fix(rtrvc): parameter issue

Author: fumiama

gui.py

@@ -172,6 +172,7 @@ def load(self):
                     data["sr_model"] = data["sr_type"] == "sr_model"
                     data["sr_device"] = data["sr_type"] == "sr_device"
                     data["pm"] = data["f0method"] == "pm"
+                    data["dio"] = data["f0method"] == "dio"
                     data["harvest"] = data["f0method"] == "harvest"
                     data["crepe"] = data["f0method"] == "crepe"
                     data["rmvpe"] = data["f0method"] == "rmvpe"
@@ -228,6 +229,7 @@ def load(self):
                     data["sr_model"] = data["sr_type"] == "sr_model"
                     data["sr_device"] = data["sr_type"] == "sr_device"
                     data["pm"] = data["f0method"] == "pm"
+                    data["dio"] = data["f0method"] == "dio"
                     data["harvest"] = data["f0method"] == "harvest"
                     data["crepe"] = data["f0method"] == "crepe"
                     data["rmvpe"] = data["f0method"] == "rmvpe"
@@ -405,6 +407,13 @@ def launcher(self):
                                     default=data.get("pm", False),
                                     enable_events=True,
                                 ),
+                                sg.Radio(
+                                    "dio",
+                                    "f0method",
+                                    key="dio",
+                                    default=data.get("dio", False),
+                                    enable_events=True,
+                                ),
                                 sg.Radio(
                                     "harvest",
                                     "f0method",
@@ -614,9 +623,10 @@ def event_handler(self):
                             # "use_jit": values["use_jit"],
                             "use_jit": False,
                             "use_pv": values["use_pv"],
-                            "f0method": ["pm", "harvest", "crepe", "rmvpe", "fcpe"][
+                            "f0method": ["pm", "dio", "harvest", "crepe", "rmvpe", "fcpe"][
                                 [
                                     values["pm"],
+                                    values["dio"],
                                     values["harvest"],
                                     values["crepe"],
                                     values["rmvpe"],
@@ -656,7 +666,7 @@ def event_handler(self):
                         self.rvc.set_index_rate(values["index_rate"])
                 elif event == "rms_mix_rate":
                     self.gui_config.rms_mix_rate = values["rms_mix_rate"]
-                elif event in ["pm", "harvest", "crepe", "rmvpe", "fcpe"]:
+                elif event in ["pm", "dio", "harvest", "crepe", "rmvpe", "fcpe"]:
                     self.gui_config.f0method = event
                 elif event == "I_noise_reduce":
                     self.gui_config.I_noise_reduce = values["I_noise_reduce"]
@@ -718,9 +728,10 @@ def set_values(self, values):
             self.gui_config.rms_mix_rate = values["rms_mix_rate"]
             self.gui_config.index_rate = values["index_rate"]
             self.gui_config.n_cpu = values["n_cpu"]
-            self.gui_config.f0method = ["pm", "harvest", "crepe", "rmvpe", "fcpe"][
+            self.gui_config.f0method = ["pm", "dio", "harvest", "crepe", "rmvpe", "fcpe"][
                 [
                     values["pm"],
+                    values["dio"],
                     values["harvest"],
                     values["crepe"],
                     values["rmvpe"],
@@ -1003,7 +1014,7 @@ def audio_callback(
                 sola_offset = sola_offset.item()
             else:
                 sola_offset = torch.argmax(cor_nom[0, 0] / cor_den[0, 0])
-            printt("sola_offset = %d", int(sola_offset))
+            # printt("sola_offset = %d", int(sola_offset))
             infer_wav = infer_wav[sola_offset:]
             if "privateuseone" in str(self.config.device) or not self.gui_config.use_pv:
                 infer_wav[: self.sola_buffer_frame] *= self.fade_in_window
@@ -1030,7 +1041,7 @@ def audio_callback(
             total_time = time.perf_counter() - start_time
             if flag_vc:
                 self.window["infer_time"].update(int(total_time * 1000))
-            printt("Infer time: %.2f", total_time)
+            # printt("Infer time: %.2f", total_time)
 
         def update_devices(self, hostapi_name=None):
             """获取设备列表"""

infer/lib/rtrvc.py

@@ -213,7 +213,6 @@ def infer(
                 pitch, pitchf = self._get_f0(
                     input_wav[-f0_extractor_frame:],
                     self.f0_up_key - self.formant_shift,
-                    self.n_cpu,
                     3,
                     f0method,
                 )
@@ -275,7 +274,7 @@ def _get_f0(
         self,
         x: torch.Tensor,
         f0_up_key: Union[int, float],
-        filter_radius: Union[int, float],
+        filter_radius: Optional[Union[int, float]] = None,
         method: Literal["crepe", "rmvpe", "fcpe", "pm", "harvest", "dio"] = "fcpe",
     ):
         if method not in self.f0_methods.keys():
@@ -302,14 +301,15 @@ def _get_f0_pm(self, x, f0_up_key, filter_radius):
         f0 = self.pm.compute_f0(x)
         return self._get_f0_post(f0, f0_up_key)
 
-    def _get_f0_harvest(self, x, f0_up_key, filter_radius):
+    def _get_f0_harvest(self, x, f0_up_key, filter_radius=3):
         if not hasattr(self, "harvest"):
             self.harvest = Harvest(
                 self.window,
                 self.f0_min,
                 self.f0_max,
                 self.sr,
             )
+        if filter_radius is None: filter_radius=3
         f0 = self.harvest.compute_f0(x, filter_radius=filter_radius)
         return self._get_f0_post(f0, f0_up_key)
 
@@ -344,8 +344,9 @@ def _get_f0_rmvpe(self, x, f0_up_key, filter_radius=0.03):
                 device=self.device,
                 use_jit=self.use_jit,
             )
+        if filter_radius is None: filter_radius=0.03
         return self._get_f0_post(
-            self.rmvpe.compute_f0(x, thred=filter_radius),
+            self.rmvpe.compute_f0(x, filter_radius=filter_radius),
             f0_up_key,
         )
 

rvc/layers/synthesizers.py

@@ -187,7 +187,7 @@ def infer(
             length = int(return_length.item())
             flow_head = torch.clamp(skip_head - 24, min=0)
             dec_head = head - int(flow_head.item())
-            m_p, logs_p, x_mask = self.enc_p(phone, pitch, phone_lengths, flow_head)
+            m_p, logs_p, x_mask = self.enc_p(phone, pitch, phone_lengths)
             z_p = (m_p + torch.exp(logs_p) * torch.randn_like(m_p) * 0.66666) * x_mask
             z = self.flow(z_p, x_mask, g=g, reverse=True)
             z = z[:, :, dec_head : dec_head + length]