Home
Creators
Search Recent Missing Random
Posts
Search DMs Popular Hash Lookup Tags Random
Importer
Import FAQ
Register Favorites Keys Review DMs Logout
Community
Telegram ThePornDude
Home Creators Posts Import Register Favorites Logout

Early access to version 0.23.2 standalone (Patreon)

Published:
2024-01-26 12:47:59
Imported:
2026-06
Tags:

Content

We received some feedback and patched a few things related to the Story Editor and Speaker Detection function - detailed infos here.

So, here are the re-compiled apps for all platforms, just for Frequent Users.

For macOS arm64 (M1, M2 etc.): use this top-secret link.

For macOS Intel: use this top-secret link.

For Windows, CUDA enabled GPUs: use this top-secret link.

For Windows, non-CUDA: use this top-secret link.

Please report if you notice any issues.

Thanks again for your support!

Comments

Ted Hayash

Hi - very much looking forward to having speakers identified, and trying out the story editor. Just installed 0.23.2 and tried to transcribe a timeline in Resolve. The script creates the job in Resolve but it doesn't render. Rendering it manually doesn't cue StoryToolkitAI to submit the transcription - it's stuck at "Waiting for Render". Any ideas?

StoryToolkitAI
>>128024022

We also noticed something similar here in the editing room, but wasn't sure if it's the new Resolve API or something we missed in the update. Will look into it asap! I opened the issue on GitHub for tracking: https://github.com/octimot/StoryToolkitAI/issues/158 As a quick workaround, cancel the job that is "Waiting for Render" in the queue and just ingest the rendered file.

Wojciech Holysz

Where can I find a decent tutorial on how to use STAI in Resolve? also How can I setup exported SRT properties - number of characters/words per line and number of lines?

StoryToolkitAI
>>131504967

A lot of these workflows are described in detail in the FEATURES page on GitHub: https://github.com/octimot/StoryToolkitAI/blob/main/FEATURES.md There, you can find a Quick how-to transcribe (using the tool from Resolve): https://github.com/octimot/StoryToolkitAI/blob/main/FEATURES.md#quick-how-to-transcribe and even how to set up the Resolve connection: https://github.com/octimot/StoryToolkitAI/blob/main/FEATURES.md#initial-setup Or do you mean a video tutorial? What are you interested in finding out about the Resolve integration?

StoryToolkitAI
>>131504967

About the SRT export settings: currently, the way you can set (some) of the settings you mentioned are only available in the ingest window (see Transcription Settings here: https://github.com/octimot/StoryToolkitAI/blob/main/FEATURES.md#transcription-settings). With version 0.24.0 we've added the ability to customize your export formats which sets the baseline for customizing the export itself. We found that it's super difficult to think about all the settings and formats that different editors or production houses work with, so we're going to try to give as much flexibility on that end in the next updates. Now, having said that, I think think the benefits of setting a a maximum number of characters/words and exporting SRTs with these kinds of hard limits is not really useful (unless you're exporting TikTok/IG Stories single-word subtitles), since you're basically breaking the flow and the readability of the subtitles. This is why automated subtitles in other software also doesn't work too well for final picture distribution in my opinion. A human still needs to double check and reshuffle if necessary. In other words, there needs to be another layer - possibly AI - that is able to take decisions regarding where to break subtitles into new lines so that the viewer isn't confused by hard cut subtitles. Does this make sense?

Wojciech Holysz
>>131504967

Thank you! About the export settings - sure, it always requires manual work BUT, whatever an editors preference/spec is, if they could narrow down the number of words/letter per line, and the number of lines, they could set it up so that with their specific workflow, each would have AS LITTLE manual work as possible.

Andrzej Wisniewski

I tried in version 0.23.2 to identify speakers, but in 5/5 cases, it Failed. I tried to change the detection number from 0.3 to 0.1 to 0.9. None of these numbers resulted in the success of finding the speakers. How can I make it work?

StoryToolkitAI
>>158557605

This sounds unusual! Could you send your log file here on Patreon via DMs or open up an issue on GitHub? This way we can try to figure out what is happening. Thanks!!

Sky Martinez

Its been over a year. Any update on this project?

Sky Martinez
>>164615804

i thought this was the link to buy the product but its not. where can i download the thing at? Keep the $12, i hope it supports this project and it is fruitful into something useful.

Pavlo Yavorskyi

The programme crashes when i activate API I have the Davinci studio version. Pyhton 3.11.9 2025-02-24 02:18:18,049 - StAI - DEBUG: MotsResolve module initialized. (mots_resolve.py:42) 2025-02-24 02:18:18,077 - StAI - DEBUG: Trying to load DaVinciResolveScript module... (mots_resolve.py:160) 2025-02-24 02:18:18,078 - StAI - DEBUG: Found DaVinci Resolve at the default location: C:\Program Files\Blackmagic Design\DaVinci Resolve\ (mots_resolve.py:225) 2025-02-24 02:18:18,078 - StAI - DEBUG: Unable to find module DaVinciResolveScript from PYTHONPATH - trying default locations next (mots_resolve.py:237) 2025-02-24 02:20:12,307 - StAI - DEBUG: A version 0.25.1 standalone It won't even start. No such file transcription_template_example.yaml