diff --git a/buzz/transcriber/file_transcriber.py b/buzz/transcriber/file_transcriber.py
index 4c52f13a..34c4429a 100755
--- a/buzz/transcriber/file_transcriber.py
+++ b/buzz/transcriber/file_transcriber.py
@@ -80,7 +80,13 @@ class FileTranscriber(QObject):
                 si = subprocess.STARTUPINFO()
                 si.dwFlags |= subprocess.STARTF_USESHOWWINDOW
                 si.wShowWindow = subprocess.SW_HIDE
-                result = subprocess.run(cmd, capture_output=True, startupinfo=si, env=app_env)
+                result = subprocess.run(
+                    cmd,
+                    capture_output=True,
+                    startupinfo=si,
+                    env=app_env,
+                    creationflags=subprocess.CREATE_NO_WINDOW
+                )
             else:
                 result = subprocess.run(cmd, capture_output=True)
 
diff --git a/buzz/transcriber/openai_whisper_api_file_transcriber.py b/buzz/transcriber/openai_whisper_api_file_transcriber.py
index dd7f7ac5..2daeb3f8 100644
--- a/buzz/transcriber/openai_whisper_api_file_transcriber.py
+++ b/buzz/transcriber/openai_whisper_api_file_transcriber.py
@@ -56,7 +56,13 @@ class OpenAIWhisperAPIFileTranscriber(FileTranscriber):
             si = subprocess.STARTUPINFO()
             si.dwFlags |= subprocess.STARTF_USESHOWWINDOW
             si.wShowWindow = subprocess.SW_HIDE
-            result = subprocess.run(cmd, capture_output=True, startupinfo=si, env=app_env)
+            result = subprocess.run(
+                cmd,
+                capture_output=True,
+                startupinfo=si,
+                env=app_env,
+                creationflags = subprocess.CREATE_NO_WINDOW
+            )
         else:
             result = subprocess.run(cmd, capture_output=True)
 
@@ -83,7 +89,14 @@ class OpenAIWhisperAPIFileTranscriber(FileTranscriber):
             si.wShowWindow = subprocess.SW_HIDE
 
             duration_secs = float(
-                subprocess.run(cmd, capture_output=True, check=True, startupinfo=si, env=app_env).stdout.decode("utf-8")
+                subprocess.run(
+                    cmd,
+                    capture_output=True,
+                    check=True,
+                    startupinfo=si,
+                    env=app_env,
+                    creationflags=subprocess.CREATE_NO_WINDOW
+                ).stdout.decode("utf-8"),
             )
         else:
             duration_secs = float(
@@ -126,7 +139,14 @@ class OpenAIWhisperAPIFileTranscriber(FileTranscriber):
                 si = subprocess.STARTUPINFO()
                 si.dwFlags |= subprocess.STARTF_USESHOWWINDOW
                 si.wShowWindow = subprocess.SW_HIDE
-                subprocess.run(cmd, capture_output=True, check=True, startupinfo=si, env=app_env)
+                subprocess.run(
+                    cmd,
+                    capture_output=True,
+                    check=True,
+                    startupinfo=si,
+                    env=app_env,
+                    creationflags=subprocess.CREATE_NO_WINDOW
+                )
             else:
                 subprocess.run(cmd, capture_output=True, check=True)
 
diff --git a/buzz/transformers_whisper.py b/buzz/transformers_whisper.py
index 75f87284..0164fd2f 100644
--- a/buzz/transformers_whisper.py
+++ b/buzz/transformers_whisper.py
@@ -220,6 +220,8 @@ class TransformersWhisper:
             text = chunk['text']
 
             # Last segment may not have an end timestamp
+            if start is None:
+                start = 0
             if end is None:
                 end = start + 0.1
 
diff --git a/buzz/whisper_audio.py b/buzz/whisper_audio.py
index 4463e07e..ce3850b7 100644
--- a/buzz/whisper_audio.py
+++ b/buzz/whisper_audio.py
@@ -1,8 +1,11 @@
 import subprocess
 import numpy as np
 import sys
+import os
 import logging
 
+from buzz.assets import APP_BASE_DIR
+
 SAMPLE_RATE = 16000
 
 N_FFT = 400
@@ -10,6 +13,8 @@ HOP_LENGTH = 160
 CHUNK_LENGTH = 30
 N_SAMPLES = CHUNK_LENGTH * SAMPLE_RATE  # 480000 samples in a 30-second chunk
 
+app_env = os.environ.copy()
+app_env['PATH'] = os.pathsep.join([os.path.join(APP_BASE_DIR, "_internal")] + [app_env['PATH']])
 
 def load_audio(file: str, sr: int = SAMPLE_RATE):
     """
@@ -48,7 +53,13 @@ def load_audio(file: str, sr: int = SAMPLE_RATE):
         si = subprocess.STARTUPINFO()
         si.dwFlags |= subprocess.STARTF_USESHOWWINDOW
         si.wShowWindow = subprocess.SW_HIDE
-        result = subprocess.run(cmd, capture_output=True, startupinfo=si)
+        result = subprocess.run(
+            cmd,
+            capture_output=True,
+            startupinfo=si,
+            env=app_env,
+            creationflags=subprocess.CREATE_NO_WINDOW
+        )
     else:
         result = subprocess.run(cmd, capture_output=True)