search

technisculpt / blender-text-to-speech

blender-text-to-speech
GNU General Public License v3.0
12 stars 3 forks source link

Addon doesn't seem add any menu option (What Blender version is used? and how to fix?) #5

Closed iGodCollective closed 7 months ago

iGodCollective commented 7 months ago

My screen, blender 2.92.0

Not seeing the menu option after plugin is installed.

image

iGodCollective commented 7 months ago

I saw that in the video you used 3.1.0. So I'll try that and report.

technisculpt commented 7 months ago

It gets added to the sequence editor. Try opening with the default set up for video editor. It should work in version 2.x

iGodCollective commented 7 months ago

It gets added to the sequence editor. Try opening with the default set up for video editor. It should work in version 2.x

Thanks.

It's interesting that I made a whole open source Ai plugin for blender, and I saw SEQUENCE_EDITOR > UI > Text to Speech in your addon desc, and didn't bother checking for that

I did get a 2.92 warning from 2.91 just @ addon install, which prompted me to ask the original question.

iGodCollective commented 7 months ago

It gets added to the sequence editor. Try opening with the default set up for video editor. It should work in version 2.x

Okay, tried it in 2.91, found the drop down. When I enter text and hit "text to speech" it reports an "output path doesn't exist" error. Will dig down to find out the issue.

image

image

technisculpt commented 7 months ago

Can you screen shot the full error please?

On Thu, Mar 14, 2024, 10:00 AM iGodCollective @.***> wrote:

It gets added to the sequence editor. Try opening with the default set up for video editor. It should work in version 2.x

Okay, tried it in 2.91, found the drop down. When I enter text and hit "text to speech" it reports an "output path doesn't exist" error. Will dig down to find out the issue.

image.png (view on web) https://github.com/technisculpt/blender-text-to-speech/assets/161035122/98005bbe-c2c2-4608-b4b2-952b6669291b

— Reply to this email directly, view it on GitHub https://github.com/technisculpt/blender-text-to-speech/issues/5#issuecomment-1996047780, or unsubscribe https://github.com/notifications/unsubscribe-auth/AHUH75HKYXUMKUHHK6WHRDTYYDLARAVCNFSM6AAAAABEU7QQ2KVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSOJWGA2DONZYGA . You are receiving this because you commented.Message ID: @.***>

technisculpt commented 7 months ago

Oh what have you set as your render path?

iGodCollective commented 7 months ago

Can you screen shot the full error please?

Thanks. It's the full error, "Output path doesn't exist"

Oh what have you set as your render path?

Nothing was there, thanks. I set that path with your advice, but still same error. Still digging down.

image

Will try setting sound paths in ui above, and in code, printing out "bpy.context.scene.render.filepath" from text_to_sound.py

technisculpt commented 7 months ago

Do you get any further information in the console?

On Thu, Mar 14, 2024, 12:20 PM iGodCollective @.***> wrote:

Can you screen shot the full error please?

Thanks. It's the full error, "Output path doesn't exist"

Oh what have you set as your render path?

Nothing was there, thanks. I set that path with your advice, but still same error. Still digging down.

— Reply to this email directly, view it on GitHub https://github.com/technisculpt/blender-text-to-speech/issues/5#issuecomment-1996221524, or unsubscribe https://github.com/notifications/unsubscribe-auth/AHUH75AYULKHZATQMHVYDHLYYD3NZAVCNFSM6AAAAABEU7QQ2KVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSOJWGIZDCNJSGQ . You are receiving this because you commented.Message ID: @.***>

iGodCollective commented 7 months ago

Do you get any further information in the console?

Thanks for reply again. After setting the Sound path, a new error associated with line 61 of text_to_sound.py, line 29 of caption.py, and line 154 of eventually operators.py emerged:

image

iGodCollective commented 7 months ago

Do you get any further information in the console?

I fixed the error

iGodCollective commented 7 months ago

Do you get any further information in the console?

Quick fix (I'm working with python 3.7x for blender by the way):

In text_to_sound.py:

  1. From imports: After swapping out "from sys import platform" with "import sys as system"
  2. From sound_strip_from_text func: After swapping out {if sys.platform == "darwin"} with " {if system.platform == "darwin"}
  3. Also: system. call on platform was added to lines 11 and 13 as well.

The tool generated the audio successfully

image

technisculpt commented 7 months ago

ok, this is my bad. in my last commit, I removed "import sys" erroneously, I saw "from sys import platform". When i start a project I import the entire package then eventually i refactor to only import the modules used. However I missed that i was still calling "sys.platform" instead of just "platform"... i.e i started to refactor and didn't finish or something, dumb and naughty to commit breaking changes without testing & without automated tests a lesson to not do that even if it seems trivial...

so i'm sorry and thankyou for bringing it to my attention. reverting back a commit until i have time to test

iGodCollective commented 7 months ago

ok, this is my bad. in my last commit, I removed "import sys" erroneously, I saw "from sys import platform". When i start a project I import the entire package then eventually i refactor to only import the modules used. However I missed that i was still calling "sys.platform" instead of just "platform"... i.e i started to refactor and didn't finish or something, dumb and naughty to commit breaking changes without testing & without automated tests a lesson to not do that even if it seems trivial...

so i'm sorry and thankyou for bringing it to my attention. reverting back a commit until i have time to test

No worries bro, we're all human for now, we make little miniscule errors from time to time

Thanks for guiding me along the way, and thanks for this beautiful project

technisculpt commented 7 months ago

thanks for the kind words. looking forward to checking out your a.i addon :)

iGodCollective commented 7 months ago

thanks for the kind words. looking forward to checking out your a.i addon :)

Cool. Thanks!!!

  1. Right now, the blender/scene and addon seen here, integrates gpt2 conversational/llm unit to enable a text response associated with the 3d Aeon being.

  2. The being is a 3d model generated by ai, currently little expressions, she just moves her eyes to act like she's reading user input before ai generated text response is shown.

  3. With your project I will add a voice to the being, beyond text response. After that I will rig the being to be able to have more emotions.

So I will just play the generated sound that your project produces, based on the text response.

What's good about the ai is that it runs locally, with no need for any api key, since it's a pretrained model/weight unit.

technisculpt commented 7 months ago

sounds like a lot of fun and something I would be interested in contributing to if i wasn't so busy.

You should follow @tin2tin. He made a blender addon for screenplay writing and helped me out with text-to-speech. But he is prolific in researching open source a.i projects and my github feed is constantly filled with a tonne of really interesting repos that he has recently starred mostly to do with LLMs, text to voice, voice to text, text to image, text to mesh object, text to video etc etc. Really good guy to follow.

technisculpt commented 7 months ago

oh i didn't even realise he has made this huge addon called Paladium which is insane and very much up your alley. I must play with this.

g0dEngineer commented 6 months ago

oh i didn't even realize he has made this huge addon called Paladium which is insane and very much up your alley. I must play with this.

Thanks.

I ended up using pyttsx3 for the text to speech, as it was compatible with my version of blender, combined with Rhubarb LipSync Plugin for lipsyncing, though the lip-syncing could use some refinement.

Albeit it works and my synth finally moves her mouth and speaks words via Ai:

image

~Paladium seems very intriguing. I will find a way to integrate that too, or just introduce that to students.