I jusst started using this plugin and it worked fine at first. But I updated davinchi from 18.6.5. to 18.6.6 and it stop working.
Now i get an error message saying: String Index Out of Range

I tried following the #30 but that didn't help.

Hi!
I tried to install all the various components which seems successful (I checked for each one if it is), but somehow in DaVinci, it still can't detect ffmpeg, and thus doesn't work. I'm on Mac (Sonoma 14.4.1) so did it through Brew. I've tried multiple things, like reinstalling it, installing the latest version, also downloading ffmepg from the website, but that didn't work (there is a security lock on the file). I tried to find more info on Reddit but haven't found anyone with that specific issue. Any ideas?

Basically, Im on a laptop and I don't have full size of windows that are openned, so I can't use the advanced section.
Thank you for making this script.

So the script timed captions right, at least i think so, but text isnt added, all it says is "template" which i used to create the template

This is related to beta version of DVR so maybe not all scripting features are enabled.

auto-subs script does not work with newest 19.0b1 beta version of DVR - it does not display any text+ templates in the select component.

Hi all,

Brew being difficult to install as well as ffmpeg in some cases, i found a fix for this error on youtube. Here is the link to properly install brew and ffmpeg : https://www.youtube.com/watch?v=dJ8y-VlMNAo&ab_channel=ProgrammingKnowledge

Once completely followed, auto-sub should work just fine.

In addition, a picture of what the error looks like :

Sorry for using the issues channel as it is already fixed, but i thought it could be helpful to others.

function: 0x0287be3188
[string "???"]:8: attempt to index global 'cmp' (a nil value)
Traceback (most recent call last):
File "/Users/mymac/Library/Application Support/Blackmagic Design/DaVinci Resolve/Fusion/Scripts/Utility/auto-subs.py", line 1, in
import stable_whisper
ModuleNotFoundError: No module named 'stable_whisper'
[string "???"]:8: attempt to index global 'cmp' (a nil value)
Traceback (most recent call last):
File "/Users/mymac/Library/Application Support/Blackmagic Design/DaVinci Resolve/Fusion/Scripts/Utility/auto-subs.py", line 1, in
import stable_whisper
ModuleNotFoundError: No module named 'stable_whisper'
MediaIn1 cannot get frame for time 398
MediaOut1 cannot get Parameter for Input at time 398
MediaOut1 failed at time 398
MediaIn1 cannot get frame for time 398
MediaOut1 cannot get Parameter for Input at time 398
MediaOut1 failed at time 398
Traceback (most recent call last):
File "/Users/mymac/Library/Application Support/Blackmagic Design/DaVinci Resolve/Fusion/Scripts/Utility/auto-subs.py", line 1, in
import stable_whisper
ModuleNotFoundError: No module named 'stable_whisper'
Traceback (most recent call last):
File "/Users/mymac/Library/Application Support/Blackmagic Design/DaVinci Resolve/Fusion/Scripts/Utility/auto-subs.py", line 1, in
import stable_whisper
ModuleNotFoundError: No module named 'stable_whisper'
Traceback (most recent call last):
File "/Users/mymac/Library/Application Support/Blackmagic Design/DaVinci Resolve/Fusion/Scripts/Utility/auto-subs.py", line 1, in
import stable_whisper
ModuleNotFoundError: No module named 'stable_whisper'

I followed all the steps without error messages, but when I go to Workspace > Scripts > auto-subs nothing opens. I checked the console in DaVinci Resolve to see if Python 3 was working, and it is, but there's an error message showing:

File "/Library/Application Support/Blackmagic Design/DaVinci Resolve/Fusion/Scripts/Utility/auto-subs-light.py", line 107
    <title>auto-subs/auto-subs-light.py at main · tmoroney/auto-subs · GitHub</title>
                                                ^
SyntaxError: invalid character '·' (U+00B7)

Anyone knows what the issue and how to solve it?

OS: macOS Sonoma 14.5
DaVinci Resolve 19 beta

Hi tmoroney!

Thanks for the script. If it would work I would be very happy. Unfortunately, no luck.

Running Windows 10 Pro, Davinci Resolve 18.6. I used auto install which ended with the message that I could find it under "Scripts". Now each time I open Davinci I get the error davinci resolve unable to load plugin: fusionoperators.dll Ofter clicking OK. Davinci crashes. As this happens every time, Davinci has become unusable for me 😢

I tried running Davinci as Administrator, but that did nothing.

Has this happen before (to someone else) and do you know a solution?

Greets,
Dustin Watts

Windows 10 Pro
Resolve 18.6

I get this error in the Davinci Console after trying to generate the subs. The auto-subs menu is also stuck in the "Transcribing Audio...".

I've tried both, the auto and manual install route. Should i try to unnistall the NumPy 2.0 version and install the 1.0?

Here is the Resolve Console Log:

A module that was compiled using NumPy 1.x cannot be run in
NumPy 2.0.0 as it may crash. To support both 1.x and 2.x
versions of NumPy, modules must be compiled with NumPy 2.0.
Some module may need to rebuild instead e.g. with 'pybind11>=2.12'.

If you are a user of the module, the easiest solution will be to
downgrade to 'numpy<2' or try to upgrade the affected module.
We expect that some modules will need time to support NumPy 2.

Traceback (most recent call last):  File "C:\ProgramData\Blackmagic Design\DaVinci Resolve\Fusion\Scripts\Utility\auto-subs.py", line 26, in <module>
    import stable_whisper
  File "C:\Users\Admin\AppData\Local\Programs\Python\Python312\Lib\site-packages\stable_whisper\__init__.py", line 1, in <module>
    from .whisper_word_level import *
  File "C:\Users\Admin\AppData\Local\Programs\Python\Python312\Lib\site-packages\stable_whisper\whisper_word_level\__init__.py", line 2, in <module>
    from .cli import cli
  File "C:\Users\Admin\AppData\Local\Programs\Python\Python312\Lib\site-packages\stable_whisper\whisper_word_level\cli.py", line 11, in <module>
    from ..result import WhisperResult
  File "C:\Users\Admin\AppData\Local\Programs\Python\Python312\Lib\site-packages\stable_whisper\result.py", line 11, in <module>
    from .stabilization import suppress_silence, get_vad_silence_func, VAD_SAMPLE_RATES
  File "C:\Users\Admin\AppData\Local\Programs\Python\Python312\Lib\site-packages\stable_whisper\stabilization\__init__.py", line 7, in <module>
    from .nonvad import NONVAD_SAMPLE_RATES, audio2loudness, wav2mask, visualize_mask
  File "C:\Users\Admin\AppData\Local\Programs\Python\Python312\Lib\site-packages\stable_whisper\stabilization\nonvad.py", line 8, in <module>
    from ..audio.utils import audio_to_tensor_resample
  File "C:\Users\Admin\AppData\Local\Programs\Python\Python312\Lib\site-packages\stable_whisper\audio\__init__.py", line 7, in <module>
    from .utils import (
  File "C:\Users\Admin\AppData\Local\Programs\Python\Python312\Lib\site-packages\stable_whisper\audio\utils.py", line 7, in <module>
    import torchaudio
  File "C:\Users\Admin\AppData\Local\Programs\Python\Python312\Lib\site-packages\torchaudio\__init__.py", line 13, in <module>
    from . import (  # noqa: F401
  File "C:\Users\Admin\AppData\Local\Programs\Python\Python312\Lib\site-packages\torchaudio\compliance\__init__.py", line 1, in <module>
    from . import kaldi
  File "C:\Users\Admin\AppData\Local\Programs\Python\Python312\Lib\site-packages\torchaudio\compliance\kaldi.py", line 22, in <module>
    EPSILON = torch.tensor(torch.finfo(torch.float).eps)
C:\Users\Admin\AppData\Local\Programs\Python\Python312\Lib\site-packages\torchaudio\compliance\kaldi.py:22: UserWarning: Failed to initialize NumPy: _ARRAY_API not found (Triggered internally at ..\torch\csrc\utils\tensor_numpy.cpp:84.)
  EPSILON = torch.tensor(torch.finfo(torch.float).eps)
Using model -> [ small.en ]
MarkIn: 216000
Rendering Audio for Transcription...
Progress:  0 %
Progress:  100 %
Audio Rendering Complete!
Transcribing Audio...
Numpy is not available

After the script finishes transcribing it says this:

And if i go to the destination it talks about, i can see the file with subtitles.

The console outputs this error while transcribing the audio
[WinError 2] The system cannot find the file specified

DR 18.6.5 Studio
Windows 11 Pro

The script installed itself correctly. no errors.
But when I fire it up on DaVinci, it doesn't start at all.

To install the script I chose for the auto installer in PowerShell

Just creating this issue because after hours of troubleshooting, I figured out how to get the script to work and will update this issue soon when I have the time.

im completly lost im not sure what to do, it always renders the audio but stops before transcribing it

Auto-Subs v1.8
Davinci Resolve 18.6
MacOS Sonoma 14.2.1

When I place In- and Out-Points and start the process with "Generate Subtitles" everything is fine until the script places the Text+ elements. They are placed not exactly one hour behind my In- and Out-Points. I assume this is due to the fact that Davinci starts with "already 1 hour on the clock". The timeline starts at 1:00:00:00 and this causes confusion.

Example:

srt file is
01:15:38,670 --> 01:15:38,930

text+ ist placed on
02:15:34:04

subtitles overview in the auto-subs window:
03:15:45:26

Something is very off regarding the timecodes

In addition: is there an option in the current version to just create the SRT file and skip the placing of the Text+ element like it was possible in the old version?

Thank you very much. I can only imagine how big of a pain in the butt this was to code.

I tried to uninstall / reinstall the script and python and i still get the same error message. Please help

I installed the script, but it seems that it doesn't work for me.
Auto subs light works.

Traceback (most recent call last):
File "", line 1, in
NameError: name 'i' is not defined. Did you mean: 'id'?
Traceback (most recent call last):
File "/Library/Application Support/Blackmagic Design/DaVinci Resolve/Fusion/Scripts/Utility/auto-subs.py", line 1, in
import stable_whisper
ModuleNotFoundError: No module named 'stable_whisper'

I asked in the stable-whisper github but apparently they have no idea

Invoke-WebRequest : The underlying connection was closed: An unexpected error occurred on a send.
At line:1 char:20

• ... Expression (Invoke-WebRequest -Uri "https://raw.githubusercontent.com ...

•             ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  

  • CategoryInfo : InvalidOperation: (System.Net.HttpWebRequest:HttpWebRequest) [Invoke-WebRequest], WebExc
    eption
  • FullyQualifiedErrorId : WebCmdletWebResponseException,Microsoft.PowerShell.Commands.InvokeWebRequestCommand

Hello, I've added the text+ template to the media pool and its blank everything else works but this.

its just blank and nothing happens if I click on it

You already answered my comment on Youtube. So I just wanted to add this here to make it easier for you to keep track of the issues that can be fixed in future updates.

At the moment "max. characters" doesn't seem to work properly. For example I set the max. characters to 11 and max. words to 3. Now I still get lines that look like this: "usually not what". These are 3 words but obviously more than 11 characters. So it doesn't really work the way I expected it to.

You said it might be caused by a setting called "regroup" from the latest update. It regroups words together using an algorithm so that the sentences flow better, but it seems to be interfering with these other parameters. I'm looking forward to the updated version. 👍

I have this problem when i try to generate the subs, it gives me an srt file but it doesn't do it automaticly and when i import it mysefl the presets on the text+ doesnt aply, srry if this is a super easy thing to fix, i don't know much about programing, ty for ur time c:

Using model -> [ small ]
MarkIn: 86400
Rendering Audio for Transcription...
Progress: 0 %
Progress: 0 %
Progress: 0 %
Progress: 100 %
Audio Rendering Complete!
Transcribing Audio...

Transcribe: 0%| | 0/76.61 [00:00<?, ?sec/s]

Detected language: spanish

Transcribe: 0%| | 0/76.61 [00:06<?, ?sec/s]
Transcribe: 35%|###4 | 26.58/76.61 [00:33<01:02, 1.24s/sec]
Transcribe: 73%|#######3 | 55.94/76.61 [00:54<00:19, 1.07sec/s]
Transcribe: 100%|##########| 76.61/76.61 [01:10<00:00, 1.15sec/s]
Transcribe: 100%|##########| 76.61/76.61 [01:10<00:00, 1.09sec/s]
Saved: C:\Users\cinde\AppData\Roaming\Blackmagic Design\DaVinci Resolve\Support\Fusion\Scripts\Utility\audio.srt
Transcription Complete!
Subtitles saved to -> [ C:\Users\cinde\AppData\Roaming\Blackmagic Design\DaVinci Resolve\Support\Fusion\Scripts\Utility\audio.srt ]
Adjusting timestamps by 0:00:00
Found 73 subtitles in SRT file
list index out of range

Using model -> [ small ]
MarkIn: 90000
Rendering Audio for Transcription...
Progress:  100 %
Audio Rendering Complete!
Transcribing Audio...

Transcribe:   0%|          | 0/23.93 [00:00<?, ?sec/s]
                                                      
Detected language: english

Transcribe:   0%|          | 0/23.93 [00:01<?, ?sec/s]
Transcribe: 100%|##########| 23.93/23.93 [00:12<00:00,  1.89sec/s]
Transcribe: 100%|##########| 23.93/23.93 [00:12<00:00,  1.89sec/s]
Saved: C:\Users\beniz\AppData\Roaming\Blackmagic Design\DaVinci Resolve\Support\Fusion\Scripts\Utility\audio.srt
Transcription Complete!
Subtitles saved to -> [ C:\Users\USER\AppData\Roaming\Blackmagic Design\DaVinci Resolve\Support\Fusion\Scripts\Utility\audio.srt ]
Adjusting timestamps by 0:00:00
Found 25 subtitles in SRT file
Text+ selected as template
Adding template subtitles...
Modifying subtitle text content...
Subtitles added to timeline!

BUT, there are NO subtitles in my timeline and there is also no Subtitle Content in the AutoSubs-Console.

I have seen that others have had a similar problem where subtitles fail to transcribe to the timeline due to the text template selection not having options as described in #31 A possible solution is also mentioned in #36 however, after testing with the changes made, the bug remained with the same error result

Using model -> [ small ]
MarkIn: 108000
Rendering Audio for Transcription...
Progress: 100 %
Audio Rendering Complete!
Transcribing Audio...

Transcribe: 0%| | 0/4.25 [00:00<?, ?sec/s]

Detected language: spanish

Transcribe: 0%| | 0/4.25 [00:02<?, ?sec/s]
Transcribe: 100%|##########| 4.25/4.25 [00:10<00:00, 2.57s/sec]
Transcribe: 100%|##########| 4.25/4.25 [00:10<00:00, 2.57s/sec]
Saved: C:\Users\MRgro_0\AppData\Roaming\Blackmagic Design\DaVinci Resolve\Support\Fusion\Scripts\Utility\audio.srt
Transcription Complete!
Subtitles saved to -> [ C:\Users\MRgro_0\AppData\Roaming\Blackmagic Design\DaVinci Resolve\Support\Fusion\Scripts\Utility\audio.srt ]
Adjusting timestamps by 0:00:00
Found 4 subtitles in SRT file
list index out of range

I followed the instructions specified for automatic setup

I keep running into issues with the script not running,

Audio Rendering Complete!
Transcribing Audio...
DecodingOptions.init() got an unexpected keyword argument 'regroup'

After removing the "regroup" and "only_voice_freq"(which also didn't let my code go through), it still gives out an error for:

'dict' object has no attribute 'split_by_punctuation'
'dict' object has no attribute 'merge_by_gap'
'dict' object has no attribute 'split_by_length'

Removing those steps, I just got stuck on this:

'dict' object has no attribute 'to_srt_vtt'

Any help on what's up is appreciated!

https://i.postimg.cc/633Jg4Gb/2023-11-19-220720.png
https://i.postimg.cc/kgm5ZwMj/2023-11-19-220724.png

Hello! Let me just start by saying I've tested this plugin extensively and on the small transcription model is far superior to the default subtitles in everything except the timings. Really good job!
The issue im having is when im trying to use it to subtitle multiple tracks (one at a time) with different subtitles templates. Its very finicky if its gonna detect the Text+ in the folder or not. My request is if it would be possible to add a way for us to select which text+ we want to use with a drop down menu, similar to how the Snapcaptions plugin does it instead of dragging it to the media pool.

Thank you so much

Hi, I from Poland and I want to import polish model to this program? How can I do this? Also big love to the creator of this plugin <3

Using model -> [ medium.en ]
Rendering Audio for Transcription...
Progress:  0 %
Progress:  100 %
Audio Rendering Complete!
Transcribing Audio...
[WinError 193] %1 is not a valid Win32 application

Not sure where this would be caused because the lack of a traceback in the console.

Well, Step 2: Install Stable-TS ... Run this command. ...
Where should I run this command?

In PowerShell? It does not work ... (The term 'pip' is not recognized ...)
In Python? It does not work ... (invalid syntax)
In Resolve Console? It does not work ... (invalid syntax)

Its throwing errors that say the above and its not putting any subtitles anywhere, the clips I'm using have audio I have no clue why its not working. Pls help

If you have the [Format Text] set to ALL UPPERCASE and have words typed into [Censored Words] in lowercase it will keep the edges of the censored word lowercase
Not a program breaking bug, but a smallfix to make these edges capitalized would help especially in the case a user is using a long csv of lowercase words to censor

• step by step Video to setup on macos
• Simple or more clear steps on point to set up

(https://github.com/tmoroney/auto-subs/assets/102491821/bc6030a9-5a0d-4277-a7cf-edeac71161fa)

Hello,
I have being testing the script multiple times, but I always have this error in the console.
The when I click generate it goes into the exporting tab but doesn't export (puts the video in the queue but does not export it)
Using model -> [ small.en ]
MarkIn: 86400
Rendering Audio for Transcription...
Audio Rendering Complete!
Transcribing Audio...
FFmpeg failed to read "C:\Users\Leona\AppData\Roaming\Blackmagic Design\DaVinci Resolve\Support\Fusion\Scripts\Utility\audio.mov".

I want them to be longer, we can currently set the max word, but can we set the min word/duration?
like premiere pro

So, for example here, the bottom one (that has a black background), is the one I created manually, and I want the auto subs length to be something like that (I set the gap here to 5 second)
just ignore the language/text, it's indonesian/bahasa

or for example in this youtube video subtitle, it has some nice 2 line subtitle.

Sometimes, when I use the plugin and split a generated text box, one part of the split box will glitch, causing the whole screen to go dark, and not allow the final project to be rendered.

Transcribing Audio...
Failed to load audio: ffmpeg version 6.0-essentials_build-www.gyan.dev Copyright (c) 2000-2023 the FFmpeg developers
built with gcc 12.2.0 (Rev10, Built by MSYS2 project)
configuration: --enable-gpl --enable-version3 --enable-static --disable-w32threads --disable-autodetect --enable-fontconfig --enable-iconv --enable-gnutls --enable-libxml2 --enable-gmp --enable-lzma --enable-zlib --enable-libsrt --enable-libssh --enable-libzmq --enable-avisynth --enable-sdl2 --enable-libwebp --enable-libx264 --enable-libx265 --enable-libxvid --enable-libaom --enable-libopenjpeg --enable-libvpx --enable-libass --enable-libfreetype --enable-libfribidi --enable-libvidstab --enable-libvmaf --enable-libzimg --enable-amf --enable-cuda-llvm --enable-cuvid --enable-ffnvcodec --enable-nvdec --enable-nvenc --enable-d3d11va --enable-dxva2 --enable-libmfx --enable-libgme --enable-libopenmpt --enable-libopencore-amrwb --enable-libmp3lame --enable-libtheora --enable-libvo-amrwbenc --enable-libgsm --enable-libopencore-amrnb --enable-libopus --enable-libspeex --enable-libvorbis --enable-librubberband
libavutil 58. 2.100 / 58. 2.100
libavcodec 60. 3.100 / 60. 3.100
libavformat 60. 3.100 / 60. 3.100
libavdevice 60. 1.100 / 60. 1.100
libavfilter 9. 3.100 / 9. 3.100
libswscale 7. 1.100 / 7. 1.100
libswresample 4. 10.100 / 4. 10.100
libpostproc 57. 1.100 / 57. 1.100
C:\Users\Suoxe\AppData\Roaming\Blackmagic Design\DaVinci Resolve\Support\Fusion\Scripts\Utility\audio.unknown: No such file or directory

I got it to transcribe once. But then the rest failed. Even with the new pushed version, the Format always changes to wave but MP3 codec stays the same.

Hey,

When transcribing this issue appears. I get this issue with whatever audio file i use. Some help would be greatly aprreciated. Thank you

First at all, in the setup of lite version there is a mistake.
"Skip step 1 of the installation guide" is not true.
You have to install Python, which is part of step 1.

After installation i load [auto-subs-light.py] into the Resolve Studio (18.6), i load the srt file ... but nothing happends.

Sorry bro ... nothing works here.

Hi,
I love this! It works so well, and saves an enormous amount of time. I appreciate it.

It'd be even better if clips would be named as their captions like so:

You immediately see the content in the timeline, as opposed to having to click and check:

I do word-for-word subtitles, so it's hard to find specific words sometimes.

Thanks!

I did a sample and work fine. The second one I'm working with, works fine and add the Text+, but can't see the subtitles in the video.

Hi, I'm amused with your hard work, and I appreciate it a lot 😍

However, the problem is that it does not support non-English characters when I generate the timeline text with i.e. Polish subtitles. I have to fix that manually, which can be cumbersome.

I'll be grateful, if you manage to fix that issue 💕

Received:

Expected:

When using one word per line, the subs are off by anywhere from 1 to 7 frames on average, with some subs being off by a few seconds. By the end of a 1 minute clip, the subtitles will be off by several seconds in total. The only solution seems to be manually adjusting 80% or more of the subtitle positions, and because they aren't consistent (I.e all off by the same amount of frames), it is very time consuming to fix.

Is there any best practices for improving the timing? Does it have to do with project framerate, or the settings of the initial audio export by the script?

https://streamable.com/qg5c2v

I don't know why, but it doesn't want to recognize start and end points. Davinci version 18.5

idk much but heres all my trubleshooting with chatgpt https://chatgpt.com/share/e4a040e0-1f62-438e-addb-a3e7a8e26acc