stftpshift overhaul + fixes

fixed cli traceback + implemented formant shifting
fixed batch conversion

formantshiftcfg/f2m.txt

@@ -1,2 +1,2 @@
 8.0
--1.2
+0.8

formantshiftcfg/random.txt

@@ -1,2 +1,2 @@
-16.0
+32.0
 9.8

infer-web.py

@@ -388,6 +388,7 @@ def vc_multi(
             info, opt = vc_single(
                 sid,
                 path,
+                None,
                 f0_up_key,
                 None,
                 f0_method,
@@ -2134,8 +2135,9 @@ with gr.Blocks(theme=gr.themes.Soft(), title='Mangio-RVC-Web 💻') as app:
                         
                         qfrency = gr.Slider(
                                 value=Quefrency,
+                                info="Default value is 1.0",
                                 label="Quefrency for formant shifting",
-                                minimum=-16.0,
+                                minimum=0.0,
                                 maximum=16.0,
                                 step=0.1,
                                 visible=bool(DoFormant),
@@ -2144,8 +2146,9 @@ with gr.Blocks(theme=gr.themes.Soft(), title='Mangio-RVC-Web 💻') as app:
                             
                         tmbre = gr.Slider(
                             value=Timbre,
+                            info="Default value is 1.0",
                             label="Timbre for formant shifting",
-                            minimum=-16.0,
+                            minimum=0.0,
                             maximum=16.0,
                             step=0.1,
                             visible=bool(DoFormant),

my_utils.py

@@ -4,6 +4,7 @@ import numpy as np
 # import praatio
 # import praatio.praat_scripts
 import os
+import random
 
 import sqlite3
 
@@ -25,39 +26,61 @@ def load_audio(file, sr, DoFormant, Quefrency, Timbre):
         file_formanted = file.strip(" ").strip('"').strip("\n").strip('"').strip(" ")
         cursor.execute("SELECT Quefrency, Timbre, DoFormant FROM formant_data")
         Quefrency, Timbre, DoFormant = cursor.fetchone()
-        print(f"dofor={bool(DoFormant)} timbr={Timbre} quef={Quefrency}\n")
+        #print(f"dofor={bool(DoFormant)} timbr={Timbre} quef={Quefrency}\n")
         if bool(DoFormant):
+            numerator = round(random.uniform(1,4), 4)
             # os.system(f"stftpitchshift -i {file} -q {Quefrency} -t {Timbre} -o {file_formanted}")
             # print('stftpitchshift -i "%s" -p 1.0 --rms -w 128 -v 8 -q %s -t %s -o "%s"' % (file, Quefrency, Timbre, file_formanted))
             
             if not file.endswith(".wav"):
-                print(f"\nfile = {file}\n")
-                converting = (
-                    ffmpeg.input(file, threads = 0)
-                    .output(f"{file_formanted}.wav")
-                    .run(
-                        cmd=["ffmpeg", "-nostdin"], capture_stdout=True, capture_stderr=True
+                
+                if not os.path.isfile(f"{file_formanted}.wav"):
+                    #print(f"\nfile = {file}\n")
+                    #print(f"\nfile_formanted = {file_formanted}\n")
+                    converting = (
+                        ffmpeg.input(file_formanted, threads = 0)
+                        .output(f"{file_formanted}.wav")
+                        .run(
+                            cmd=["ffmpeg", "-nostdin"], capture_stdout=True, capture_stderr=True
+                        )
                     )
-                )
-            print("formanting...")
+                else:
+                    pass
+            
+            
+            
+            file_formanted = f"{file_formanted}.wav" if not file_formanted.endswith(".wav") else file_formanted
+            
+            
+            
+            print(f" · Formanting {file_formanted}...\n")
+            
+            
+            
             os.system(
-                'stftpitchshift.exe -i "%s" -q %s -t %s -o "%sFORMANTED"'
-                % (file_formanted, Quefrency, Timbre, file_formanted)
+                'stftpitchshift.exe -i "%s" -q "%s" -t "%s" -o "%sFORMANTED_%s.wav"'
+                % (file_formanted, Quefrency, Timbre, file_formanted, str(numerator))
             )
-            print("formanted!")
+            
+            
+            
+            print(f" · Formanted {file_formanted}!\n")
+            
+            
+            
             # filepraat = (os.path.abspath(os.getcwd()) + '\\' + file).replace('/','\\')
             # file_formantedpraat = ('"' + os.path.abspath(os.getcwd()) + '/' + 'formanted'.join(file_formanted) + '"').replace('/','\\')
+            #print("%sFORMANTED_%s.wav" % (file_formanted, str(numerator)))
             
             out, _ = (
-                ffmpeg.input("%sFORMANTED%s" % (file_formanted, ".wav"), threads=0)
+                ffmpeg.input("%sFORMANTED_%s.wav" % (file_formanted, str(numerator)), threads=0)
                 .output("-", format="f32le", acodec="pcm_f32le", ac=1, ar=sr)
                 .run(
                     cmd=["ffmpeg", "-nostdin"], capture_stdout=True, capture_stderr=True
                 )
             )
 
-            os.remove("%sFORMANTED%s" % (file_formanted, ".wav"))
-            os.remove(f"{file_formanted}.wav")
+            
             
         else:
             out, _ = (
@@ -70,5 +93,10 @@ def load_audio(file, sr, DoFormant, Quefrency, Timbre):
     except Exception as e:
         raise RuntimeError(f"Failed to load audio: {e}")
     
+    try: os.remove("%sFORMANTED_%s.wav" % (file_formanted, str(numerator)))
+    except OSError: pass; print("couldn't remove formanted type of file")
+    try: os.remove(file_formanted)
+    except OSError: pass; print("couldn't remove converted type of file")
+    
     conn.close()
     return np.frombuffer(out, np.float32).flatten()