diff --git a/data/main.lua b/data/main.lua
index 7a89220..d861a5c 100644
--- a/data/main.lua
+++ b/data/main.lua
@@ -37,10 +37,16 @@ function normal_update()
     cls(20)
     view(10,10,140,100)
     cls(3)
+    color(5)
     prnt("HOLA",0,0)
     origin(70,50)
     prnt("ORIGIN",0,0)
     prnt(text,x,10)
+    color(10)
+    bcolor(15)
+    fillp(0x5a5a, false);
+    circfill(20,20,10);
+    fillp(0xffff,false);
 end
 
 redefinekeys = {
diff --git a/lua.cpp b/lua.cpp
index c701a87..fedd061 100644
--- a/lua.cpp
+++ b/lua.cpp
@@ -53,6 +53,12 @@ extern "C" {
         return 0;
     }
 
+    static int cpp_bcolor(lua_State *L) {
+        uint8_t col = luaL_optinteger(L, 1, 6);
+        bcolor(col);
+        return 0;
+    }
+
     static int cpp_loadpal(lua_State *L) {
         const char* str = luaL_checkstring(L, 1);
         uint32_t *pal = loadpal(str);
@@ -126,6 +132,31 @@ extern "C" {
         lua_pushinteger(L, gettrans());
         return 1;
     }
+    static int cpp_subpal(lua_State *L) {
+        const int numargs = lua_gettop(L);
+        uint8_t index, index2, color;
+        switch (numargs) {
+            case 0:
+                reset_subpal();
+                break;
+            case 1:
+                index = luaL_checkinteger(L, 1);
+                subpal(index,index);
+                break;
+            case 2:
+                index = luaL_checkinteger(L, 1);
+                color = luaL_checkinteger(L, 2);
+                subpal(index, color);
+                break;
+            case 3:
+                index = luaL_checkinteger(L, 1);
+                index2 = luaL_checkinteger(L, 2);
+                color = luaL_checkinteger(L, 3);
+                for (int i=index;i<=index2;++i) subpal(i, color);
+                break;            
+        }
+        return 0;
+    }
 /*    static int cpp_pal(lua_State *L) {
         int numargs = lua_gettop(L);
         switch (numargs) {
@@ -842,6 +873,7 @@ void push_lua_funcs() {
 
     lua_pushcfunction(L,cpp_cls);       lua_setglobal(L, "cls");
     lua_pushcfunction(L,cpp_color);     lua_setglobal(L, "color");
+    lua_pushcfunction(L,cpp_bcolor);     lua_setglobal(L, "bcolor");
 
     lua_pushcfunction(L,cpp_loadpal);   lua_setglobal(L, "loadpal");
     lua_pushcfunction(L,cpp_setpal);    lua_setglobal(L, "setpal");
@@ -849,6 +881,7 @@ void push_lua_funcs() {
     lua_pushcfunction(L,cpp_getcolor);  lua_setglobal(L, "getcolor");
     lua_pushcfunction(L,cpp_settrans);  lua_setglobal(L, "settrans");
     lua_pushcfunction(L,cpp_gettrans);  lua_setglobal(L, "gettrans");
+    lua_pushcfunction(L,cpp_subpal);    lua_setglobal(L, "subpal");
 
     lua_pushcfunction(L,cpp_pset);      lua_setglobal(L, "pset");
     lua_pushcfunction(L,cpp_pget);      lua_setglobal(L, "pget");
diff --git a/mini.cpp b/mini.cpp
index 19dd471..694d173 100644
--- a/mini.cpp
+++ b/mini.cpp
@@ -48,14 +48,18 @@ bool        file_ignore_comma=true;
 #define TILES(x, y) map_surface->p[x+y*map_surface->w]
 #define CURRENT(x, y) surfaces[i].p[(x)+(y)*surfaces[i].w]
 
+void (*do_pset)(int,int);
+
 namespace ds {
     uint8_t     pen_color = 6;
+    uint8_t     back_color = 0;
     int         cam[2] = {0, 0};
     int         clip[4] = {0, 0, screen_width, screen_height};
     int         clp[4] = {0, 0, screen_width-1, screen_height-1};
     uint8_t     trans = 0;
     uint16_t    fill_pattern = 0b1111111111111111;
     bool        fill_trans = false;
+    uint8_t     draw_palette[256];
 }
 
 bool turbo_mode = true;
@@ -128,8 +132,24 @@ void read_ini() {
     fclose(f);
 }
 
+void pset_fast(int x, int y) {
+    if (ds::trans != ds::pen_color) DEST(x, y) = ds::draw_palette[ds::pen_color];
+}
+
+void pset_pattern(int x, int y) {
+    int pbx = x % 4, pby = y % 4;
+    int pb = pbx+pby*4;
+    if (ds::fill_pattern & (1 << pb)) {
+        if (ds::trans != ds::pen_color) DEST(x, y) = ds::draw_palette[ds::pen_color];
+    } else {
+        if (!ds::fill_trans) DEST(x, y) = ds::draw_palette[ds::back_color];
+    }
+}
+
 void reinit() {
+    do_pset = pset_fast;
     ds::pen_color = 6;
+    ds::back_color = 0;
     ds::cam[0] = ds::cam[1] = 0;
     ds::clip[0] = ds::clip[1] = 0; ds::clip[2] = screen_width; ds::clip[3] = screen_height;
     ds::clp[0] = ds::clp[1] = 0; ds::clp[2] = screen_width-1; ds::clp[3] = screen_height-1;
@@ -140,6 +160,7 @@ void reinit() {
         if (surfaces[i].p != NULL) free(surfaces[i].p);
         surfaces[i].p = NULL;
     }
+    for (int i=0;i<256;++i) ds::draw_palette[i]=i;
     if (file!=NULL) fclose(file);
     file = NULL;
 }
@@ -386,9 +407,10 @@ void simple_pset(int x, int y, uint8_t color) {
 }
 
 void cls(uint8_t color) {
+    const uint8_t col = ds::draw_palette[color];
     for (int y=ds::clip[1]+ds::cam[1]; y<ds::clip[3]+ds::cam[1];++y) {
         for (int x=ds::clip[0]+ds::cam[0]; x<ds::clip[2]+ds::cam[0];++x) {
-            simple_pset(x,y,color);
+            simple_pset(x,y,col);
         }
     }
     //SDL_memset(dest_surface->p, color, dest_surface->size);
@@ -398,6 +420,10 @@ void color(uint8_t color) {
     ds::pen_color=color;
 }
 
+void bcolor(uint8_t color) {
+    ds::back_color=color;
+}
+
 uint32_t *loadpal(const char* filename) {
     int size;
     uint8_t *buffer = (uint8_t*)file_getfilebuffer(filename, size);
@@ -442,6 +468,14 @@ uint8_t gettrans() {
     return ds::trans;
 }
 
+void subpal(uint8_t index, uint8_t color) {
+    ds::draw_palette[index] = color;
+}
+
+void reset_subpal() {
+    for (int i=0;i<256;++i) ds::draw_palette[i]=i;
+}
+
 /*void pal() {
     for (int i=0; i<16; ++i) {
         ds::draw_palette[i] = i;
@@ -471,21 +505,9 @@ void palt(uint8_t col, bool t) {
 void pset(int x, int y) {
     x -= ds::cam[0]; y -= ds::cam[1];
     if (x < ds::clip[0] || x >= ds::clip[2] || y < ds::clip[1] || y >= ds::clip[3]) return;
-    if (ds::trans != ds::pen_color) DEST(x, y) = ds::pen_color;
+    do_pset(x,y);
 }
 
-/*void pset(int x, int y) {
-    x -= ds::cam[0]; y -= ds::cam[1];
-    if (x < ds::clp[0] || x > ds::clp[2] || y < ds::clp[1] || y > ds::clp[3]) return;
-    int pbx = x % 4, pby = y % 4;
-    int pb = pbx+pby*4;
-    if (ds::fill_pattern & (1 << pb)) {
-        if (!ds::trans[ds::pen_color & 0xf]) DEST(x, y) = ds::draw_palette[ds::pen_color & 0xf];
-    } else {
-        if (!ds::fill_trans) DEST(x, y) = ds::draw_palette[ds::pen_color >> 4];
-    }
-}*/
-
 void pset(int x, int y, uint8_t color) {
     ds::pen_color = color;
     pset(x, y);
@@ -586,6 +608,7 @@ void rectfill(int x0, int y0, int x1, int y1, uint8_t color) {
 void fillp(uint16_t pat, bool transparent) {
     ds::fill_trans = transparent;
     ds::fill_pattern = pat;
+    do_pset=(pat==0xffff?pset_fast:pset_pattern);
 }
 
 void print_symbol(char sym, int x, int y) {
diff --git a/mini.h b/mini.h
index 0e96f09..253854f 100644
--- a/mini.h
+++ b/mini.h
@@ -126,6 +126,7 @@ void setmap(uint8_t surface);
 
 void cls(uint8_t color=0);
 void color(uint8_t color=6);
+void bcolor(uint8_t color=0);
 
 uint32_t *loadpal(const char* filename);
 void setpal(uint32_t *pal);
@@ -133,6 +134,8 @@ void setcolor(uint8_t index, uint32_t color);
 uint32_t getcolor(uint8_t index);
 void settrans(uint8_t index);
 uint8_t gettrans();
+void subpal(uint8_t index, uint8_t color);
+void reset_subpal();
 
 /*void pal();
 void pal(uint8_t c0, uint8_t c1, uint8_t p = 0);
diff --git a/version.h b/version.h
index b02db80..b490715 100644
--- a/version.h
+++ b/version.h
@@ -1,3 +1,3 @@
 #pragma once
 
-#define MINI_VERSION "0.9.73d"
+#define MINI_VERSION "0.9.78d"