Rokkuban is a Thunder Force inspired doujin shoot'em up.
Project halted on 30th June 2008
Original post: I tried to come up with an original ship, but I want to go sure it's not just an unintentional rip-off. The first ship I designed looked WAY too much like a X-Wing from Star Wars.
What do you guys think? Is this one original enough or did I create a copy of an already existing ship without knowing?
Sketch of my ship "Lobster 6" (click to enlarge)
Last edited by FSR on Mon Jun 30, 2008 9:23 pm, edited 7 times in total.
The sprites are for a retro style game I'm currently working on.
I'm holding back with screen caps and a demo at the moment because there isn't much to see. To be precise, while I have about 10 functioning basic enemies and a boss, I still haven't completed a single stage.
The ship design is great, but the sprites have too many colours for their small size: what would be nice details at large sizes and/or with enough time to appreciate them become noise and complication when they move fast in a game.
You could simplify the colours (e.g. the gratuitous stripes), specular highlights, and shapes, or make bigger sprites (how retro is the screen size?), or both.
I agree that most sprites have too many colours at this point and I will do my best to reduce them to 12 colours for each sprite. I don’t want to use a lower number as I’m having a hard time maintaining the level of detail I want to sprites to have, and it wouldn’t look like a 16bit game anymore.
Ok, enough babbling, here are two new screenshots:
Stage 3; (Gaah, contrast is too low! This looked so much better on my other computer)
My initial idea was to make ever 6th enemy give a special score bonus (base score value * 6). However, this leads to a very slow paced gameplay that heavily puts focus on memorizing: Destroy five ‘100 points’ enemies, evade until the nexr‘1000 points’ enemy appears and destroy him for 6000 points.
And this is what I came up with today: Every 6th enemy drops a special coin. This coin is used as a score multiplicator (Score = special coins * remaining boss battle time)
I’m aware this sound as innovative as EA’s summer lineup, but I favor this score system because
- The position for the special coin isn’t preset, so reflexes and keeping an eye on the counter (counts the number of destroyed enemies) are the key to high scores.
- The scoring system is more intuitve: Collect the valuable coins, kill the boss as fast s possible. Basically what everyone thinks when playing a shmup without reading about the scoring mechanics first.
Is there an obvious flaw in my thoughts or do you agree?
You could leave the counter out, most shmups aren't very clear on what you need to make more points, and I think people dig this.
If you are coing to use dropping coins, you could go as far as making a medal-chaining alike score system, like Battle Garegga. Each coin gives you a score boost and raises the value of the next coin. If you drop one coin, its value is resetted.
Maybe you don't want to copy the score system from another game, but it's an idea.
I decided to remove the weapon energy entirely as I wasn't able to find a good balance.
name: Rokkuban
engine: Game Maker 6.1
speed: 60 FPS
players: 1
controls: can be changed in the options menu, default is:
- arrows: move
- Q: bomb
- W: fire
- E: change sub weapon
bullet count: medium (think Batsugun normal version)
weapons: 4 (Main: vulcan, Secondary: laser, side shot, arrow)
Power up the vulcan cannon 6 times by collection power ups. Secondary weapons can be powered up 2 times by collecting energy crystalls. All weapons have unlimited use.
stages: 6
bombs: yes, 3 by default
rank: yes
difficulty levels: 2 (easy, normal)
resolution: 'fake' 320 x 240 (The game uses upscaled graphics and actually runs in 640 x 480.This is required as Game Maker blurrs all games that run in 320 x 240 when using certain graphic cards.)
score system:
- special '6' emblems give a chain bonus
- special '6' coins give a bonus after completing the stage (similar to Batsugun's 'V' emblems)
Last edited by FSR on Wed Oct 10, 2007 12:20 pm, edited 1 time in total.
I've been working on a Shmup in GM myself as well. The only problem is, having a lot of bullets on the screen at once tends to create an annoying lag (I'm making a Danmaku). So I'm thinking of switching to a different engine, or even write my own.
How does it work out for you with having many instances on screen at once?
TThanks again for the comments, I really appreciate it
How does it work out for you with having many instances on screen at once?
My game runs smooth with up to 500 moving instances on screen, all backgrounds activated and music playing. I don't know if that is enough for Cave or Raizing style games though (I never got to count the number of instances on the screen in said games).
In my game I usually have 200-300 instance on screen at once. This might appear a bit high for a semi-traditional game, but I have to use sprites as backgrounds because game maker's built-in backgrounds mechanisms don't give me enough freedom.
Anyway, if ~500 objects at once are enough for the Danmaku game you're envisioning, I can send you a demo version with a benchmark stage this friday or saturday. Should my game run without lag on your computer I can try to assemble a detailed list that ontains all changes I made to speed my game up.
FSR wrote:TThanks again for the comments, I really appreciate it
How does it work out for you with having many instances on screen at once?
My game runs smooth with up to 500 moving instances on screen, all backgrounds activated and music playing. I don't know if that is enough for Cave or Raizing style games though (I never got to count the number of instances on the screen in said games).
In my game I usually have 200-300 instance on screen at once. This might appear a bit high for a semi-traditional game, but I have to use sprites as backgrounds because game maker's built-in backgrounds mechanisms don't give me enough freedom.
Anyway, if ~500 objects at once are enough for the Danmaku game you're envisioning, I can send you a demo version with a benchmark stage this friday or saturday. Should my game run without lag on your computer I can try to assemble a detailed list that ontains all changes I made to speed my game up.
Thanks for your explaination.
I think the reason why it doesn't run very smoothly when I use that many instances is because a lot of my bullets use timers for special patterns. I'm also not sure on the best way to handle collision detection (I'm using pixel perfect at the moment).
My computer is fast enough to handle even the newest games, so should the demo not run smooth, it's GM's fault.
But yeah, knowing some good ways of speeding games like these up is always useful.
Shion wrote:
I'm also not sure on the best way to handle collision detection (I'm using pixel perfect at the moment).
That's your problem right there. Pixel perfect collisions are EXTREMELY demanding on computer, no matter which language you use. Using traditonal hitboxes should be more than enough for shmup.
EDIT: another quick way to do collision detection is to use point_distance together with instance_nearest (i'm too lazy to write example) but it only works when there's no difference between colliding with one object and with multiple objects.
"A game isn't bad because you resent it. A game is bad because it's shitty."
Shion wrote:
I'm also not sure on the best way to handle collision detection (I'm using pixel perfect at the moment).
That's your problem right there. Pixel perfect collisions are EXTREMELY demanding on computer, no matter which language you use. Using traditonal hitboxes should be more than enough for shmup.
EDIT: another quick way to do collision detection is to use point_distance together with instance_nearest (i'm too lazy to write example) but it only works when there's no difference between colliding with one object and with multiple objects.
Thanks for the tip! My game now handles well up to 2000 bullets until fps drops below 60. It's still running without sound and all though.
worstplayer wrote:
EDIT: another quick way to do collision detection is to use point_distance together with instance_nearest (i'm too lazy to write example) but it only works when there's no difference between colliding with one object and with multiple objects.
That would probably be double processing each object. Instance nearest must iterate through all objects. Adding point distance would just be recalculating anyways. It would be quickest to use point_distance with a break statement to exit the loop if a collision is reached.
Some of the best shmups don't actually end in a vowel.
No, this game is not Space Invaders.
It looks like scores numbers are really low huh? This is a nice change from the typical "My hi-score is 950 trillion billion million points" score systems
Shatterhand wrote:It looks like scores numbers are really low huh? This is a nice change from the typical "My hi-score is 950 trillion billion million points" score systems
The lumps of zeroes at the beginning of the score might be excessive, chopping off one or two of them would allow a more pleasant spacing to the left and right of the score; the top of the screen is unusually crowded (but very tasteful).
Great new screens, reminds me of Thunder force series but with bigger numbers! I wish you good luck in making this game since any mine attempts fail at verts
@Ixmucane
Thanks. This has low priority now, but it’s noted. I’ll try to calculate the highest possible score and remove some zeros according the result.
@ Shion
The gameplay bothers me most at the moment, to be honest. I’m content with the ship’s weapon power, but some enemies feel too strong, others too weak. This will take some time to balance out.
@Indsutry
Indeed. The game has no tate support and is designed to be a horizontal shmup.
@Kaiser
Thunder Force was a major inspiration for this, so I guess someone noticing this is a good sign. Unlike Thunder Force, this game features bombs and no speed management (this was initally planned, but 4 buttons made things too complicated)
As for creating a shmup, I totally see where you’re coming from. It’s a tiring process and takes forever. It took me weeks to create two stages, and it really hurt when I had to delete them because they sucked beyond repair.
@battlegorge
Thanks, I’ll change the color palette when I have the whole stage done.
I was on the verge of announcing a demo last time I posted, but now I’m glad I didn’t (see my reply to Kaiser) The last thing I want is a bad demo ruining this game’s name before it’s even out.
Sorry for the compression and the video quality in general, today was the first time I worked with video editing software.
Anyway, I hope some of you enjoy this video and look forward to a demo.
Other news:
- Game has now full joystick support
- Fixed various bugs and removed some slowdown that was caused by the joystick script
- Finalized the score system
- Created about 10 new enemies and 2 new bosses
- Included a system to change the stage order
