INPUT magazine

Previous topic - Next topic

erico

I kind of didn´t understand, you are talking about doing an image per room on the classic adventure style (like those on coco), right?
Yep 20+ rooms can be a burden, but could be solved by working with photographs and then process and draw with it to get there.

I once (probably said this before on another post) thought about making a hired guns mechanics style game on a 3d world, all made of blocks with Z level.
It would have a bladerunner theme.

I have  also thought of the same mechanics towards an ´alien 2´style, ye colony on alien planet, where there would be a lot of interaction with the world.

This last example is something I think we could work out. What do you think?

matchy

Digitizing could work if need be. I'm open to whatever and am only motivated if you supply material, as a non-designer.

Please, post some Youtube samples of the games you mentioned!

erico

Ok, good to know, then we should do the alien colony thing (in and out doors).
The best example of this is definently hired guns, here a video, skip to 1:30 to see gameplay, check that it has z levels, which I find super.


As of visual style, I´d have to think.
If I mimick the coco visuals (be any mode) then things are sure faster to pull out. Low res and color count also helps too.
Gimme a while here since I think I did some visual work for this idea at some point and that could be usefull speedwise.

Both coco adventures I mentioned, Major Istar and Escape 2020 have this visuals, check:


erico

#123
Just an extra thing about visuals, I did create a set of brushes for photoshop in order to produce image that corresponds visually to what one would expect from a coco hi res mode. Of course, it cheats a little as it dosen´t transform a black and white image into a color one when ntsc by the way of pixel position.

I simply extracted all visual patterns that produce any kind of color and transformed then into some kind of brush.
The result is what you see on my game thread here:
http://www.glbasic.com/forum/index.php?topic=9354.135

Which is a conversion from my current game image.

Now, I do have the patterns spread into a possible palette match according to approximate colors those patters generate (by blur), and from there I can have some color ramps. Here I append the image with the original pattern and the blurred equivalent all already on a possible ramp. For real use, I´d skip some of them, like the brick styled ones.

Still, considering those patterns as colors, one can easily convert real images to what would be, somehow, more or less, close to what to expect on the real hardware. I have been really happy using those to convert some images and it would be funky to try the assets for this possible adventure engine using those. Here a png, so you can use it too for whatever you need.

edit: for a conversion test, force downgrade your original image to an indexed palette that corresponds to the lower full colors, then, get back to rgb space and use a color picker to exchange each color for the appropriate dithered ones. This could be scripted, or actioned, but a manual work helps when some of the patterns don´t quite work. ;)

edit2: yes, there is green on the coco land, just rarely used. :D

matchy

Dungeons of Daggorath is another perfect example where there's no need to worry about the colors or background artwork. How important is it that there is?

erico

#125
Quote from: matchy on 2015-Oct-23
...I'm open to whatever and am only motivated if you supply material, as a non-designer.
...
Fine, but there may always be a a "design" to actually create something, at least an idea or general direction.
In which case, it is easier for me to delve into things I already tried and plan to make at some point with same art or modified.

Quote from: matchy on 2015-Oct-24
Dungeons of Daggorath is another perfect example where there's no need to worry about the colors or background artwork. How important is it that there is?

Sure is, but I´m also sure you got the point, as for the question with infinite answers, one may be the motivation to drive such work.

Here is what I have so far:

Looking at a dungeon crawler solution, 3d maps (with Z level), just like the mechanics on ´Hired Guns´. Possibly 4 players.

I did a bunch of research on the many crawlers while looking for something that would attend in and out doors.
Its dominant perspective seems to be the 'Dungeon Master' way, ie you are kind of looking down, but here I prefer the style used on ´Perihilion´ or 'Ishar III' as appended. With the former, it all takes a more vertical shifted camera looks and seems to be a stronger on drama. It could even accommodate 6 players. :)

Here the most probable perspective to be used so far.

erico

#126
I have been giving some more extensive thoughts on this.

Certainly, like Matchy once said, the choice of 2d/3d drawing is a strong issue.
It seems to me currently, that drawing in 3d is the obvious choice.

But for exercise, I will try pull it out 2d (probably drawed on a 3d package to try speed things up and have the power of recycling).
My current worry is to convert a 3d array into the visual display, the composition.
I have done some stuff in this front with 2d arrays, 3d is new and should present a harder challenge.
I´d probably finish the 2d array game first before seriously delving into this, but here is some pre-conclusions.

It is a lot of (huge) work to make things happen in 2d considering the results on the mentioned games before, and it gets even worse as you add details.
So the first thing would be to have something simple kicking.

The simplest I have come to think so far is on the lines of Dagorath (wireframe kind of), keep photography off unless it is just plain painting like those coco II examples or some shading dictated by depth (fog).
This way, you could even dump a variation of full screens (like pipes) and work your levels within the pipes you have. It seems to be the easiest/fastest way.
If one goes the way of tiling the map, then you can recycle perpendicular tiles according to Z level and probably flip others.
Of course you can have it all as cubes, but it sure won´t impress the girls.

For this purpose (not the girls but testing) I decided to use a filled distorted wireframes. It will kill my ability to recycle some of the assets but it is still flippable none the less. The distortion is also unique on this genre and I hope it brings a claustrophobic sensation to the players (ye, even though you can see more this way).

One great problem you face when you choose this way, is about the FOV.
If you check many references, you will come to the conclusion that the smaller the FOV the easier to pull, but less the immersion.
Some of the very classic games have extreme small FOV.

Considering this and also considering a now center perspective (like hired guns, I changed my mind), I´m currently worrying about how much is shown and how it may fit the perspective itself. This is really hard to pull out. :(

So far, I believe that, considering you on the center block, 2 blocks each side, 2 blocks up and down, and 4 blocks as depth.
If the final output resolution is 340x200 (don´t ask about where this proportion came from...) then the depth is fine, if I go lower, I can safely reduce the depth to 3 blocks. It all may sound a little low res, but I would like 4 simultaneous players on the test, and between this res and a quarter of it, it makes a lot easier to pull out art for according to the pipeline I´m establishing. 

So ye, a 5x5x5 FOV 2d wire tiles should come out of this, 4 players playable in a simple test code.
Here a set of attachments of the many tests to figure things out, these are the better ones. :-[

erico

#127
Ok, so here is the target plan grid so far for the proposal.

Within this, the amount of sprites to compose the scene, considering flipping X and Y, is about the central blocks towards the up left.
That will be 45 sprite slots straight, 37 as I take off the ones not seen by been close to camera.
And this is all considering a single tillable block, the cube, and that I´m just filling the array´s content with either full or empty.

So, so far, each x/y symmetrical tile requires 37 images so to play a part. :O
The cube is going to be the shape.

Depending on the game´s resolution and other stuff like FOG, etc , I can reduce the amount of sprite slots (like last post, the FOV).
Now I understand why Hired Guns had to have such a blocky kind of textures.

Let´s test it out.
First I have to find a way to automate the slot´s sprite creation in a way that it will spit images that are ready usable or at least with the least trouble to be used by the code.  :sick:


erico

#128
Got the tile spitter kind of working.
I tuned the FOV, so 36 tiles are needed to cover, the one you stand does not count.

I´m rendering each tile as a fullscreen+alpha for the sake of quick testing should anything perspective wise change.
This way I don´t need to compose the tile on screen, it is just a matter of overlaying the image with the desired tile at 0x0 coords.
For the sake of this method, I will keep the resolution kind of low (and I like it  :P) so 170x100.

The way of assembling the final image is in accordance to the sprite generated, back to front, top to down, left to right for each quarter of the screen.

Now I guess this needs life by a bit of coding.
I will try a 4 players split screen thing so we can at least walk around and see the world.
Anyone willing to give this a go may go ahead and do anyway you please.
Should you want to take a look at the 3d files, you are welcome too.

Here the images, use BLK001.png for the tiles. Both guides should give a good visual idea.

MrPlow

Cool - advanced stuff here :) v nice!
Comp:
Speccy-48k, Speccy-128k, Amigas, PCs

erico

#130
Yeah! After bumping my head on the wall a few times cause of the 3d array and how to rotate the fov within it...I finally made it.
It works fine, 1 player for now, I even added a dithered fog to it so it dosen´t look too bland.
Will try do the other players and a shooting thing.

Thanks Plow!
I´m sure I´m doing the array the worst way possible, will look into it at some point.

edit:a better image.

Ian Price

Looking awesome there erico. Nice one  :good:
I came. I saw. I played.

erico

Thanks, it is just a tech demo more or less, exercise.
I do have some plans for a couple games of these kind and using 3d to draw screen, even with the same low 3d style and movement, is way more practical I guess.
I will wrap it up and post the code as soon as I get other players dancing and some shooting.

erico

#133
oh heck, I guess I won´t be able to kick it up to an at least playable thing. :(
But I don´t want the idea lost and won´t post it here as it is off topic, will do a proper snippet post.

Like said before, I do have some ideas for crawlers to boot.:)
But it will have to wait the one games I plan to release.

Will edit with the snippet post soon...

here the link, have fun!
http://www.glbasic.com/forum/index.php?topic=10556.msg93207#msg93207

erico

Here a neat ludum crawler:
http://ludumdare.com/compo/ludum-dare-34/?action=preview&uid=13139

and here an article about dungeon master, I haven´t read yet but plan to.
http://www.filfre.net/2015/12/dungeon-master-part-1-the-making-of/