Libbet the boulder girl

This is an archive of a topic from NESdev BBS, taken in mid-October 2019 before a server upgrade.
View original topic
Libbet the boulder girl
by on (#226947)
One day, Libbet was rearranging her basement when she discovered a passage to an empty hall whose floor had a peculiar pattern. She rolled in to investigate.
Re: Libbet the boulder girl
by on (#226952)
Very "Tepple like" game :lol:

Is Libbet has any meaning in English? Just curious.
Re: Libbet the boulder girl
by on (#226958)
The closest analogue is probably the Terries and Fermies from Uncle Scrooge (and as adapted to the DuckTales TV series and games). This is my design in the same vein while spending no cash.

As for the name, I guess it's one of the many, many names derived from Hebrew Elisheva.
Re: Libbet the boulder girl
by on (#226960)
nocash via PM a few days ago wrote:
Would be also interesting to see if somebody does (or doesn't) come up with completely different tile designs or the like.

Okay, that new design is pretty. I had never thought about animating the boulder, or giving it a head that can pop out when not moving.
The game rules in Magic Floor allow moving between identical fields (eg. white-to-white), but you won't get points/arrows for that moves.
Re: Libbet the boulder girl
by on (#226963)
The original Magic Floor used the obscure "1-chip cartridge" setup (where the cartridge doesn't have a CHR chip and repurposes part of the name table memory for patterns), didn't it? Does this version retain that characteristic?
Re: Libbet the boulder girl
by on (#226964)
Yes, as far as I know tepples is actually aiming at a 1-chip cartridge withou CHR chip, but not as how you might think. At the moment it seems to be only a gif/concept without any chip(s) yet. The design does neatly fit in (or extend) with what I had in mind when writing the original Magic Floor story. And, don't know if it's just a coincidence, but Libbet's white face and dark hair is roughly reminding me about the icon and figurehead that I had used in unlaunch.dsi's background graphics : )
Re: Libbet the boulder girl
by on (#226972)
Glad to see you like it.

Displaying the third arrow was a mistake, I admit. No code has been written yet.
Re: Libbet the boulder girl
by on (#226984)
"One chip" cartridge using the internal VRAM as CHR-RAM is limited to 64 total tiles, which can freely be updated during vblank time just like any other CHR-RAM game. You could have a fully animated main character just like Battletoads or Nightmare on Elm Street.
Re: Libbet the boulder girl
by on (#227109)
I ran into a problem with animating the hop move (A+Up, A+Down, A+Left, A+Right).

Attachment:
spritesheet-side.png
spritesheet-side.png [ 909 Bytes | Viewed 8132 times ]


I had been trying to draw her from the side, as many NES "overhead" RPGs did. The roll moves were fine, as was hopping over a tile sideways. But it didn't look right when I applied the same approach to hop up or down.

Attachment:
hop_test.gif
hop_test.gif [ 9.34 KiB | Viewed 8132 times ]


I think traveling up and down exposes the incongruity between side view and overhead view. So I'll have to throw her design into Blender in order to experiment with camera angles. It reminds me of the ongoing stop-motion discussion.
Re: Libbet the boulder girl
by on (#227111)
Given that the heroine will have to do a lot of moving from floor tile to floor tile, how about cutting her some slack and making her armless instead of legless this time around?
Re: Libbet the boulder girl
by on (#227115)
Looks like she's sniffing her own butthole.
Re: Libbet the boulder girl
by on (#227119)
@Bananmos
She'll be moving mostly by rolling with the occasional hop. (Another comparison is Geodude from Pokémon.) Is there a reference for an armless biped-type boulder creature that could fold up for something like that?

But with the help of some acquaintances on Discord, I figured out that part of the problem was lack of shadow.
Re: Libbet the boulder girl
by on (#227143)
Pretty! Looks good even without the shadow. And the shadow is making it even better. I like the animations for left/right and down.

The animation for up doesn't look so well. I think the arms are a bit too long and too vertical. Maybe you could spread the arms a bit to the sides in the middle of the move, and swing the arms forward at the end of the move?

Btw. did you consider continous moves in same (or different) directions? The animations aren't showing that, but it looks as if you have already designed it so that Libbet could keep rolling continously, and that she'd lift her head only if the player stops pushing dpad buttons, right?

Most versions of Magic Floor are using 24x24pix floor cells (for NES it's 32x32pix because NES BG color attributes don't allow 24x24pix). So far I've been cheaply using a single 8x8pix sprite as boulder. But now I am wondering if 10x10pix wouldn't look better on 24x24pix cells (and for NES, maybe 12x12pix on 32x32pix cells).

Or in other words: The boulder could be as big as possible (if possible leaving a small gap between boulder and arrows). Although, to be geometrically correct, the perimeter of the boulder should be roughly same as the width/height of the floor cells (so it takes one rotation per floor cell, unless you would want Libbet to end up with the nose on the ground after 1.25 rotations).
Re: Libbet the boulder girl
by on (#227144)
Shadow should get smaller the higher she goes.
Re: Libbet the boulder girl
by on (#227148)
nocash wrote:
it looks as if you have already designed it so that Libbet could keep rolling continously, and that she'd lift her head only if the player stops pushing dpad buttons

That's correct. A buffered move will execute without delay.

nocash wrote:
(for NES it's 32x32pix because NES BG color attributes don't allow 24x24pix)

Yet the ZX81 tileset works without need for attributes.

nocash wrote:
Although, to be geometrically correct, the perimeter of the boulder should be roughly same as the width/height of the floor cells (so it takes one rotation per floor cell, unless you would want Libbet to end up with the nose on the ground after 1.25 rotations).

It's an animation cheat, I admit. Games are full of them. Look at how small the houses are in outdoor scenes in many RPGs.
Re: Libbet the boulder girl
by on (#227157)
pubby wrote:
Shadow should get smaller the higher she goes.

Whatever looks best. Technically it could get bigger, or smaller, or stay same size - depending on the light source.
I guess bigger shadow wouldn't look so well, same size looks good, and, yes, maybe smaller size could look even better.

tepples wrote:
That's correct. A buffered move will execute without delay.

Okay, just wanted to be sure that you've kept it in mind.

tepples wrote:
Yet the ZX81 tileset works without need for attributes.

Yes, but that's using dithering instead of colors. The NES version does actually have some source code option for 24x24pix cells with dithering, too. The dither in NES version is rather imperfectly implemented though.
On the other hand, dither+palette might help to get better contrast on NES. Or, using different colors (like replacing four green-shades by black, dark-green, yellow-green, yellow) might also help on better contrast (on color TV at least).
Anyways, 32x32pix cells aren't so bad for NES. With smaller sizes you'd get even bigger floors with more moves (making the game more difficult to solve).

tepples wrote:
It's an animation cheat, I admit. Games are full of them. Look at how small the houses are in outdoor scenes in many RPGs.

No problem, I just meant that it might become a problem if the dimensions are totally wrong (and even then it might be fine if it doesn't look too obvious).
Re: Libbet the boulder girl
by on (#234839)
I started with a Scale2x of the existing sprite (using this GIMP plug-in) and added a bit of depth to produce a double-size rolling animation. Anything I could improve?
Re: Libbet the boulder girl
by on (#234849)
I didn't read the whole thread so i might've missed something, but how about giving her a bit of directional shade? I think that could help balance the impression, what with all the intensive changes of rolling, and reinforce her sphere-iness.
Re: Libbet the boulder girl
by on (#234850)
tepples wrote:
It's an animation cheat, I admit. Games are full of them.

Some cheats are easy to get away with. Others... are horrendous. See below for a gruesome mistake i made at one point:

Attachment:
disturbingknees.gif
disturbingknees.gif [ 13.89 KiB | Viewed 4823 times ]


In roller derby, it is often beneficial to step-roll 360 degress around your own axis to escape a troublesome blocker of the opposing team. If you're doing it successfully, each leg shouldn't rotate around its own axis or swap hip joints :shock:
Re: Libbet the boulder girl
by on (#234860)
Disney animators routinely make the same "gruesome mistake" with Mickey Mouse's ears that you made with your skater's legs. But perhaps it appears less gruesome because Mickey is less often seen rotating.

FrankenGraphics wrote:
how about giving her a bit of directional shade?

I didn't try directional shade earlier because I feared loss of ability to reuse vertically flipped tiles. This doesn't hurt the jumping cels as much as the rolling cels. But now that I look at the CHR set that my sprite sheet tool produces, only six in-game tiles are reused vertically flipped: four in side view rolling and two in front/back view rolling. So that's a good idea.

Attachment:
File comment: Sprite sheet
Libbet.png
Libbet.png [ 952 Bytes | Viewed 4801 times ]

The horizontal line at far right is the shadow that shows during a jump. The squarish looking thing is used for setting up impossible triangle priority on the animation for falling into a hole. The priority trick works the same on Game Boy as on NES, with the caveat that when running monochrome software, the PPU reorders sprites with smaller X coordinates to the front.
Re: Libbet the boulder girl
by on (#235818)
In this post, darryl.revok wrote:
tepples wrote:
There are practical problems with asymmetric highlighting on a sprite that's going to be flipped.

I've [...] decided at the end of the day, nobody but me is ever going to care that the highlight might reverse sides. It's like when a character faces the other way, their handedness changes.

That's fine for a character who doesn't often turn upside down, such as Milca from my aborted attempt at a Cookie Clicker port. Libbet is a little bit different, as she sticks her head down and rolls. If you take art shaded to imply a light source in one direction and try to rotate it, she'll look like a cardboard cutout.
Re: Libbet the boulder girl
by on (#235822)
nocash wrote:
Although, to be geometrically correct, the perimeter of the boulder should be roughly same as the width/height of the floor cells (so it takes one rotation per floor cell, unless you would want Libbet to end up with the nose on the ground after 1.25 rotations).

I remember people making jokes about π=2 for Stephen's Sausage Roll, and I felt like this impossible physics was an incredibly good feature for a game to have.

FrankenGraphics wrote:
If you're doing it successfully, each leg shouldn't rotate around its own axis or swap hip joints :shock:

This too I think is great and probably better than a realistic turn. :)