diff --git a/CUETools.Codecs.FLACCL/FLACCLWriter.cs b/CUETools.Codecs.FLACCL/FLACCLWriter.cs
index 07493fd..c525c71 100644
--- a/CUETools.Codecs.FLACCL/FLACCLWriter.cs
+++ b/CUETools.Codecs.FLACCL/FLACCLWriter.cs
@@ -754,6 +754,32 @@ namespace CUETools.Codecs.FLACCL
 			frame.writer.writebits(8, crc);
 		}
 
+		unsafe int measure_residual(FlacFrame frame, FlacSubframeInfo sub, int pos, int cnt, int k)
+		{
+			int q = 0;
+			for (int i = pos; i < pos + cnt; i++)
+			{
+				int v = sub.best.residual[i];
+				v = (v << 1) ^ (v >> 31);
+				q += (v >> k);
+			}
+			return (k + 1) * cnt + q;
+		}
+
+		unsafe int measure_residual(FlacFrame frame, FlacSubframeInfo sub)
+		{
+			// partition order
+			int porder = sub.best.rc.porder;
+			int psize = frame.blocksize >> porder;
+			//assert(porder >= 0);
+			int size = 6 + (4 << porder);
+			size += measure_residual(frame, sub, sub.best.order, psize - sub.best.order, sub.best.rc.rparams[0]);
+			// residual
+			for (int p = 1; p < (1 << porder); p++)
+				size += measure_residual(frame, sub, p * psize, psize, sub.best.rc.rparams[p]);
+			return size;
+		}
+
 		unsafe void output_residual(FlacFrame frame, FlacSubframeInfo sub)
 		{
 			// rice-encoded block
@@ -806,6 +832,12 @@ namespace CUETools.Codecs.FLACCL
 			output_residual(frame, sub);
 		}
 
+		unsafe uint
+		measure_subframe_lpc(FlacFrame frame, FlacSubframeInfo sub)
+		{
+			return (uint)(sub.best.order * sub.obits + 9 + sub.best.order * sub.best.cbits + measure_residual(frame, sub));
+		}
+
 		unsafe void
 		output_subframe_lpc(FlacFrame frame, FlacSubframeInfo sub)
 		{
@@ -898,8 +930,8 @@ namespace CUETools.Codecs.FLACCL
 			task.nResidualTasks = 0;
 			task.nTasksPerWindow = Math.Min(32, eparams.orders_per_window);
 			task.nResidualTasksPerChannel = _windowcount * task.nTasksPerWindow + 1 + (eparams.do_constant ? 1 : 0) + eparams.max_fixed_order - eparams.min_fixed_order;
-			if (task.nResidualTasksPerChannel >= 4)
-				task.nResidualTasksPerChannel = (task.nResidualTasksPerChannel + 7) & ~7;
+			//if (task.nResidualTasksPerChannel >= 4)
+			//    task.nResidualTasksPerChannel = (task.nResidualTasksPerChannel + 7) & ~7;
 			task.nAutocorTasksPerChannel = _windowcount;
 			for (int iFrame = 0; iFrame < nFrames; iFrame++)
 			{
@@ -973,20 +1005,20 @@ namespace CUETools.Codecs.FLACCL
 						}
 						task.nResidualTasks++;
 					}
-					// Filler
-					while ((task.nResidualTasks % task.nResidualTasksPerChannel) != 0)
-					{
-						task.ResidualTasks[task.nResidualTasks].type = (int)SubframeType.Verbatim;
-						task.ResidualTasks[task.nResidualTasks].channel = ch;
-						task.ResidualTasks[task.nResidualTasks].obits = (int)bits_per_sample + (channels == 2 && ch == 3 ? 1 : 0);
-						task.ResidualTasks[task.nResidualTasks].abits = task.ResidualTasks[task.nResidualTasks].obits;
-						task.ResidualTasks[task.nResidualTasks].blocksize = blocksize;
-						task.ResidualTasks[task.nResidualTasks].residualOrder = 0;
-						task.ResidualTasks[task.nResidualTasks].samplesOffs = ch * FLACCLWriter.MAX_BLOCKSIZE + iFrame * blocksize;
-						task.ResidualTasks[task.nResidualTasks].residualOffs = task.ResidualTasks[task.nResidualTasks].samplesOffs;
-						task.ResidualTasks[task.nResidualTasks].shift = 0;
-						task.nResidualTasks++;
-					}
+					//// Filler
+					//while ((task.nResidualTasks % task.nResidualTasksPerChannel) != 0)
+					//{
+					//    task.ResidualTasks[task.nResidualTasks].type = (int)SubframeType.Verbatim;
+					//    task.ResidualTasks[task.nResidualTasks].channel = ch;
+					//    task.ResidualTasks[task.nResidualTasks].obits = (int)bits_per_sample + (channels == 2 && ch == 3 ? 1 : 0);
+					//    task.ResidualTasks[task.nResidualTasks].abits = task.ResidualTasks[task.nResidualTasks].obits;
+					//    task.ResidualTasks[task.nResidualTasks].blocksize = blocksize;
+					//    task.ResidualTasks[task.nResidualTasks].residualOrder = 0;
+					//    task.ResidualTasks[task.nResidualTasks].samplesOffs = ch * FLACCLWriter.MAX_BLOCKSIZE + iFrame * blocksize;
+					//    task.ResidualTasks[task.nResidualTasks].residualOffs = task.ResidualTasks[task.nResidualTasks].samplesOffs;
+					//    task.ResidualTasks[task.nResidualTasks].shift = 0;
+					//    task.nResidualTasks++;
+					//}
 				}
 			}
 			if (sizeof(FLACCLSubframeTask) * task.nResidualTasks > task.residualTasksLen)
@@ -1029,6 +1061,17 @@ namespace CUETools.Codecs.FLACCL
 							ulong csum = 0;
 							for (int i = task.frame.subframes[ch].best.order; i > 0; i--)
 								csum += (ulong)Math.Abs(coefs[i - 1]);
+
+#if DEBUG
+							// check size
+							if (_settings.GPUOnly)
+							{
+								uint real_size = measure_subframe_lpc(task.frame, task.frame.subframes[ch]);
+								if (real_size != task.frame.subframes[ch].best.size)
+									throw new Exception("size reported incorrectly");
+							}
+#endif
+
 							if ((csum << task.frame.subframes[ch].obits) >= 1UL << 32 || !_settings.GPUOnly)
 							{
 								if (!unpacked) unpack_samples(task, task.frameSize); unpacked = true;
@@ -1039,8 +1082,18 @@ namespace CUETools.Codecs.FLACCL
 								int pmin = get_max_p_order(eparams.min_partition_order, task.frame.blocksize, task.frame.subframes[ch].best.order);
 								int pmax = get_max_p_order(eparams.max_partition_order, task.frame.blocksize, task.frame.subframes[ch].best.order);
 								uint bits = (uint)(task.frame.subframes[ch].best.order * task.frame.subframes[ch].obits) + 4 + 5 + (uint)task.frame.subframes[ch].best.order * (uint)task.frame.subframes[ch].best.cbits + 6;
-								//uint oldsize = task.frame.subframes[ch].best.size;
-								task.frame.subframes[ch].best.size = bits + calc_rice_params(task.frame.subframes[ch].best.rc, pmin, pmax, task.frame.subframes[ch].best.residual, (uint)task.frame.blocksize, (uint)task.frame.subframes[ch].best.order);
+#if KLJLKJLKJL
+								uint oldsize = task.frame.subframes[ch].best.size;
+								RiceContext rc1 = task.frame.subframes[ch].best.rc;
+								task.frame.subframes[ch].best.rc = new RiceContext();
+#endif
+								task.frame.subframes[ch].best.size = bits + calc_rice_params(task.frame.subframes[ch].best.rc, pmin, pmax, task.frame.subframes[ch].best.residual, (uint)task.frame.blocksize, (uint)task.frame.subframes[ch].best.order);								
+								task.frame.subframes[ch].best.size = measure_subframe_lpc(task.frame, task.frame.subframes[ch]);
+#if KJHKJH
+								// check size
+								if (_settings.GPUOnly && oldsize > task.frame.subframes[ch].best.size)
+									throw new Exception("unoptimal size reported");
+#endif
 								//if (task.frame.subframes[ch].best.size > task.frame.subframes[ch].obits * (uint)task.frame.blocksize &&
 								//    oldsize <= task.frame.subframes[ch].obits * (uint)task.frame.blocksize)
 								//    throw new Exception("oops");
@@ -1137,19 +1190,19 @@ namespace CUETools.Codecs.FLACCL
 			task.cudaComputeLPC.SetArg(0, task.cudaResidualTasks);
 			task.cudaComputeLPC.SetArg(1, task.cudaAutocorOutput);
 			task.cudaComputeLPC.SetArg(2, task.cudaLPCData);
-			task.cudaComputeLPC.SetArg(3, (uint)task.nResidualTasksPerChannel);
+			task.cudaComputeLPC.SetArg(3, task.nResidualTasksPerChannel);
 			task.cudaComputeLPC.SetArg(4, (uint)_windowcount);
 
 			task.cudaQuantizeLPC.SetArg(0, task.cudaResidualTasks);
 			task.cudaQuantizeLPC.SetArg(1, task.cudaLPCData);
-			task.cudaQuantizeLPC.SetArg(2, (uint)task.nResidualTasksPerChannel);
+			task.cudaQuantizeLPC.SetArg(2, task.nResidualTasksPerChannel);
 			task.cudaQuantizeLPC.SetArg(3, (uint)task.nTasksPerWindow);
 			task.cudaQuantizeLPC.SetArg(4, (uint)eparams.lpc_min_precision_search);
 			task.cudaQuantizeLPC.SetArg(5, (uint)(eparams.lpc_max_precision_search - eparams.lpc_min_precision_search));
 
 			task.cudaCopyBestMethod.SetArg(0, task.cudaBestResidualTasks);
 			task.cudaCopyBestMethod.SetArg(1, task.cudaResidualTasks);
-			task.cudaCopyBestMethod.SetArg(2, (uint)task.nResidualTasksPerChannel);
+			task.cudaCopyBestMethod.SetArg(2, task.nResidualTasksPerChannel);
 
 			task.cudaCopyBestMethodStereo.SetArg(0, task.cudaBestResidualTasks);
 			task.cudaCopyBestMethodStereo.SetArg(1, task.cudaResidualTasks);
@@ -2384,8 +2437,8 @@ namespace CUETools.Codecs.FLACCL
 			cudaComputeAutocor.SetArg(1, cudaSamples);
 			cudaComputeAutocor.SetArg(2, cudaWindow);
 			cudaComputeAutocor.SetArg(3, cudaResidualTasks);
-			cudaComputeAutocor.SetArg(4, (uint)nAutocorTasksPerChannel - 1);
-			cudaComputeAutocor.SetArg(5, (uint)nResidualTasksPerChannel);
+			cudaComputeAutocor.SetArg(4, nAutocorTasksPerChannel - 1);
+			cudaComputeAutocor.SetArg(5, nResidualTasksPerChannel);
 
 			int workX = max_prediction_order / 4 + 1;
 			int workY = nAutocorTasksPerChannel * channelsCount * frameCount;
@@ -2406,7 +2459,7 @@ namespace CUETools.Codecs.FLACCL
 		{
 			cudaChooseBestMethod.SetArg(0, cudaResidualTasks);
 			cudaChooseBestMethod.SetArg(1, cudaResidualOutput);
-			cudaChooseBestMethod.SetArg(2, (uint)nResidualTasksPerChannel);
+			cudaChooseBestMethod.SetArg(2, nResidualTasksPerChannel);
 
 			openCLCQ.EnqueueNDRangeKernel(cudaChooseBestMethod, 2, null, new int[] { 32, channelsCount * frameCount }, new int[] { 32, 1 });
 		}
diff --git a/CUETools.Codecs.FLACCL/flac.cl b/CUETools.Codecs.FLACCL/flac.cl
index 5ee17e6..d8e09d7 100644
--- a/CUETools.Codecs.FLACCL/flac.cl
+++ b/CUETools.Codecs.FLACCL/flac.cl
@@ -621,18 +621,18 @@ void cudaChooseBestMethod(
 	tasks[tid + taskCount * get_group_id(1)].data.size = shared.length[tid];
 
     int l1 = shared.length[tid];
-    for (int sh = 4; sh > 0; sh --)
+    for (int l = 16; l > 0; l >>= 1)
     {
-	if (tid < (1 << sh))
+	if (tid < l)
 	{
-	    int l2 = shared.length[tid + (1 << sh)];
-	    shared.index[tid] = shared.index[tid + ((l2 < l1) << sh)];
+	    int l2 = shared.length[tid + l];
+	    shared.index[tid] = shared.index[tid + select(0, l, l2 < l1)];
 	    shared.length[tid] = l1 = min(l1, l2);
 	}
 	barrier(CLK_LOCAL_MEM_FENCE);
     }
     if (tid == 0)
-	tasks[taskCount * get_group_id(1)].data.best_index = taskCount * get_group_id(1) + shared.index[shared.length[1] < shared.length[0]];
+	tasks[taskCount * get_group_id(1)].data.best_index = taskCount * get_group_id(1) + shared.index[0];
 }
 
 __kernel __attribute__((reqd_work_group_size(64, 1, 1)))
@@ -921,13 +921,14 @@ void cudaFindPartitionOrder(
     if (get_local_id(0) < 32)
 	shared.index[get_local_id(0)] = get_local_id(0);
     barrier(CLK_LOCAL_MEM_FENCE);
+    //atom_min(shared.index[get_local_id(0)],);
     int l1 = get_local_id(0) <= max_porder ? shared.length[get_local_id(0)] : 0xfffffff;
-    for (int sh = 3; sh >= 0; sh --)
+    for (int l = 8; l > 0; l >>= 1)
     {
-	if (get_local_id(0) < (1 << sh))
+	if (get_local_id(0) < l)
 	{
-	    int l2 = shared.length[get_local_id(0) + (1 << sh)];
-	    shared.index[get_local_id(0)] = shared.index[get_local_id(0) + ((l2 < l1) << sh)];
+	    int l2 = shared.length[get_local_id(0) + l];
+	    shared.index[get_local_id(0)] = shared.index[get_local_id(0) + select(0, l, l2 < l1)];
 	    shared.length[get_local_id(0)] = l1 = min(l1, l2);
 	}
 	barrier(CLK_LOCAL_MEM_FENCE);
diff --git a/CUETools.FLACCL.cmd/Program.cs b/CUETools.FLACCL.cmd/Program.cs
index 252c930..87bdcea 100644
--- a/CUETools.FLACCL.cmd/Program.cs
+++ b/CUETools.FLACCL.cmd/Program.cs
@@ -65,6 +65,7 @@ namespace CUETools.FLACCL.cmd
 			TextWriter stdout = Console.Out;
 			Console.SetOut(Console.Error);
 
+			var settings = new FLACCLWriterSettings();
 			DateTime start = DateTime.Now;
 			TimeSpan lastPrint = TimeSpan.FromMilliseconds(0);
 			bool debug = false, quiet = false;
@@ -79,9 +80,8 @@ namespace CUETools.FLACCL.cmd
 				orders_per_window = -1,
 				blocksize = -1;
 			int level = -1, padding = -1, vbr_mode = -1;
-			bool do_md5 = true, do_seektable = true, do_verify = false, gpu_only = true;
+			bool do_seektable = true;
 			bool buffered = false;
-			int cpu_threads = 0;
 			bool ok = true;
 
 			for (int arg = 0; arg < args.Length; arg++)
@@ -93,17 +93,27 @@ namespace CUETools.FLACCL.cmd
 				else if ((args[arg] == "-q" || args[arg] == "--quiet"))
 					quiet = true;
 				else if (args[arg] == "--verify")
-					do_verify = true;
+					settings.DoVerify = true;
 				else if (args[arg] == "--no-seektable")
 					do_seektable = false;
 				else if (args[arg] == "--slow-gpu")
-					gpu_only = false;
+					settings.GPUOnly = false;
 				else if (args[arg] == "--no-md5")
-					do_md5 = false;
+					settings.DoMD5 = false;
 				else if (args[arg] == "--buffered")
 					buffered = true;
 				else if (args[arg] == "--cpu-threads")
-					ok = (++arg < args.Length) && int.TryParse(args[arg], out cpu_threads);
+				{
+					int val = settings.CPUThreads;
+					ok = (++arg < args.Length) && int.TryParse(args[arg], out val);
+					settings.CPUThreads = val;
+				}
+				else if (args[arg] == "--group-size")
+				{
+					int val = settings.GroupSize;
+					ok = (++arg < args.Length) && int.TryParse(args[arg], out val);
+					settings.GroupSize = val;
+				}
 				else if ((args[arg] == "-o" || args[arg] == "--output") && ++arg < args.Length)
 					output_file = args[arg];
 				else if ((args[arg] == "-s" || args[arg] == "--stereo") && ++arg < args.Length)
@@ -141,7 +151,7 @@ namespace CUETools.FLACCL.cmd
 				else if ((args[arg] == "-v" || args[arg] == "--vbr"))
 					ok = (++arg < args.Length) && int.TryParse(args[arg], out vbr_mode);
 				else if (args[arg] == "--orders-per-window")
-					ok = (++arg < args.Length) && int.TryParse(args[arg], out orders_per_window);					
+					ok = (++arg < args.Length) && int.TryParse(args[arg], out orders_per_window);
 				else if ((args[arg] == "-b" || args[arg] == "--blocksize") && ++arg < args.Length)
 					ok = int.TryParse(args[arg], out blocksize);
 				else if ((args[arg] == "-p" || args[arg] == "--padding") && ++arg < args.Length)
@@ -202,10 +212,7 @@ namespace CUETools.FLACCL.cmd
 
 			try
 			{
-				(encoder.Settings as FLACCLWriterSettings).GPUOnly = gpu_only;
-				(encoder.Settings as FLACCLWriterSettings).CPUThreads = cpu_threads;
-				(encoder.Settings as FLACCLWriterSettings).DoVerify = do_verify;
-				(encoder.Settings as FLACCLWriterSettings).DoMD5 = do_md5;
+				encoder.Settings = settings;
 				if (level >= 0)
 					encoder.CompressionLevel = level;
 				if (stereo_method != null)
@@ -310,15 +317,15 @@ namespace CUETools.FLACCL.cmd
 			if (debug)
 			{
 				Console.SetOut(stdout);
-				Console.Out.WriteLine("{0}\t{1}\t{2}\t{3}\t{4}..{5}\t{6}..{7}\t{8}..{9}\t{10}\t{11}",
+				Console.Out.WriteLine("{0}\t{1}\t{2}\t{3}\t{4} ({5})\t{6} ({7})\t{8}..{9}\t{10}\t{11}",
 					encoder.TotalSize,
 					encoder.UserProcessorTime.TotalSeconds > 0 ? encoder.UserProcessorTime.TotalSeconds : totalElapsed.TotalSeconds,
 					encoder.StereoMethod.ToString().PadRight(15),
 					encoder.WindowFunction.ToString().PadRight(15),
-					encoder.MinPartitionOrder,
 					encoder.MaxPartitionOrder,
-					encoder.MinLPCOrder,
+					settings.GPUOnly ? "GPU" : "CPU",
 					encoder.MaxLPCOrder,
+					encoder.OrdersPerWindow,
 					encoder.MinPrecisionSearch,
 					encoder.MaxPrecisionSearch,
 					encoder.BlockSize,