanded - camera tilted down
If you'd like more possible angles, you canhook two of your pistons back to back
in each corner of the platform.
This cwould give you thrr positions in the front and three positions in the
back, for a total of 9 tilt position/angles.
>
> The platform is manipulated by four pistons, each connected upside-down to a
> corner of the platform. One switch manipulates two pistons (a set in the front
> and a set in the back), allowing me to raise, lower or tilt the platform up or
> down. A fairly wide range of motion - I would guess around 30 degrees.
>
> <http://brickshelf.com/gallery/Buachaille/camtilt/pistons.jpg>
Yes it will give you a wide range of motion, but pneumatic pistons controlled by
switches are binary in nature, so you only really have 4 steady state
position/angles for the camera.
If you control the platform pistons using pistons instead of switches, you can
have a more analog situation. You can hook two pistons together. By expanding
and contracting one piston, you can expand/contract the other piston.
>
> The pistons are shown here, connected to the platform. I'm sorry if the
> pictures
> are unclear where the camera sits; I had to use it to take the photos. :)
>
> <http://brickshelf.com/gallery/Buachaille/camtilt/cam.jpg>
>
> This photo shows where the camera sits. The lense aperture fits right next to
> the 1x2 white slope on the left side. The LCD viewer is located just above the
> 2x2 tiles on the back, allowing me to see what I'm doing while I'm manipulating
> switches.
>
> <http://brickshelf.com/gallery/Buachaille/camtilt/switches.jpg>
>
> I kept it simple and only used three switches. The left and right switches
> control a set of pistons each; the right switch controls the front and the left
> switch controls the back. Moving a switch to the UP position raises the
> platform, moving it down obviously lets it down. A wide variety of camera
> angles
> are possible by manipulating both switches together.
>
> The middle switch is the main control valve. The line coming from the airtank
> (shown below) plugs into this. One output line from this switch feeds the other
> two switches and the other output line is a release valve. I've fount I can
> pump
> the hand pumps at least 80 times before I pop a hose. Remember that the pumps
> are connected together, so that's 240 single pump compressions. In a very
> scientific test, I found that the airtank, fully pressurized, could flip (from
> one state to another) one set of pistons with no load around 75 times and a
> single small piston around 390 times.
>
> <http://brickshelf.com/gallery/Buachaille/camtilt/tank.jpg>
>
> An 8oz/237ml bottle with a couple of hard tubes hotglued in, connected to three
> hand pumps. It's quite a job to fully pressurize but once you do it just keeps
> going...and going...and going...well, 390 times anyway. I found the hotglue to
> work surprisingly well - I haven't had a single leak from the airtank. (my
> hoses
> tend to leak a lot, but only because the cat likes to get her claws into them)
>
> The last three photos show the range of motion available to the camera:
>
> <http://brickshelf.com/gallery/Buachaille/camtilt/camrange1.jpg>
> <http://brickshelf.com/gallery/Buachaille/camtilt/camrange2.jpg>
> <http://brickshelf.com/gallery/Buachaille/camtilt/camrange3.jpg>
>
> As always, PLMKWYT,
Fun project! Thanks for sharing!
>
> ~Mike "hoping-this-overlong-post-goes-through" Thorn
Kevin
| | | | | | | | | | | | | | | | | |
| |
|
At 08:20 AM 10/6/2003, you wrote:
> I was curious why you needed three switches. It seems like a front switch
> and a
> back switch would be sufficient.
>
> It seems like you have four possible positions for the camera:
>
> Front contracted, back contracted - camera low but level
>
> Front expanded, back contracted - camera tilted up
>
> Front expanded, back expanded - camera high but level
>
> Front contracted, back expanded - camera tilted down
Yes, you're right - two switches would have been entirely sufficient. I
just wanted another switch so I could add a release valve, that's all. :) I
think originally I wanted something in there so I could bypass the tank and
pump directly to the pistons, but somehow that got implemented as a release
valve instead. Don't know how that happened. I may add a bypass link in
there someday.
> If you'd like more possible angles, you canhook two of your pistons back
> to back
> in each corner of the platform.
Except that I only have two more pistons. :)
> This cwould give you thrr positions in the front and three positions in the
> back, for a total of 9 tilt position/angles.
Nine binary positions, anyway. I still need the intermediate positions to
make the final tracking shot look smooth. I wish there was a way to only
allow a specific quantity of air into the pistons.
> Yes it will give you a wide range of motion, but pneumatic pistons
> controlled by
> switches are binary in nature, so you only really have 4 steady state
> position/angles for the camera.
Correct - and they'll be used quite a bit, as a solid position to hold the
camera. I really ought to CAD this whole thing with LSynth so I can put it
back together eventually if I take it apart.
> If you control the platform pistons using pistons instead of switches, you can
> have a more analog situation. You can hook two pistons together. By
> expanding
> and contracting one piston, you can expand/contract the other piston.
That's a good idea, something that hadn't occurred to me. This would
eliminate the pumps and the airtank and the switches, and a lot more of the
tubing. However, like I said earlier, I don't have enough pistons to do
this. I really need two control pistons to flip the two 'working' pistons.
There isn't enough air contained in one piston to fully flip two working
pistons, especially under load.
I could, though, use my remaining pistons together to just move one side of
the platform. I don't think I need to do any full tilts (fully up to fully
down) at any point in the film, so I could rig it for just one half of the
platform and switch the control pistons to the other set when I get to that
point. I like that idea, I'll try it.
> Fun project! Thanks for sharing!
Thanks for looking at it! I appreciate the feedback.
~Mike
| | | | | | | | | | | | | | | | | |
| |
|
In lugnet.animation, Mike Thorn wrote:
> At 08:20 AM 10/6/2003, you wrote:
[snip]
>
> > If you'd like more possible angles, you canhook two of your pistons back
> > to back
> > in each corner of the platform.
>
> Except that I only have two more pistons. :)
Ah..., but you don't really need the back and the front controlled by pistons do
you?
I would think that keeping the front hinged and the back actuated would give you
what you want. With two pistons ganged together, you can achieve tilt down,
level and tilt up. Then you have four more pistons to do whatever you want.
[snip]
>
> > If you control the platform pistons using pistons instead of switches, you can
> > have a more analog situation. You can hook two pistons together. By
> > expanding
> > and contracting one piston, you can expand/contract the other piston.
>
> That's a good idea, something that hadn't occurred to me. This would
> eliminate the pumps and the airtank and the switches, and a lot more of the
> tubing. However, like I said earlier, I don't have enough pistons to do
> this. I really need two control pistons to flip the two 'working' pistons.
> There isn't enough air contained in one piston to fully flip two working
> pistons, especially under load.
Ah, but if you simply hinge the front, and control the back with a single
piston, that is driven by a single piston, you can get very fine control and
still have four pistons left over.
>
> I could, though, use my remaining pistons together to just move one side of
> the platform. I don't think I need to do any full tilts (fully up to fully
> down) at any point in the film, so I could rig it for just one half of the
> platform and switch the control pistons to the other set when I get to that
> point. I like that idea, I'll try it.
>
> > Fun project! Thanks for sharing!
>
> Thanks for looking at it! I appreciate the feedback.
You are welcome.
>
> ~Mike
Kevin
| | | | | | | | | | | | | | | | | |
| |
|
At 12:03 PM 10/6/2003, you wrote:
> [bigger snip]
>
> Ah..., but you don't really need the back and the front controlled by
> pistons do
> you?
>
> I would think that keeping the front hinged and the back actuated would
> give you
> what you want. With two pistons ganged together, you can achieve tilt down,
> level and tilt up. Then you have four more pistons to do whatever you want.
You really don't *need* both sides actuated, but it allows one to raise the
platform as well. Not really necessary, I suppose.
I presume you mean two pistons ganged end-to-end. I think this would work
much better, but so far have not found any really good ways to do this
without putting undue stress on at least some of the connection. Do you
have a good geometry in mind?
> [snip]
>
> >
> > > If you control the platform pistons using pistons instead of switches, • you can
> > > have a more analog situation. You can hook two pistons together. By
> > > expanding
> > > and contracting one piston, you can expand/contract the other piston.
> >
> > That's a good idea, something that hadn't occurred to me. This would
> > eliminate the pumps and the airtank and the switches, and a lot more of the
> > tubing. However, like I said earlier, I don't have enough pistons to do
> > this. I really need two control pistons to flip the two 'working' pistons.
> > There isn't enough air contained in one piston to fully flip two working
> > pistons, especially under load.
>
> Ah, but if you simply hinge the front, and control the back with a single
> piston, that is driven by a single piston, you can get very fine control and
> still have four pistons left over.
Very true. I'll have to try that. Thanks for the idea!
~Mike
| | | | | | | | | | | | | | | | | |
| |
|
In lugnet.animation, Mike Thorn wrote:
> At 12:03 PM 10/6/2003, you wrote:
> > [bigger snip]
> >
> > Ah..., but you don't really need the back and the front controlled by
> > pistons do
> > you?
> >
> > I would think that keeping the front hinged and the back actuated would
> > give you
> > what you want. With two pistons ganged together, you can achieve tilt down,
> > level and tilt up. Then you have four more pistons to do whatever you want.
>
> You really don't *need* both sides actuated, but it allows one to raise the
> platform as well. Not really necessary, I suppose.
>
> I presume you mean two pistons ganged end-to-end. I think this would work
> much better, but so far have not found any really good ways to do this
> without putting undue stress on at least some of the connection. Do you
> have a good geometry in mind?
It just so happens that I have a good geometry :^)
http://www.brickshelf.com/cgi-bin/gallery.cgi?f=38106
[snip]
> Very true. I'll have to try that. Thanks for the idea!
Sure. It is always a pleasure to talk to you Mike.
>
> ~Mike
Kevin
| | | | | | |
