Written by Alcator
In this final chapter of the PF Fun series, I’ll explain the construction-speeding tip you were all waiting for 🙂
PF building tip 3 – Use the ghost
It is a common practice to build some obstacles in the expected path of the car, place the car using the “car” icon on the start, and test the track this way; when the stunt doesn’t quite work, trackmaker removes the obstacle, places a different one, and tries again. This may result in dozens of failing attempts and frustration. And yet, there’s a way to at least minimize the number of failed attempts…
This is the start section of a demonstrational track on which I show how to build PF quickly:
As you can see, it has the CP that I mentioned in tip #1 (in the second chapter). It is followed by 2 red boosters and a dual-height jump ramp, which, if entered in the middle, results in clockwise rotation; the red boosters have a lifting effect on the front wheels, which further enhances the rotation (upon entering the ramp tile, the left wheel is forced up and the right front wheel actually descends a bit, which adds additional rotational momentum).
Now, observe the line of finish gates in the distance – I advice placing MANY of these in a safe distance from the PF section currently being constructed, so that the car may, after testing, quickly get to one of them; on Stadium, it is sometimes possible to reach the finish gate even if you end up on the roof. The reason for this is that you NEED to make it to the finish in order to get the (new) MediaTracker GHOST.
The process is this: record a MediaTracker ghost for the part that works so far, then build some “guess”-obstacles in the space through which the ghost travels, and then play the ghost back and forth to see where it will colide; since the ghost was recorded when the obstacle was not there, it passes through them, so you can even in some cases see how the ghost would pass through several possible obstacles at different points, and see which of those passes look most promising; then, remove any useless tiles and re-record the ghost. (if during recording the stunt doesn’t work well, scratch the attempt and return to the editor). Repeat this method until you have exactly what you want.
Now, let’s go through this process in a more thorough way:
As you can see in this picture, it took me about 13.4 seconds to get to the finish (the car lands a bit roughly, but it is possible to keep it on the wheels). Once you successfully validate, you get the MT ghost. You can re-record it by using the “Record MediaTracker ghost” option (in a menu that is accessed through the “camera” icon).
After you know roughly where the car flies through in your PF section, place some marker-tiles nearby so you can measure the car position more easily. Here, I placed several track tiles in the air and some track tiles on bridge supports below them.
Then, go to the “End race” editor (by clicking the highlighted “Edit” from the previous picture) and play the replay in slow motion until the car gets to some important point*. When it does, stop the replay – the car will freeze at the spot.
*important point: This greatly depends on what you want to achieve next – whether a bounce (collision) or landing. For bounces, the more “diagonal” the car’s orientation in space is, the better, as most of the tiles then result in spinning or rotating bounce. For landing, you want the wheels that will be landing first to be almost at the same distance from the obstacle, otherwise one side of the car lands sooner and the car may flip.
In this picture, I’ve aligned the view with the top of the bridge blocks; as can be observed, the car passes above the second bridge from the left but below the first flying circle (marked with green angles)…
The speed given by the boosters makes the flight almost straight (and upwards) – which means this is a perfect spot for placing two slope tiles, one above and one below the car. The following picture shows the two sloped tiles already placed (one with a bridge support, one floating), and another floating track tile placed a bit further. By playing the replay clip in slow motion, we can stop the replay at the moment when the car is closest to this new tile (green circle). This is good, the car doesn’t TOUCH it or go through, so this can be placed here.
Sooner or later, we will want the car to land on something. This is the way of easily placing suitable landing zone:
First, we play the replay so far (without re-recording it!) in slow motion and make notes of all places where the car has rotation, speed and direction suitable for landing (as mentioned before, the landing wheels should be almost at the same distance from the imaginary surface). Use the tiles placed in the “background” to easily remember these spots.
When such position is found, place a suitable track at that position and again watch the replay in slow motion. Look at the circled wheel on these two pictures:
1. Here, the wheel is just above the track.
2. Here, the wheel is partially buried in the track.
By moving the time slider a little bit left and right, you can pinpoint the exact spot where the car “hits” the obstacle (caution: the actual collision model of the car may be slightly different than the look of the car; such as with the Coast car, where the seemingly “free” area between wheels is actually “filled” with invisible material and may collide with obstacles. Watch the way the car goes on the jump ramps at the start from a close view in slow motion and see how the car reacts to the ramps).
You can try different position of the landing track, such as moving the above shown chicane tile one square to the right (from our view), rotating an unsymmetric segment etc.
After you place the obstacle in a promising location, re-record the ghost. Don’t let mediocre results satisfy you – remember, with enough speed, there are usually several possible obstacles or landing positions, so don’t be afraid to try all of them. Also, don’t ignore locations where the car would hit the track tiles in an off-set way; as I mentioned before, off-set collisions result in rotation, and that’s often pretty spectacular without losing much speed.
Using this method, I managed to reduce my PF-construction process to about 1/3 of the original time while producing better quality results (as the endless trial-and-error method sooner or later makes you settle for some “good” PF, while there could be ‘excelent” PF).
Hope this guide was helpful.