|
| |
Released At :
X'2008
Achievements :
C64 Demo Competition at X'2008 : #3
Credits :
SIDs used in this release :
Scrolltext and other text in this release : ()
Download :
Look for downloads on external sites:
Pokefinder.org
Trivia Info
Submitted by WVL on 12 March 2010
Some trivia :
The intro
---------
The intro is a reworked version of the code in Primary Star, but keeps the checker size constant. The ideas behind the code and memory layout evolved during several years, but at the time of Primary Star I already had it running fullscreen and at 50fps. Small updates and ideas later on made it possible to add a fullscreen scroller that's using both $d800 and a fullscreen layer of sprites to give that translucent feeling. I was horrified at how much rastertime the sprites were eating in the end :P. The sprites on the screen are 40 pixels high :)
Interference part
-----------------
The interference circles in the demo consist of a moving and a stationary part. The stationary part is exactly the same size (22*22 chars) and shape as the one in DeM (I coded a tool to draw the circles exactly like Graham's :D).
The moving circles in DeM have a size of 41*39 chars, but the ones here have a size of 46*45.25 chars, a 30% bigger area!
I tried to put some totally weird movements into the fpp routine for the moving layer, but that totally destroyed the interference feeling and look. So in the end the fpp-ing was pretty limited to keep the look nice... The fpp also was my saviour, because it allowed me to mirror the gfx so I only needed halve of the gfx in the memory.
If the part is not fpp-ing, the y-movement can be done by changing only the value of y at the start of the rastercode. The same y-movement in Graham's version must have been hell-on-6502 to do with so many different combinations of multiplex stuff and badlines... brrr!
3D scroller
-----------
I made a very fast fake 3d routine to maybe be used in a game. I was thinking of something like a pacman game over several screens but in the end I made a part out of it.. The only thing that is calculated in realtime is where the shadows have to be, but not actual 3d stuff..
The routine is fast enough to draw a fullscreen maze in 50fps if the 3d calcs are precalced.
Kefrens
-------
I made this in the last weeks before X08, because I felt the demo was a bit too short.. While thinking about effects, I thought it would be cool to do something different and make horizontal kefrens bars.. At the start I aimed for 80 bars, but in the end I had to settle for 40 (but nicely executed if I say so myself ;).. You all know how I got my butt kicked in the end. The part lasts so long because it's waiting for all the zoomer sprites being calculated in the background.
All-border-zoomer
-----------------
I kinda invented the all-border sprite zoomer in our Design Overdose demo, using my own idea for X-zooming combined with Pernod's Y stretching and shrinking from Rutig Banan and putting everything into the border.. It was pretty badly executed, because it didnt move and so it was hard to see that it was using the sideborders..
I was pretty annoyed when Graham totally stole the show with the same kind of routine in Comajob. While coding this one, I found out how hard it is to keep everything looking correctly in the borders though... What I didnt like about Graham's routine was that he couldnt shrink the picture (while my old routine in DO could shrink it to 0 pixels wide and 1/3 height).
So i just had to take the crown back ;).. This routine moves even more and better than Graham's, because the picture can move out of the top of the screen (really hard to do with sprites), and it can shrink the picture down to one single pixel!! Also it can turn the picture upside down since it's all based on sprite-fpp. As a nice addition I added turning around the y-axis aswell, so you can see the backside of the picture! When zoomed out totally, the height can be > 256 pixels easily, but that depends on the music a bit.. I think in the demo it reaches about 270 pixels max.
End part
--------
The endpart evolved around a routine I made that rendered 16 colored pictures into a displayable FLI picture. The idea was to have a picture that was continually modified. In the case of the part because of exploding bubbles over a logo or picture and leaving stuff behind. Of course the bubbles had to come from somewhere, and the idea was to have them pop out of the parallaxed craters at the bottom of the screen.
It's too bad we didnt have time to finish the part, but you can still see the bubbles (with a floating embryo inside) if you play a bit with the joystick. I removed the FLI-render code in the end because the rest wasnt finished and I needed the diskspace.. Maybe I can put it to use later. |
|
|
|
 | Search CSDb |
|
| Navigate | |
|
| Detailed Info | |
|
| Fun Stuff | |
|
| Forum | |
|
| Support CSDb | |
|
 |  |
|
9.2/10 (76 votes) 