diff --git a/papers/export-tech.md b/papers/export-tech.md
index ef569f935..77caae46f 100644
--- a/papers/export-tech.md
+++ b/papers/export-tech.md
@@ -69,6 +69,8 @@ hex | description
  .. | ...
  ef | preset delay 15
 ----|------------------------------------
+ f0 | UNUSED - placeholder used during optimization passes (3-byte nonce follows)
+ f1 | no operation
  f4 | call symbol (16-bit index follows; only used internally)
  f5 | jump to sub-block (address follows)
  f6 | go to sub-block (32-bit offset follows)
diff --git a/src/engine/cmdStream.cpp b/src/engine/cmdStream.cpp
index 6942c7bf7..f18c78646 100644
--- a/src/engine/cmdStream.cpp
+++ b/src/engine/cmdStream.cpp
@@ -115,6 +115,12 @@ bool DivCSPlayer::tick() {
         case 0xc7: case 0xc8: case 0xc9: case 0xca:
           command=next-0xb4;
           break;
+        case 0xf0: // placeholder
+          stream.readC();
+          stream.readC();
+          stream.readC();
+          break;
+        case 0xf1: // nop
         case 0xf7:
           command=stream.readC();
           break;
diff --git a/src/engine/cmdStreamOps.cpp b/src/engine/cmdStreamOps.cpp
index 98a1c482e..ac39ebc27 100644
--- a/src/engine/cmdStreamOps.cpp
+++ b/src/engine/cmdStreamOps.cpp
@@ -22,20 +22,17 @@
 
 /*
 #define WRITE_TICK(x) \
-  if (!wroteTick[x]) { \
-    wroteTick[x]=true; \
-    if (tick-lastTick[x]>255) { \
-      chanStream[x]->writeC(0xfc); \
-      chanStream[x]->writeS(tick-lastTick[x]); \
-    } else if (tick-lastTick[x]>1) { \
-      delayPopularity[tick-lastTick[x]]++; \
-      chanStream[x]->writeC(0xfd); \
-      chanStream[x]->writeC(tick-lastTick[x]); \
-    } else if (tick-lastTick[x]>0) { \
-      chanStream[x]->writeC(0xfe); \
-    } \
-    lastTick[x]=tick; \
-  }
+  if (tick-lastTick[x]>255) { \
+    chanStream[x]->writeC(0xfc); \
+    chanStream[x]->writeS(tick-lastTick[x]); \
+  } else if (tick-lastTick[x]>1) { \
+    delayPopularity[tick-lastTick[x]]++; \
+    chanStream[x]->writeC(0xfd); \
+    chanStream[x]->writeC(tick-lastTick[x]); \
+  } else if (tick-lastTick[x]>0) { \
+    chanStream[x]->writeC(0xfe); \
+  } \
+  lastTick[x]=tick; \
 */
 
 void writePackedCommandValues(SafeWriter* w, const DivCommand& c) {
@@ -205,6 +202,10 @@ void writePackedCommandValues(SafeWriter* w, const DivCommand& c) {
   }
 }
 
+void reloc(unsigned char* buf, size_t len, unsigned int sourceAddr, unsigned int destAddr) {
+  // TODO... this is important!
+}
+
 SafeWriter* DivEngine::saveCommand() {
   stop();
   repeatPattern=false;
@@ -228,7 +229,8 @@ SafeWriter* DivEngine::saveCommand() {
   
   SafeWriter* chanStream[DIV_MAX_CHANS];
   unsigned int chanStreamOff[DIV_MAX_CHANS];
-  bool wroteTick[DIV_MAX_CHANS];
+  std::vector<size_t> tickPos[DIV_MAX_CHANS];
+  int loopTick=-1;
 
   memset(cmdPopularity,0,256*sizeof(int));
   memset(delayPopularity,0,256*sizeof(int));
@@ -268,12 +270,22 @@ SafeWriter* DivEngine::saveCommand() {
 
   memset(lastTick,0,DIV_MAX_CHANS*sizeof(int));
   while (!done) {
+    for (int i=0; i<chans; i++) {
+      tickPos[i].push_back(chanStream[i]->tell());
+    }
+    if (loopTick==-1) {
+      if (loopOrder==curOrder && loopRow==curRow) {
+        if ((ticks-((tempoAccum+virtualTempoN)/virtualTempoD))<=0) {
+          logI("loop is on tick %d",tick);
+          loopTick=tick;
+        }
+      }
+    }
     if (nextTick(false,true) || !playing) {
       done=true;
       break;
     }
     // get command stream
-    memset(wroteTick,0,DIV_MAX_CHANS*sizeof(bool));
     if (curDivider!=divider) {
       curDivider=divider;
       chanStream[0]->writeC(0xfb);
@@ -306,8 +318,23 @@ SafeWriter* DivEngine::saveCommand() {
     }
     tick++;
   }
+  if (!playing || loopTick<0) {
+    for (int i=0; i<chans; i++) {
+      chanStream[i]->writeC(0xff);
+    }
+  } else {
+    for (int i=0; i<chans; i++) {
+      if ((int)tickPos[i].size()>loopTick) {
+        chanStream[i]->writeC(0xfa);
+        chanStream[i]->writeI(tickPos[i][loopTick]);
+        logD("chan %d loop addr: %x",i,tickPos[i][loopTick]);
+      } else {
+        logW("chan %d unable to find loop addr!",i);
+        chanStream[i]->writeC(0xff);
+      }
+    }
+  }
   logV("%d",tick);
-  memset(wroteTick,0,DIV_MAX_CHANS*sizeof(bool));
   cmdStreamEnabled=oldCmdStreamEnabled;
 
   int sortCand=-1;
@@ -353,7 +380,6 @@ SafeWriter* DivEngine::saveCommand() {
   }
 
   for (int i=0; i<chans; i++) {
-    chanStream[i]->writeC(0xff);
     // optimize stream
     SafeWriter* oldStream=chanStream[i];
     SafeReader* reader=oldStream->toReader();
diff --git a/src/gui/csPlayer.cpp b/src/gui/csPlayer.cpp
index d35a35816..c637fe6ed 100644
--- a/src/gui/csPlayer.cpp
+++ b/src/gui/csPlayer.cpp
@@ -94,6 +94,41 @@ String disasmCmd(unsigned char* buf, size_t bufLen, unsigned int addr) {
     case 0xec: case 0xed: case 0xee: case 0xef:
       return fmt::sprintf("qwait (%d)",(int)(buf[addr]-0xe0));
       break;
+    case 0xf0:
+      if (addr+3>=bufLen) return "???";
+      return fmt::sprintf("opt %.2x%.2x%.2x",(int)buf[addr+1],(int)buf[addr+2],(int)buf[addr+3]);
+      break;
+    case 0xf1:
+      return "nop";
+      break;
+    case 0xf4:
+      if (addr+2>=bufLen) return "???";
+      return fmt::sprintf("callsym %.4x",(int)(buf[addr+1]|(buf[addr+2]<<8)));
+      break;
+    case 0xf5:
+      if (addr+4>=bufLen) return "???";
+      return fmt::sprintf("call %.8x",(unsigned int)(buf[addr+1]|(buf[addr+2]<<8)|(buf[addr+3]<<16)|(buf[addr+4]<<24)));
+      break;
+    case 0xf6:
+      if (addr+4>=bufLen) return "???";
+      return fmt::sprintf("callb32 %.8x",(unsigned int)(buf[addr+1]|(buf[addr+2]<<8)|(buf[addr+3]<<16)|(buf[addr+4]<<24)));
+      break;
+    case 0xf7:
+      return "CMD";
+      break;
+    case 0xf8:
+      if (addr+2>=bufLen) return "???";
+      return fmt::sprintf("callb16 %.4x",(int)(buf[addr+1]|(buf[addr+2]<<8)));
+      break;
+    case 0xf9:
+      return "ret";
+      break;
+    case 0xfa:
+      return fmt::sprintf("jmp %.8x",(unsigned int)(buf[addr+1]|(buf[addr+2]<<8)|(buf[addr+3]<<16)|(buf[addr+4]<<24)));
+      break;
+    case 0xfb:
+      return fmt::sprintf("rate %.8x",(unsigned int)(buf[addr+1]|(buf[addr+2]<<8)|(buf[addr+3]<<16)|(buf[addr+4]<<24)));
+      break;
     case 0xfc:
       if (addr+2>=bufLen) return "???";
       return fmt::sprintf("waits %d",(int)(buf[addr+1]|(buf[addr+2]<<8)));