diff options
Diffstat (limited to 'home.asm')
| -rw-r--r-- | home.asm | 170 |
1 files changed, 109 insertions, 61 deletions
@@ -2269,48 +2269,55 @@ MultiplyBbyCSigned: pop af ret -MultiplyBCByEAndRoundToMostSignificantShort: ; 0x210b -; bc = round(bc * e / 256) -; b^d = sign of output +MultiplyVectorComponentByAngleFactor: ; 0x210b +; Input: bc = velocity +; e = multiplier +; d = $0 if multiplier is treated as positive, $ff if multiplier is treated as negative +; Output: bc = bc * e / 256 push af push hl ld a, b xor d ld [hSignedMathSignBuffer2], a bit 7, b - jr z, .positive1 + jr z, .positive + ; negate bc ld a, c cpl - add $1 + add 1 ld c, a ld a, b cpl - adc $0 + adc 0 ld b, a -.positive1 +.positive push bc + ; first, multiply the lo byte of the vector ld b, e call MultiplyBbyCUnsigned ld l, c ld h, b + ; round to nearest hi byte ld bc, $0080 add hl, bc ld l, h ld h, $0 pop bc + ; then, multiply the hi byte of the vector ld c, e call MultiplyBbyCUnsigned add hl, bc ld a, [hSignedMathSignBuffer2] rlca jr nc, .positive2 + ; negate hl ld a, l cpl - add $1 + add 1 ld l, a ld a, h cpl - adc $0 + adc 0 ld h, a .positive2 ld c, l @@ -2320,14 +2327,18 @@ MultiplyBCByEAndRoundToMostSignificantShort: ; 0x210b ret Cosine: ; 0x2147 - ; cos(a) + ; Input: a = angle + ; Output: e = cos(a) + ; d = 0 if cos(a) is positive, $ff if cos(a) is negative add $40 ; fall through Sine: ; 0x2149 - ; sin(a) + ; Input: a = angle + ; Output: e = sin(a) + ; d = 0 if sin(a) is positive, $ff if sin(a) is negative push hl ld [hSignedMathSignBuffer], a - and $7f ; subtract 180 degrees + and $7f ; ensure angle is between 0 and 180 degrees cp $40 jr c, .firstQuadrant ; convert angle so it's between 0 and 90 degrees @@ -2366,7 +2377,7 @@ ApplyGravityToBall: ; 0x2168 LimitBallVelocity: ; 0x2180 ; Ensures that the ball's x and y velocity are kept under a threshold. -; The ball can travel at a higher max speed when moving diagonally, since it +; The ball can travel at a greater max speed when moving diagonally, since it ; limits the x and y components independently. ld hl, wBallXVelocity + 1 call _LimitBallVelocity @@ -2376,16 +2387,16 @@ _LimitBallVelocity: ; 0x2189 ld a, [hl] bit 7, a ; is it negative velocity? (left or up) jr nz, .negativeVelocity - cp $8 + cp 8 ret c - ld a, $7 ; max positive velocity + ld a, 7 ; max positive velocity ld [hl], a ret .negativeVelocity - cp $f9 + cp -7 ret nc - ld a, $f9 ; max negative velocity + ld a, -7 ; max negative velocity ld [hl], a ret @@ -2409,24 +2420,24 @@ MoveBallPosition: ; 0x219c AddVelocityToPosition: ; 0x21c3 ld a, [de] bit 7, a - jr nz, .asm_21d1 - cp 1+4 - jr c, .asm_21da + jr nz, .negativeVelocity + cp 5 + jr c, .readLoByte ld bc, $04ff - jr .asm_21de + jr .updateBallPos -.asm_21d1 +.negativeVelocity cp -4 - jr nc, .asm_21da + jr nc, .readLoByte ld bc, -$04ff - jr .asm_21de + jr .updateBallPos -.asm_21da +.readLoByte ld b, a dec de ld a, [de] ld c, a -.asm_21de +.updateBallPos ld a, [hl] add c ld [hli], a @@ -2435,35 +2446,42 @@ AddVelocityToPosition: ; 0x21c3 ld [hl], a ret -NegateAngleAndApplyCollisionForce: ; 0x21e5 +NegateAngleAndRotateVector: ; 0x21e5 cpl inc a ; fall through -ApplyCollisionForce: ; 0x21e7 +RotateVector: ; 0x21e7 +; Rotates a vector by an angle using the standard rotation matrix calculation. Note, that +; the matrix is inverted vertically to account for negative y values mean "up" in this world. +; Input: bc = vector x component +; de = vector y component +; a = rotation angle +; Returns: bc = resulting x component = xComponent * cos(angle) + yComponent * sin(angle) +; de = resulting y component = yComponent * cos(angle) - xComponent * sin(angle) push hl - ; bc_ret = bc * cos(a) + de * sin(x) - ; de_ret = bc * cos(a) - de * sin(x) push bc push de - ld [hSineOrCosineArgumentBuffer], a + ld [hRotationAngleBuffer], a call Cosine ld a, e ld [hCosineResultBuffer], a ld a, d ld [hCosineResultBuffer + 1], a - call MultiplyBCByEAndRoundToMostSignificantShort + ; xComponent * cos(angle) + call MultiplyVectorComponentByAngleFactor ld l, c ld h, b pop bc push bc - ld a, [hSineOrCosineArgumentBuffer] + ld a, [hRotationAngleBuffer] call Sine ld a, e ld [hSineResultBuffer], a ld a, d ld [hSineResultBuffer + 1], a - call MultiplyBCByEAndRoundToMostSignificantShort - add hl, bc + ; yComponent * sin(angle) + call MultiplyVectorComponentByAngleFactor + add hl, bc ; hl = xComponent * cos(angle) + yComponent * sin(angle) pop de pop bc push hl @@ -2473,7 +2491,8 @@ ApplyCollisionForce: ; 0x21e7 ld a, [hSineResultBuffer + 1] cpl ld d, a - call MultiplyBCByEAndRoundToMostSignificantShort + ; xComponent * -sin(angle) + call MultiplyVectorComponentByAngleFactor ld l, c ld h, b pop bc @@ -2481,74 +2500,103 @@ ApplyCollisionForce: ; 0x21e7 ld e, a ld a, [hCosineResultBuffer + 1] ld d, a - call MultiplyBCByEAndRoundToMostSignificantShort - add hl, bc + ; yComponent * cos(angle) + call MultiplyVectorComponentByAngleFactor + add hl, bc ; hl = yComponent * cos(angle) - xComponent * sin(angle) ld d, h ld e, l pop bc pop hl ret -ApplyTorque: ; 0x222b +ApplyCollisionForces: ; 0x222b +; Applies the collision force to the ball's velocity and spin. +; When this function is called, the ball's velocity has already been rotated +; by the collision's normal angle, so that the velocity components are in a +; standardized coordinate system, where the normal is pointing directly upward +; at the colliding ball. If the ball is traveling away from the normal, then this +; function is a no-op. +; Input: bc = ball x velocity in rotated coordinate system +; de = ball y velocity in rotated coordinate system +; Output: bc = updated ball x velocity in rotated coordinate system +; de = updated ball y velocity in rotated coordinate system push hl - ld hl, wd7f8 + ld hl, wNoCollisionApplied ld [hl], $ff + ; Early exit if the ball is traveling away from the collision normal. bit 7, d - jr nz, .asm_2297 + jr nz, .exit ld [hl], $0 ld a, d - cp $3 - jr c, .asm_2254 + cp 3 + jr c, .applyforces + ; The ball collided hard because its y velocity is a + ; large value. Shake the rumble pack, and play a little + ; sound effect to enhance the collision. ld a, $ff ld [wRumblePattern], a - ld a, $1 + ld a, 1 ld [wRumbleDuration], a ld a, [wFlipperCollision] and a - jr nz, .asm_2254 + jr nz, .applyforces push de ld de, $0008 call PlaySFXIfNoneActive pop de -.asm_2254 +.applyforces + ; First, apply some dampening of the vertical + ; velocity component, so that walls absorb some + ; of the ball's speed. Colliding with certain objects + ; dampen the speed less that normal, like the flippers + ; and Poliwag button. + ; Divide y velocity by 4. srl d rr e srl d rr e ld h, d - ld l, e + ld l, e ; hl = yVelocity / 4 srl d - rr e - ld a, [wd7eb] + rr e ; de = yVelocity / 8 + ld a, [wSpinForceAmplification] and a - jr z, .asm_226c -.asm_2268 + jr z, .updateYVelocity +.amplify add hl, de dec a - jr nz, .asm_2268 -.asm_226c + jr nz, .amplify +.updateYVelocity + ; Negate the dampened dampened y velocity so that the ball + ; bounces off. ld d, h ld e, l ld a, e cpl - add $1 + add 1 ld e, a ld a, d cpl - adc $0 + adc 0 ld d, a ld a, [wBallSpin] + ; Divide ball spin by 2, signed, and extend to 2 bytes (hl) sra a ld l, a ld h, $0 bit 7, l - jr z, .asm_2286 + jr z, .updateXVelocity ld h, $ff -.asm_2286 +.updateXVelocity + ; hl = ballSpin / 2 + ; bc = x velocity component in rotated coordinate system + ; Add the spin to the x velocity component. add hl, bc ld b, h ld c, l push bc + ; Recalculate ball spin. + ; new ball spin = (xVelocity * 4) >> 8 sla c rl b sla c @@ -2556,7 +2604,7 @@ ApplyTorque: ; 0x222b ld a, b ld [wBallSpin], a pop bc -.asm_2297 +.exit pop hl ret @@ -2594,7 +2642,7 @@ SetBallVelocity: ; 0x22a7 pop hl ret -CheckObjectCollision: ; 0x22b5 +CheckStageCollision: ; 0x22b5 ld a, [wBallXPos + 1] sub $4 push af @@ -2801,7 +2849,7 @@ CheckObjectCollision: ; 0x22b5 jr c, .asm_23c1 .asm_23ee ld a, e - ld [wd7e9], a + ld [wIsBallColliding], a and a ret z ld a, [hLoadedROMBank] @@ -2815,7 +2863,7 @@ CheckObjectCollision: ; 0x22b5 ld hl, CollisionForceAngles add hl, de ld a, [hl] - ld [wCollisionForceAngle], a + ld [wCollisionNormalAngle], a sla e rl d ld hl, CollisionYDeltas |