running issues

[wassfila]

• all repos not on latest : fixed

Top repo

• [x] add tested csound version
• [x] add warning about long install time

Production

• [x] readme path rg_production\venv\Scripts\python rg_production\main.py

• [x] missing matplotlib dependency

• [x] c-sound version missing , downloaded version in path csound-windows_x86_64-6.16.2

• errors screen freeze and Error in setup

  17: dac17 (Kopfh├Ârer (@System32\drivers\bthhfenum.sys,#2;%1 Hands-Free A)
  PortAudio: selected output device 'Microsoft Sound Mapper - Output'
  writing 2048 sample blks of 64-bit floats to dac0
  SECTION 1:
  Error in setup
  Error in setup
  Done loading stt
  Setting up stt
  Done loading sg
  Error in setup
  Error in setup
  Error in setup

Text to Sound

• [x] add missing dependency in prod requirements asyncio

• [x] fix command line call of rg_text_to_sound to be similar to other modules calls and remove bash and wsl dependency

• [x] fix module paths errors either with sys.path.append or other

• [x] reference the local path of -r prod.txt instead of git+https://git@github.com/TheSoundOfAIOSR/rg_text_to_sound.git#'subdirectory=tts_pipeline see this section https://github.com/TheSoundOfAIOSR/project_common/blob/9758ba352c3ea5004465de393207b22e87b22fc7/setup3.8.bat#L30

• bash command not compatible with windows script source venv/bin/activate; cd tts_websocketserver/src; python -m tts_websocketserver.tts_server as directory venv/bin does not exist

• replaced with command start ../../venv/Scripts/python -m tts_websocketserver/tts_server

• run as "../../venv/Scripts/python" -m tts_websocketserver/tts_server

• No module named tts_websocketserver/tts_server

• run with "venv/Scripts/python" tts_websocketserver\src\tts_websocketserver\tts_server.py

• error : ModuleNotFoundError: No module named 'tts_websocketserver.tts_pipeline'

• activated venv

• error : ModuleNotFoundError: No module named 'tts_pipeline'

• added import sys sys.path.append('tts_websocketserver')

• error ModuleNotFoundError: No module named 'tts_pipeline.pipelines'; 'tts_pipeline' is not a package

[wassfila]

@atjessehill for info, I realised you're not in the RC's group so I invited you. The matoplotlib dependency is easy to add to the requirement.txt, I have it local, did not want to push on your repo maybe just let you do it. The freeze and the error, I'm not sure if it's csound version related or if you've seen that already ?

[wassfila]

after installing asyncio manually into the bash venv and run.bat again

D:\Dev\SoundHacking\project_common\rg_text_to_sound>bash -c "source venv/bin/activate; cd tts_websocketserver/src; python3 -m tts_websocketserver.tts_server"
Traceback (most recent call last):
  File "/usr/lib/python3.8/runpy.py", line 194, in _run_module_as_main
    return _run_code(code, main_globals, None,
  File "/usr/lib/python3.8/runpy.py", line 87, in _run_code
    exec(code, run_globals)
  File "/mnt/d/Dev/SoundHacking/project_common/rg_text_to_sound/tts_websocketserver/src/tts_websocketserver/tts_server.py", line 1, in <module>
    from tts_websocketserver.tts_pipeline import get_pipeline
  File "/mnt/d/Dev/SoundHacking/project_common/rg_text_to_sound/tts_websocketserver/src/tts_websocketserver/tts_pipeline.py", line 1, in <module>
    from tts_pipeline.pipelines.waterfall.pipeline import WaterfallPipeline
ModuleNotFoundError: No module named 'tts_pipeline'

[wassfila]

After all the bash and wsl dependency could be the source of deviations between systems so removing it and having a symilar install run like other modules could be a clean up that solves this issue

[Leofltt]

@wassfila matplotlib is in the requirements, please pull most recent rg_production into project_common

[wassfila]

For info, more details about the tts fix , files to be edited

• https://github.com/TheSoundOfAIOSR/project_common/blob/main/setup3.8.bat
• https://github.com/TheSoundOfAIOSR/project_common/blob/main/setup.bat
• https://github.com/TheSoundOfAIOSR/project_common/blob/main/run.bat

[wassfila]

Taken from PR comments

• [x] regression in speech to text as it was running before missing pyaudio

fixed with 26018d6

D:\Dev\SoundHacking\project_common\rg_speech_to_text>venv\Scripts\python TheSoundOfAIOSR\stt\interface\server_app.py       
Traceback (most recent call last):
  File "TheSoundOfAIOSR\stt\interface\server_app.py", line 9, in <module>
    from TheSoundOfAIOSR.stt.interface.wsserver import SimpleServerInterface
  File "D:\Dev\SoundHacking\project_common\rg_speech_to_text\.\TheSoundOfAIOSR\stt\interface\wsserver.py", line 4, in <module>
    from TheSoundOfAIOSR.stt.control.stt_sm import SpeechToTextSM
  File "D:\Dev\SoundHacking\project_common\rg_speech_to_text\.\TheSoundOfAIOSR\stt\control\stt_sm.py", line 6, in <module>    
    from TheSoundOfAIOSR.stt.control.speech_to_text import SpeechToText
  File "D:\Dev\SoundHacking\project_common\rg_speech_to_text\.\TheSoundOfAIOSR\stt\control\speech_to_text.py", line 7, in <module>
    from TheSoundOfAIOSR.audiointerface.capture import MicrophoneCaptureFailed
  File "D:\Dev\SoundHacking\project_common\rg_speech_to_text\.\TheSoundOfAIOSR\audiointerface\capture.py", line 3, in <module>
    import pyaudio
ModuleNotFoundError: No module named 'pyaudio'

[wassfila]

• [ ] windows anti-slash error

D:\Dev\SoundHacking\project_common>venv/Scripts/python -m tts_websocketserver/src/tts_websocketserver/tts_server
'venv' is not recognized as an internal or external command,
operable program or batch file.

I tried all variants below, and don't think I can fix this without modifying the subrepo.

test1: D:\Dev\SoundHacking\test\project_common>venv/Scripts/python -m tts_websocketserver/src/tts_websocketserver/tts_server 'venv' is not recognized as an internal or external command, operable program or batch file.

test2: D:\Dev\SoundHacking\test\project_common\rg_text_to_sound>venv\Scripts\python -m tts_websocketserver/src/tts_websocketserver/tts_server D:\Dev\SoundHacking\test\project_common\rg_text_to_sound\venv\Scripts\python.exe: No module named tts_websocketserver/src/tts_websocketserver/tts_server

test3: D:\Dev\SoundHacking\test\project_common\rg_text_to_sound>venv\Scripts\python -m tts_websocketserver/src/tts_websocketserver/tts_server.py D:\Dev\SoundHacking\test\project_common\rg_text_to_sound\venv\Scripts\python.exe: Error while finding module specification for 'tts_websocketserver/src/tts_websocketserver/tts_server.py' (ModuleNotFoundError: No module named 'tts_websocketserver/src/tts_websocketserver/tts_server')

test4: D:\Dev\SoundHacking\test\project_common\rg_text_to_sound>venv\Scripts\python tts_websocketserver/src/tts_websocketserver/tts_server.py Traceback (most recent call last): File "tts_websocketserver/src/tts_websocketserver/tts_server.py", line 1, in <module> from tts_websocketserver.tts_pipeline import get_pipeline ModuleNotFoundError: No module named 'tts_websocketserver.tts_pipeline'

[wassfila]

fixed with #26