When in Dome…

The Morehead Planetarium & Science Center Production Blog

Author Archive | Jim Kachelries

The first blog post I made I talked about making storyboards for the dome. To give a better idea what I’m doing, Jay was kind enough to take a picture of me in the process of making one of the boards for our next show solar system show.

What’s interesting about this process in the dome, is not only the challenege of imagining the scene and positioning the characters in the false 3D space, but the relationship and staging of the characters in the 2D sense. What my good friend and collegue Pete would say is that even with using 3d software like Maya or even 3d comps in After Effects, you’re still creating a flat image for film.

I apply that philosophy when boarding in the dome, trying to make the board pleasant across the whole surface for layout and staging.

There are some advantages and disadvantages to making boards in the dome versus flat, uncurved surface. An obvious disadvantage is that having access to a projector and dome is not a readily available option. Drawing on the dome also takes a little getting used to. What we’re seeing in the picture is myself drawing in the open source program Gimp, on a 1024×768 canvas, to match the projector’s resolution. When in full screen mode, I’m able to adjust the positioning so that it covers the whole dome surface. Since the projector has a correct distortion to fit the curved surface, I need to match that while drawing in Gimp. The result is a correctly distorted image, that will show up properly on the dome. Here’s a couple of examples of finished frames.

In this sequence our heroine is hanging up on her holographic father. This sort of board is a little more flat than what they normally would look like, as there aren’t any extreme angles. It would be very feasible to even make boards like these without the use of a dome to project on, just needing to remember where the front, right, left, and back positions are.

This is a more extreme example. This particular board would best be done inside the dome environment.

The advantages are obvious, there is no guesswork, and the director would get to know exactly what the shots will look like before its brought further down the production line. A frequent question asked when boards are made on a 4:3 or widescreen is, “Well what’s the back look like?” This method is a way to show what that back actually would look like, or a close aproximation to.

Starting on our new show, we begin with our most favorite (mine anyways) portion of the production process:


With our script getting a coat of finishing wax, our characters have been essentially locked down. With no major changes to be made to them, we can begin our character sketches! The following are some examples of our rough character concepts. The next step would be to make more locked down turn-arounds when we move onto the modeling in our 3d program of choice.










So we animated our character, created a cycle in the trax editor, and increased the number of times that our cycle will repeat. When we look at our trax editor, we should see something like this, where we can see the length of our original cycle, and the number of times the cycle repeats itself. It’s represented through the lighter colored ‘phantom’ cycles.


Here’s where things are going to start to get a little weird. You may think at first that now you can apply the time warp, as I had thought myself. But the time warp only affects and represents the cycle, it does not include the post-cycle repeats.



Since the goal is to have our character slow down and then speed up again, we’ll need to make this one long animation clip. We can do this by merging what we have, making it a new animation clip. We’ll want to select our clip and go to:

Edit > Merge > Dialogue Box


Now we can pick what we want the name for our clip to be. We can just leave the Add to Trax radio button clicked so that the trax editor will be automatically updated with our new merged animation.


With our new merged animation, its time to do some speed adjustments with time warp! Go ahead and select our clip and go to:

Create > Time Warp > Dialogue Box


Go ahead and hit apply (Option to Enable Time Warp Curve checked), and open up the graph editor to check out our new keys for the time warp. With the animation clip selected in the trax editor go to:

Window > Animation Editors > Graph Editor

And select the Time Warp in your clip so that we can see just those keys.


Having applied a time warp to the clip, Maya will automatically create two keys. No matter how long you were wanting your animation to be, there will always be a key at 0,0 and 100,100. What’s very important to know is that this represents the percentage of the clip. The X axis is no longer represents the frames of our animation, but the percentage value of where we are in the trax editor clip. Frame 25 in the graph editor is 25% the way into our clip. Frame 75 is 75% into it, and frame 100 is the end of the trax editor clip, or in other words 100%.

The same happened to our Y axis. The Y axis has switched to percentages now, and those percentages represent the animation inside our trax editor clip. When the X represents what time the change happens in the clip as it plays, the Y represents the progression through the animation inside the clip.

So we have our normal graph here:


We’ll never want to exceed 100 on the X axis, as you can’t have greater than 100% of the clip. If you were wanting to suddenly halt the animation halfway through, it would look like this:


What this says is that once we reach the halfway point of the entire clip, we will no longer progress through the clip. Even though we’re still continuing along the X axis, the animation inside the clip is holding at the 50% mark, or the halfway point in the clip.

If we want to double the speed with no variation, we would make our last key end at 50,100.

TimeWarpGraphDoubleThe reason it ends at 50% on the X axis, is because it’s moving double the speed, it will reach the end in half the time.

If you were wanting to get fancy, and have your animation halt, ease to moving in reverse, and then suddenly going forward while easing into a halt, you’d look at something like this:


As you could imagine, this unfortunately can’t be an exact science. If you’re needing to be very precise on which frames you want to have these speed ups and slow downs happen, I would recommend making your animation clip divisible by 100 to make the math simple. Hopefully this will help you if you’re a little lost with using time warps.

So in part one we created the walk cycle and our character set. Now it’s time to bring that cycle into the trax editor!

First we want to make sure our character set is selected. You’ll want to click and hold onto that down arrow button to the right of your timeline length numbers.


The selected character is shown to the right of that, as evident through the screenshot. Something to note is that I forgot I had an object or control already named Coyote, so maya automatically added a 1 after it. It’s ok and shouldn’t affect anything that we’ll be doing. I left that mistake in there to let you know that it would still be ok if you accidentally did the same.

Now that we have our character set selected, it’s time to make that cycle actually cycle! Your trax editor probably still looks like what it did before. All we see is a soundtrack. To be able to see our new character in the trax editor we have to load it. With your character set selected (check the bottom right), press the Load Selected Character button found here:



A layer has been created inside of the trax editor, with the name of our character set. It can’t be seen yet, but this is where our cycle will be kept in. With our character set still being selected (bottom right arrow still red with the name of your character set written), go to:

Create > Animation Clip > Dialogue Box


This presents us with some new options!


The name is what we want to call the clip. In this case its a walking animation, so I just use Walking. The next important option is to choose where to put the clip. How trax editor works is by keeping the clips we create in what’s called the Visor.

Since we want to use the clip already, we can skip that extra step of dragging and dropping from the visor, and let the clip be put into the trax editor for us. The time range is how long the cycle is. If you know what your frame range is you can specify it here.


Sweet! Now we have our cycle! All the keys you have on your character should have disappeared, and we now have our clip in the trax editor! Now if you were to extend the timeline so that its longer than your clip, you should see that.. it still stops after the first one! That’s because we have to extend the clip so that it can cycle! There’s a couple different ways to do that. You can do it in the trax editor, or by selecting the clip in the trax editor, and bringing up your attribute editor (alt-a).

Let’s say we want our Coyote clip to extend 100 frames. We can do that by moving our mouse to the right hand side of the clip in the trax editor, holding shift, and then clicking and dragging. What it does is add more Post Cycle‘s, which is how you would extend it via the attribute editor.


Now the bulk of the work is out of the way. Unfortunately that doesn’t mean it’s time for the easy part. Next installment is going to be the actual use and practice of using the time warp. Things are only going to get more confusing from here on out, but bare with me and it’ll be hopefully worth it in the end! Stay tuned!

This isn’t directly dome related, but I thought I’d post about it anyways. I ran into a recent problem in Maya having to do with the trax editor. I wasn’t that well-versed in the trax editor’s inner-workings, so there was a steep learning curve for me. I wanted to write about what I learned, so that someone else who hasn’t ever used the trax editor, might get that extra step in the right direction.

Non-Linear Animation

These are the fancy words used to describe the purpose of the trax editor. It may be a little confusing at first, but it basically just means cycles.



So now that we can establish the difference between Linear and Non-Linear animation, we’ll have to think which would suit us best for what it is we want to do. In my case I was needing to have a character walk. This is perfect for cycles, so Non-Linear animation is the way to go. Trax editor here I come!

But first we need a walk cycle!

Here we go! Just threw together this walk cycle very quickly as an example of what you may have.


This was a quick walk cycle I threw together as an example of what you may have.  Coyote’s feeling pretty good about himself, and rightfully so! What we’re going to do is turn him into a:

Character Set

In Coyote’s particular case, his cycle is 31 frames long. Your cycle is most likely different from this length, so you’ll just want to remember to adjust the numbers so that they’ll fit your own animations. What we’ll want to do is open the Trax Editor. You can open it if you haven’t already by going to:

Window > Animation Editors > Trax Editor

You’ll be presented by this lovely window:


Now that we have Trax Editor open, its time to select our character so that we can make that character set! You’ll want to select everything that has a key on it. In Coyote’s case, it’s the curves that are used to control his joints.


Now that we have everything with keys selected, it’s time to turn him into our Character Set. We can do this by going to the trax editor and clicking:

Modify > Create Character Set > Dialog Box


What this does is present us with our options for our character set.


So the important thing is to give your Character Set a name. It can be whatever you’d like it too be. In this case I called it Coyote. The other options depends on what you’d want. Usually the default is the way to go. If you also keyed the scale, you’ll want to be sure to uncheck that one. I’ve left mine how you see it here. After you hit ok, you shouldn’t need to touch your controls or joints again!

Now that we have our character set created, we can move onto to making the cycle in the trax editor, which is continued in part two!

We may have briefly touched on this idea, but I wanted to address this more directly.

A lot of techniques used in flat screens naturally do not translate very well into the dome. Wipes used for a flat screen would work differently for domes, linear or radial. Things that should be kept in mind is having to account for the distortion of the projection onto the curved surface. It doesn’t mean you can’t do these sorts of transitions, you just have to approach them with a different understanding.

When knew we wanted to do a linear wipe for a particular scene in Earth, Moon, and Sun; for that show we used an animatic for the flat screen. We didn’t think it to be a problem until we started production. The wipe in question was one coming from the top of the frame to the bottom. The thing was that since you can see above you in the dome, the wipe wouldn’t really be possible, or effective. We had to scramble to figure out another way to have this linear wipe, and to make it fun and interesting at the same time.

The solution was to treat the dome as a sphere, and have our hero, Coyote, flip us onto our backs, letting the wipe come from behind us to the front.


Transitions from one shot to another when not using cuts must really be thought through very carefully for the dome. Creativity can really be challenged, and lots of new and interesting ideas can come from unlikely places.

Seams can be a problem not only in the dome master frames, as Pete illustrated in previous posts, but also in the projectors themselves. I learned from our resident dome expert, Eric Knisely, that seams can present themselves in the projectors themselves. Constant maintenance is needed to keep these projectors calibrated, and not every planetarium may be keeping up with theirs. Eric let me know how using solid colors can really make these seams stand out. He suggested that noise be added to them to break up the constant color, where even gradients may fall short. I can show in these quick illustrations below the scene that brought this problem to light, pardon the pun. The seams I put in are of my own doing, but I tried to make it reflect what we saw in the dome the best I could. The dome at RENCI uses 4 cameras, so the seam across the top was a cross shape.

The first example is what it would look like had it been with just a single projector.


The second is the same image projected using 4 projectors.


The third is with the noise applied to the sky, breaking up the values to help hide the seams.


Text treatment can be pretty specific for flat screen, and even more so for working in dome space. Its always good practice to steer clear of serifs, as they can be hard to read on the flat screen. Television and Film this is kept in mind a lot, and should also be something to think about when projecting on the dome surface. We’re currently working on the creation of the credits for our conversion show The Magic Tree House, and learned a couple things from the experience. The distance the credits or text can play a big factor in ledgibility. The closer to the camera the more distortion we see. A good rule of thumb we’ve found is that if you keep the size of the text no larger than one of the cameras (a 90 degree section of the 360 dome), there isn’t much distortion. Another big part to remember is that the resolution of the dome can vary significantly from planetarium to planetarium. Although the text might look really nice and crisp in the 4k version, those planetaria that have 1k domes may not be able to make out the text very easily.

In our conversion of The Magic Tree House, there is a sequence of shots that the visuals are being re-done. One part of that sequence is when we are on the surface of Mars following the Sojourner rover, but we ran into a hitch. There were two goals we were wanting to achieve for this section, which is to have the rover exit from the lander, and to end with an impression of the rover exploring the surface of Mars. Since the audio commentary was to remain unchanged, we were fairly constrained in what options we had to visually tell the story. To keep the number of shots to an absolute minimum so we could fit it in the already predetermined sequence length, we had to look to using some film techniques we weren’t sure would translate to a dome.

Needing to show passage of time to make sense for the following shot of the rover driving off into the martian sunset, we lowered the sun over a series of dissolves, while still keeping the same camera dolly in.  The reason we felt it’d translate well for the dome is that with the continued motion forward we can continue having parallax motion with the rocks and boulders to show distance, and the growing length of shadows combined with the sky’s hue and saturation change, can help to really create some immersion. Check out the video below:


The camera rig we’ve used for Earth, Moon, and Sun has undergone a slight change for the Magic Tree House. We use a 5 camera setup, with Sky-Skan’s DomeXF proprietary plugin for AfterEffects to stitch all the cameras together. From what you can see in the picture, we got all the cameras pointed the correct directions with appropriate Angle of View. We went ahead and included use background shaded planes for the appropriate cameras so we don’t have to render stuff we won’t see. Sometimes however we need to blur certain elements in which case we would turn off those planes so that there won’t be a feathering on the master frames. In order to manipulate these cameras to where we need them to be, without letting them drift independently, a supermover holds the group node of all 5 cameras. A nurbs arrow shows what direction front is.


Now, this camera setup is what we used for Earth, Moon, and Sun. With EMS, we’d have our starfield referenced in seperately, but this time for MTH we decided to just combine the two and make it easier on our end. One of the first things I learned about astronomy since working at Morehead is that when you move from one planet to another in our solar system, there is hardly any star movement, if any at all. Essentially the stars are locked in space, and only move when we’re rotating the camera around. Since we didn’t want our starfield to shift in space, we applied a point constraint to keep the stars in the same position, relative to where the cameras are, but would also allow us to rotate the stars to accurately reflect where we are.