|
Post by binarystar on May 25, 2014 13:37:20 GMT -5
Tony, That effect is called Parallax Scrolling, a great effect in 2d games but it does not work for proper 3d movement. Gliptitude, can you explain what the 'twists and turns' are?
|
|
|
Post by gliptitude on May 29, 2014 23:01:14 GMT -5
.. Sorry for delayed response. By twists and turns i mean that the view through the stars is moving in multiple directions, changing directions and changing orientation, like in Star Wars. .. This as opposed to continuously moving straight forward with the same orientation.
.. In Star Fire Spirits I actually agree with Tony's description and it does resemble Parallax Scrolling. It is only a moderately convincing effect in this game, however it is a bit impressive that the player can actually control directional movement, unlike Star Wars (the starfield movements in Star Wars are basically scripted).
.. By "orientation" now i mean basically which way is up in relation to the "background". If you imagine the "camera" inside a ship and the ship flipping upside down while maintaining the exact same trajectory through space, this would be a change in orientation, (but not a a change in directional movement).
|
|
|
Post by binarystar on May 30, 2014 7:41:00 GMT -5
Thanks for the reply. For this you would need to rotate the viewing angle on the Y axis. My demo uses a fixed viewing angle (straight ahead) but you can move easily move left/right/up/down/forwards/backwards within the 3d starfield. I think it'd need a bit of a redesign to support changing the view angle.
|
|
|
Post by mikiex on May 31, 2014 3:47:17 GMT -5
I tried it the other week, sorry not had a chance to reply. It looks like a good effect, but its not correct - not that it matters if the user does not notice, it is after all an effect and has to run fast if you want to do a game as well. What I mean by not correct is that the dots start off fast and slow as they go towards the edge of the screen. Anyway I like it as it is
|
|
|
Post by binarystar on May 31, 2014 6:15:22 GMT -5
I am actually speeding them up but the scaling slows them down again it could use some fine-tuning but it was just a lunch hour proof-of concept.
|
|
|
Post by gliptitude on May 8, 2020 11:53:19 GMT -5
I knocked up a quick test 3d starfield, you can check out the ROM below. OK so far, I'll post (messy) source once I'm happier with it. Each 'star' (64 of them) uses 6 bytes (x,y,z,speed) so its not the most efficient implementation, but it works. <edit> Now attached improved starfield ROM. Analog stick allows you to move the starfield forwards, back, left and right. I don't think I noticed the joystick movement feature back then because I was amazed when I saw it in action during one of Malban's Vide tutorial videos the other day. I feel like an idiot for mistaking it for a different starfield demo. Sorry for not fully recognizing this awesome demo back then. Considering it is a "lunch hour" effort, I wonder what you could do if revisited, or if someone else picks it up.
|
|
|
Post by gtoal on Jun 29, 2020 3:14:25 GMT -5
I hacked up a quick 3D starfield in C: gtoal.com/vectrex/starfield.c.htmlcursor keys to rotate the ship around x,y axes; buttons 3 & 4 to rotate around z axis. Buttons 1 and 2 go forward or backwards (ie tailgunner mode). It's not perfect but I'm sure could be improved with a little effort - some of the movements don't always work too well in conjunction with some of the others (although most do compose OK). I originally tried writing a 3d starfield that followed proper rules rather than a 'looks about right' hack, but the maths were just way too expensive to calculate. supports 32 stars at reasonable speed, fewer if you need to do anything else. Haven't tried compiling with any -O options yet. If I managed to upload successfully just now, bin should be in gtoal.com/vectrex/starfield.bin
|
|
|
Post by gliptitude on Jul 2, 2020 16:12:34 GMT -5
I tried it out on Vectrex. It works pretty good. Twisting and thrusting and panning at the same time is fun.
Thanks for making and sharing this.
|
|
|
Post by Peer on Jul 3, 2020 15:18:56 GMT -5
Just a quick question, the first comment in the source code reads:
// does not work with any -O options. Immediately obvious when using button 1
Is that so? I could not spot any visual difference with -O0, -O1, -O2 and -O3. The sizes of the binaries differ, but on the screen, everything seems to behave identically.
Cheers, Peer
|
|