MxTransitionManager Pixelation transition (#209)

* Pixelation transition and improved variable names

* More accurate pointer math for the 16-bit case
This commit is contained in:
MS 2023-10-14 15:15:51 -04:00 committed by GitHub
parent 5ddfaab9e7
commit e08717f496
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23

View file

@ -127,34 +127,34 @@ void MxTransitionManager::Transition_Dissolve()
memset(&ddsd, 0, sizeof(ddsd)); memset(&ddsd, 0, sizeof(ddsd));
ddsd.dwSize = sizeof(ddsd); ddsd.dwSize = sizeof(ddsd);
HRESULT res = m_ddSurface->Lock(NULL, &ddsd, 1, NULL); HRESULT res = m_ddSurface->Lock(NULL, &ddsd, DDLOCK_WAIT, NULL);
if (res == DDERR_SURFACELOST) { if (res == DDERR_SURFACELOST) {
m_ddSurface->Restore(); m_ddSurface->Restore();
res = m_ddSurface->Lock(NULL, &ddsd, 1, NULL); res = m_ddSurface->Lock(NULL, &ddsd, DDLOCK_WAIT, NULL);
} }
if (res == DD_OK) { if (res == DD_OK) {
SubmitCopyRect(&ddsd); SubmitCopyRect(&ddsd);
for (MxS32 i = 0; i < 640; i++) { for (MxS32 col = 0; col < 640; col++) {
// Select 16 columns on each tick // Select 16 columns on each tick
if (m_animationTimer * 16 > m_columnOrder[i]) if (m_animationTimer * 16 > m_columnOrder[col])
continue; continue;
if (m_animationTimer * 16 + 15 < m_columnOrder[i]) if (m_animationTimer * 16 + 15 < m_columnOrder[col])
continue; continue;
for (MxS32 j = 0; j < 480; j++) { for (MxS32 row = 0; row < 480; row++) {
// Shift the chosen column a different amount at each scanline. // Shift the chosen column a different amount at each scanline.
// We use the same shift for that scanline each time. // We use the same shift for that scanline each time.
// By the end, every pixel gets hit. // By the end, every pixel gets hit.
MxS32 ofs = (m_randomShift[j] + i) % 640; MxS32 x_shift = (m_randomShift[row] + col) % 640;
// Set the chosen pixel to black // Set the chosen pixel to black
if (ddsd.ddpfPixelFormat.dwRGBBitCount == 8) { if (ddsd.ddpfPixelFormat.dwRGBBitCount == 8) {
((MxU8*)ddsd.lpSurface)[j * ddsd.lPitch + ofs] = 0; ((MxU8*)ddsd.lpSurface)[row * ddsd.lPitch + x_shift] = 0;
} else { } else {
((MxU16*)ddsd.lpSurface)[j * ddsd.lPitch + ofs] = 0; ((MxU16*)ddsd.lpSurface)[row * ddsd.lPitch + x_shift] = 0;
} }
} }
} }
@ -164,7 +164,7 @@ void MxTransitionManager::Transition_Dissolve()
if (VideoManager()->GetVideoParam().flags().GetFlipSurfaces()) { if (VideoManager()->GetVideoParam().flags().GetFlipSurfaces()) {
LPDIRECTDRAWSURFACE surf = VideoManager()->GetDisplaySurface()->GetDirectDrawSurface1(); LPDIRECTDRAWSURFACE surf = VideoManager()->GetDisplaySurface()->GetDirectDrawSurface1();
surf->BltFast(NULL, NULL, m_ddSurface, &g_fullScreenRect, 0x10); surf->BltFast(NULL, NULL, m_ddSurface, &g_fullScreenRect, DDBLTFAST_WAIT);
} }
m_animationTimer++; m_animationTimer++;
@ -230,10 +230,121 @@ void MxTransitionManager::Transition_None()
EndTransition(TRUE); EndTransition(TRUE);
} }
// OFFSET: LEGO1 0x1004bed0 STUB // OFFSET: LEGO1 0x1004bed0
void MxTransitionManager::Transition_Pixelation() void MxTransitionManager::Transition_Pixelation()
{ {
// TODO if (m_animationTimer == 16) {
m_animationTimer = 0;
EndTransition(TRUE);
return;
}
if (m_animationTimer == 0) {
// Same init/shuffle steps as the dissolve transition, except that
// we are using big blocky pixels and only need 64 columns.
for (MxS32 i = 0; i < 64; i++) {
m_columnOrder[i] = i;
}
for (i = 0; i < 64; i++) {
MxS32 swap = rand() % 64;
MxU16 t = m_columnOrder[i];
m_columnOrder[i] = m_columnOrder[swap];
m_columnOrder[swap] = t;
}
// The same is true here. We only need 48 rows.
for (i = 0; i < 48; i++) {
m_randomShift[i] = rand() % 64;
}
}
// Run one tick of the animation
DDSURFACEDESC ddsd;
memset(&ddsd, 0, sizeof(ddsd));
ddsd.dwSize = sizeof(ddsd);
HRESULT res = m_ddSurface->Lock(NULL, &ddsd, 1, NULL);
if (res == DDERR_SURFACELOST) {
m_ddSurface->Restore();
res = m_ddSurface->Lock(NULL, &ddsd, 1, NULL);
}
if (res == DD_OK) {
SubmitCopyRect(&ddsd);
for (MxS32 col = 0; col < 64; col++) {
// Select 4 columns on each tick
if (m_animationTimer * 4 > m_columnOrder[col])
continue;
if (m_animationTimer * 4 + 3 < m_columnOrder[col])
continue;
for (MxS32 row = 0; row < 48; row++) {
MxS32 x_shift = 10 * ((m_randomShift[row] + col) % 64);
// To do the pixelation, we subdivide the 640x480 surface into
// 10x10 pixel blocks. At the chosen block, we sample the top-leftmost
// color and set the other 99 pixels to that value.
// Find the pixel to sample
MxS32 sample_ofs = 10 * row * ddsd.lPitch + x_shift;
MxS32 bytesPerPixel = ddsd.ddpfPixelFormat.dwRGBBitCount / 8;
// Save this cast from void* to save time.
// Seems to help accuracy doing it this way.
MxU8 *surface = (MxU8*)ddsd.lpSurface;
MxU8 *source = surface + sample_ofs * bytesPerPixel;
MxU32 sample = bytesPerPixel == 1 ? *source
: *(MxU16*)source;
for (MxS32 k = 10*row; k < 10*row + 10; k++) {
if (ddsd.ddpfPixelFormat.dwRGBBitCount == 8) {
// TODO: This block and the next don't match, but they are
// hopefully correct in principle.
MxU16 color_word = MAKEWORD(LOBYTE(sample), LOBYTE(sample));
MxU32 new_color = MAKELONG(color_word, color_word);
MxU8 *pos = surface + k * ddsd.lPitch + x_shift;
MxU32 *dest = (MxU32*)pos;
// Sets 10 pixels (10 bytes)
dest[0] = new_color;
dest[1] = new_color;
MxU16 *half = (MxU16*)(dest+2);
*half = new_color;
} else {
MxU32 new_color = MAKELONG(sample, sample);
// You might expect a cast to MxU16* instead, but lPitch is
// bytes/scanline, not pixels/scanline. Therefore, we just
// need to double the x_shift to get to the right spot.
MxU8 *pos = surface + k * ddsd.lPitch + 2*x_shift;
MxU32 *dest = (MxU32*)pos;
// Sets 10 pixels (20 bytes)
dest[0] = new_color;
dest[1] = new_color;
dest[2] = new_color;
dest[3] = new_color;
dest[4] = new_color;
}
}
}
}
SetupCopyRect(&ddsd);
m_ddSurface->Unlock(ddsd.lpSurface);
if (VideoManager()->GetVideoParam().flags().GetFlipSurfaces()) {
LPDIRECTDRAWSURFACE surf = VideoManager()->GetDisplaySurface()->GetDirectDrawSurface1();
surf->BltFast(NULL, NULL, m_ddSurface, &g_fullScreenRect, DDBLTFAST_WAIT);
}
m_animationTimer++;
}
} }
@ -263,10 +374,10 @@ void MxTransitionManager::Transition_Wipe()
memset(&ddsd, 0, sizeof(ddsd)); memset(&ddsd, 0, sizeof(ddsd));
ddsd.dwSize = sizeof(ddsd); ddsd.dwSize = sizeof(ddsd);
HRESULT res = m_ddSurface->Lock(NULL, &ddsd, 1, NULL); HRESULT res = m_ddSurface->Lock(NULL, &ddsd, DDLOCK_WAIT, NULL);
if (res == DDERR_SURFACELOST) { if (res == DDERR_SURFACELOST) {
m_ddSurface->Restore(); m_ddSurface->Restore();
res = m_ddSurface->Lock(NULL, &ddsd, 1, NULL); res = m_ddSurface->Lock(NULL, &ddsd, DDLOCK_WAIT, NULL);
} }
if (res == DD_OK) { if (res == DD_OK) {