diff --git a/source/nand/aes.c b/source/nand/aes.c
index cf2c1f3..6bad542 100644
--- a/source/nand/aes.c
+++ b/source/nand/aes.c
@@ -85,7 +85,7 @@ void add_ctr(void* ctr, uint32_t carry)
     uint32_t sum;
     int32_t i;
 
-    for(i = 0; i < 4; i++) {
+    for(i=0; i < 4; i++) {
         counter[i] = ((uint32_t)outctr[i*4+0]<<24) | ((uint32_t)outctr[i*4+1]<<16) | ((uint32_t)outctr[i*4+2]<<8) | ((uint32_t)outctr[i*4+3]<<0);
     }
 
@@ -110,6 +110,46 @@ void add_ctr(void* ctr, uint32_t carry)
     }
 }
 
+void ctr_decrypt_boffset(void *inbuf, void *outbuf, size_t size, size_t off, uint32_t mode, uint8_t *ctr)
+{
+    size_t bytes_left = size;
+    size_t off_fix = off % AES_BLOCK_SIZE;
+    uint8_t temp[AES_BLOCK_SIZE];
+    uint8_t ctr_local[16];
+    uint8_t *in  = inbuf;
+    uint8_t *out = outbuf;
+    uint32_t i;
+    
+    for (i=0; i<16; i++) // setup local ctr
+        ctr_local[i] = ctr[i];
+    add_ctr(ctr_local, off / AES_BLOCK_SIZE);
+    
+    if (off_fix) // handle misaligned offset (at beginning)
+    {
+        for (i=off_fix; i<AES_BLOCK_SIZE; i++)
+            temp[i] = *(in++);
+        ctr_decrypt(temp, temp, 1, mode, ctr_local);
+        for (i=off_fix; i<AES_BLOCK_SIZE; i++)
+            *(out++) = temp[i];
+        bytes_left -= AES_BLOCK_SIZE - off_fix;
+    }
+    
+    ctr_decrypt(in, out, bytes_left / AES_BLOCK_SIZE, mode, ctr_local);
+    in += AES_BLOCK_SIZE * (uint32_t) (bytes_left / AES_BLOCK_SIZE);
+    out += AES_BLOCK_SIZE * (uint32_t) (bytes_left / AES_BLOCK_SIZE);
+    bytes_left -= AES_BLOCK_SIZE * (uint32_t) (bytes_left / AES_BLOCK_SIZE);
+    
+    if (bytes_left) // handle misaligned offset (at end)
+    {
+        for (i=0; i<bytes_left; i++)
+            temp[i] = *(in++);
+        ctr_decrypt(temp, temp, 1, mode, ctr_local);
+        for (i=0; i<bytes_left; i++)
+            *(out++) = temp[i];
+        bytes_left = 0;
+    }
+}
+
 void ctr_decrypt(void *inbuf, void *outbuf, size_t size, uint32_t mode, uint8_t *ctr)
 {
     size_t blocks_left = size;
diff --git a/source/nand/aes.h b/source/nand/aes.h
index 87b8ec1..ff46726 100644
--- a/source/nand/aes.h
+++ b/source/nand/aes.h
@@ -54,6 +54,7 @@ void set_ctr(void* iv);
 void add_ctr(void* ctr, uint32_t carry);
 void aes_decrypt(void* inbuf, void* outbuf, size_t size, uint32_t mode);
 void ctr_decrypt(void* inbuf, void* outbuf, size_t size, uint32_t mode, uint8_t *ctr);
+void ctr_decrypt_boffset(void *inbuf, void *outbuf, size_t size, size_t off, uint32_t mode, uint8_t *ctr);
 void aes_cmac(void* inbuf, void* outbuf, size_t size);
 void aes_fifos(void* inbuf, void* outbuf, size_t blocks);
 void set_aeswrfifo(uint32_t value);
diff --git a/source/virtual/sddata.c b/source/virtual/sddata.c
index ebca9a1..daa7e6d 100644
--- a/source/virtual/sddata.c
+++ b/source/virtual/sddata.c
@@ -52,44 +52,6 @@ FilCryptInfo* fx_find_cryptinfo(FIL* fptr) {
     return info;
 }
 
-void fx_crypt(FilCryptInfo* info, void* buff, FSIZE_t off, UINT bt) {
-    u32 mode = AES_CNT_CTRNAND_MODE;
-    u8 ctr[16] __attribute__((aligned(32)));
-    u8 buff16[16];
-    u8* buffer = buff;
-    
-    // copy CTR and increment it
-    memcpy(ctr, info->ctr, 16);
-    add_ctr(ctr, off / 16);
-    
-    // setup the key
-    setup_aeskeyY(0x34, info->keyy);
-    use_aeskey(0x34);
-    
-    // handle misaligned offset (at beginning)
-    if (off % 16) {
-        memcpy(buff16 + (off % 16), buffer, 16 - (off % 16));
-        ctr_decrypt(buff16, buff16, 1, mode, ctr);
-        memcpy(buffer, buff16 + (off % 16), 16 - (off % 16));
-        buffer += 16 - (off % 16);
-        bt -= 16 - (off % 16);
-    }
-    
-    // de/encrypt the data
-    ctr_decrypt(buffer, buffer, bt / 16, mode, ctr);
-    buffer += 16 * (UINT) (bt / 16);
-    bt -= 16 * (UINT) (bt / 16);
-    
-    // handle misaligned offset (at end)
-    if (bt) {
-        memcpy(buff16, buffer, bt);
-        ctr_decrypt(buff16, buff16, 1, mode, ctr);
-        memcpy(buffer, buff16, bt);
-        buffer += bt;
-        bt = 0;
-    }
-}
-
 FRESULT fx_open (FIL* fp, const TCHAR* path, BYTE mode) {
     int num = alias_num(path);
     FilCryptInfo* info = fx_find_cryptinfo(fp);
@@ -122,8 +84,11 @@ FRESULT fx_read (FIL* fp, void* buff, UINT btr, UINT* br) {
     FilCryptInfo* info = fx_find_cryptinfo(fp);
     FSIZE_t off = f_tell(fp);
     FRESULT res = f_read(fp, buff, btr, br);
-    if (info && info->fptr)
-        fx_crypt(info, buff, off, btr);
+    if (info && info->fptr) {
+        setup_aeskeyY(0x34, info->keyy);
+        use_aeskey(0x34);
+        ctr_decrypt_boffset(buff, buff, btr, off, AES_CNT_CTRNAND_MODE, info->ctr);
+    }
     return res;
 }
 
@@ -132,12 +97,14 @@ FRESULT fx_write (FIL* fp, const void* buff, UINT btw, UINT* bw) {
     FSIZE_t off = f_tell(fp);
     FRESULT res = FR_OK;
     if (info && info->fptr) {
+        setup_aeskeyY(0x34, info->keyy);
+        use_aeskey(0x34);
         *bw = 0;
         for (UINT p = 0; (p < btw) && (res == FR_OK); p += SDCRYPT_BUFFER_SIZE) {
             UINT pcount = min(SDCRYPT_BUFFER_SIZE, (btw - p));
             UINT bwl = 0;
             memcpy(SDCRYPT_BUFFER, (u8*) buff + p, pcount);
-            fx_crypt(info, SDCRYPT_BUFFER, off + p, pcount);
+            ctr_decrypt_boffset(SDCRYPT_BUFFER, SDCRYPT_BUFFER, pcount, off + p, AES_CNT_CTRNAND_MODE, info->ctr);
             res = f_write(fp, (const void*) SDCRYPT_BUFFER, pcount, &bwl);
             *bw += bwl;
         }