Don't know how to help you directly, but you might consider using a different codec. Some codecs will be more lossy in their compression than others.

There is a freeware screen capture program called CamStudio that has a lossless codec available for download. Once installed, you may be able to choose the lossless codec for your video export, giving you a file that is gigantic in size and hard to work with, but which may then be transcoded successfully to another format with better results.

I am guessing, but it's worth a shot.

@mitcorb said:

Question- what is the function of the U bar suspended down below?

The U-bar is used by the tractor driver to gauge the height of the plants he is running over. The IR sensor rack is raised so the bottom of that bar brushes over the tops of the plants. I think that determines the area that the sensors can "see".

OK. Here is a machine I simulated in SketchUp/Sketchyphysics.
It is a mechanism to hold a set of infrared sensors above a cotton field. The IR reflectance of plants indicates some aspect of plant health, and so the sensors report plant reflectance to a small palmtop computer, which uses the data to apply fertilizer to the specific part of the field that is nutrient-poor.

Our machine is mounted on a Landini Mythos tractor. I slightly modified a similar-looking tractor that i downloaded from the 3D Warehouse, and then mounted the sensor mechanism. The elevation control on the model, as in the actual equipment, is controlled using a single piston that pushes on a pantograph, so the sensors remain level as they move up and down.

One surprise for me was that when the piston is extended from its lowest position, the swinging linker arm rotates forward and then suddenly rotates back. On later examination of video of the real machine, I found that the mechanism does the same thing in the field. This was rather satisfying to see.

Here is an embedded video of the machine and its model:

[flash=640,320:1py3cvuq]http://www.youtube.com/v/ZUL6S5JMxls[/flash:1py3cvuq]

I made up a bunch of videos related to modeling a house for an engineering course.
Check YouTube, and search for "ET365"

Could you describe your intended model? From what I have read, you want to make some kind of base with a motor on it. The motor is connected to another object that will rotate with the motor.

You have constructed a 3D box and grouped its parts together, making the base. You then make a motor and connect it to the base box somehow (by grouping?). You then make a second 3D object and try to use the Joint Connector to connect this object to the motor.

Is this correct?
If so, one question is whether your base box needs to be able to move or be moved. If it is just sitting there, it really doesn't need to be grouped and attached to the motor; the motor can be placed arbitrarily in space, and then joint-connected to your rotating object. The motor will not move, since it is itself a static object.

OK. I made up a video.
I recommend looking at the thread on Hinges and Joints for details about creating solids and joining them.

If the instructions are not clear, please comment here. I will try to fix any problems to make the tutorial as understandable as necessary.

[flash=640,320:3e8p8bw6]http://www.youtube.com/v/J5IApgKEZCM[/flash:3e8p8bw6]

(EDIT: Now with audible voiceover!)

OK. I can't make it work using the procedure that I can see in the video. It may be that earlier versions of SketchyPhysics were not as picky about some things. Dunno.

I have made a model that does the same thing in the end, but it required more steps, and the solids do not interact with each other. They are able to pass through each other, which looks OK, but I personally find it unacceptable.

So, I will see if I can make up instructions for making a piston-engine-thingy that satisfies my own perferences. I'll update here as soon as I have the video uploaded. Stay tuned.

The video shows a nice construction, but there are so many details that are left unspoken. I'll see if I can produce something similar that has a voiceover. It would be a good exercise in creating a tutorial. Stay tuned!

One limitation of SketchyPhysics is that objects that are below a certain size will not move. If your individual door panels are very thin, for instance, they will not move. If you connect several thin panels together with joints, then group them all into a single group, that group will be large enough to move in the SketchyPhysics environment, but the panels themselves will be frozen relative to each other.

Sometimes, you need to scale everything up to make a simulation work, even though your dimensions will be proportionately larger than in real life.

Yes, the brain is very colorful, but as soon as it is exposed to air, it reverts to a uniform grey color.

Or, maybe that's the "color by layer" effect at work?

I truly didn't think about that when I uploaded the model. I guess a colorful brain is kind of strange.

OK, another plug for Miku Miku Dance:

Take a look at this video

[flash=425,344:15sbrqiy]http://www.youtube.com/v/DU08c8h6a2U[/flash:15sbrqiy]

This was made using the open source software "Miku Miku Dance", which is able to import SketchUp models. The trains, I think, are actually driven by the rotation of their wheels, and a "Camera Follow" effect is used to keep the view on one of the trains.

Something like this can be made using SketchyPhysics, I'm sure.

My wife is taking a neuroscience class, and so I looked for brain models. I stumbled on the Scalable Brain Atlas from the Netherlands. This site has brain scans of macaque, mouse and human, in which different labeled regions of the brain can be downloaded as 3D surfaces in .wrl (VRML) format.

Downloaded the full set for the human brain, and converted from .wrl to .3ds using MeshLab. The .3ds files were then imported into SketchUp 8 free, with each model named according to the model online, and then placed into a layer with its pair from the opposite hemisphere.

The result is a pretty nice, if a bit chunky looking, brain model in which different regions can be isolated by changing the visibility of different layers. This, combined with SketchUp's Section Plane tool and Color by Layers option makes for a potentially very useful model.

Following the label at the Scalable Brain Atlas site, the model is called Human_LPBA40_LoRes_BrainAtlas. I emailed the contact person at the Scalable Brain Atlas site to ask for permission to make this model public, and he gave me some attribution information to add. These are in the model description, and are included in the model file itself.

I hope this is of use to neuroscientists and students.

Note: I am told that the Hippocampus component is not correct (lacking part of its arch), and the Amygdala is missing. I'll have to see if I can find them.

One idea is to make a series of SP machines to hone your skills with the plug in. For instance, making a steerable motorized car is a common exercise. Another is to learn the behavior of the various joints and their controls, or to learn the limits of the dimensions of objects, to see how small or large they can be made before their physics break down.

An idea that I had, which I haven't tried to do, is to make an electric motor using passive hinges and the magnet property. Can the magnetic attraction be turned on and off in a rhythmic way that emulates the electromagnetic switching in an electric motor's rotor? Can this switching be linked to 'touching' between two objects in a manner similar to the way a motor's brushes turn the electricity on and off? While the SketchyPhysics motor joint already exists and works out of the box, an exercise of creating a motor from these principles would be quite a challenge. In the process of making such a model, you will learn many skills in the use of SketchUp and SketchyPhysics.

You may want to see the video I prepared for a different thread in this form.

I think you need to make your blocks into groups in order to make them interact with physics. Also, there is a lower limit to the size of an object that will work with physics. Once you have separately grouped each of your blocks, they should fall under gravity when you start the simulation.

There was a forum thread called Groups, Components and Joints in which I embedded a video of basic joint attachment. You might find it to be useful, since I tried to show some very basic characteristics of SketchyPhysics.

AAahhhh! The hinge joint has no controller, since it does not actually have any power. You want a "servo" joint, which is designed to be rotated around one complete turn. If you want it to spin, use a "motor".

Similarly, a slider joint probably has no controller, while a piston does.

If you're renaming the controller through the (forget what it's called. The UI button on the toolbar), you can't delete the entire name given in the dialog box. You can only change the name that is inside brackets and quote marks:

controllername1234("controllername1234")
becomes
controllername1234("NewNameForControl")

If I misunderstood you, please forgive me.

There may be some trick about how you connect the pin to the rest of your model that is relevant. I can't remember exactly how it's done, though. The idea is that if you make your objects as regular SketchUp objects, then select them all and group them, their collision geometry will be generalized, as mentioned before. That is, SP will just envelope the entire group in a single invisible rectangle that will define its physical boundaries.

You may want to make your main body as one group, and set its characteristics as "nocollision", then make a second group, consisting of your pin. Now group both of these together. The two groups may then retain their original collision boundaries.

I'll do a few experiments and take notes, then post them here. It's a common problem, and I am tired of having to figure it out from scratch every time, so having it written out would be good for me, too.

Rather than try to make your landing gear retract on your model, you should experiment with some basic SketchyPhysics exercises on simple models.

I assume you've installed SP already. Have you tried connecting two solids together with a hinge? This is a good first step to learning how the hinge joint and other rotating joints work.

You will also need to learn whether you want your model's main body to be static, so it won't fall when you start the simulation.

Next, you can connect a static object to a free object using a Servo joint, and then control the movement of the free object using a slider control. This should put you well on your way to making your landing gear work in a basic way.

@tspco said:

. . . Perhaps those of us who use it, should populate the 3D warehouse with working SP models, and while we are at it populate You-Tube with videos of working SP model. I have a few on You-Tube, but there needs more new ones.

OK, I will if you will.
Should we agree on a keyword tag for this initiative?
Maybe SketchyPhysicsPromo