Speaker A
In this video, I'll show you how to turn audio to text in Audacity for free using OpenVINO AI. Welcome to this video. My name is David, and I hope you're well wherever you are around the world. Now, we all know Audacity, or if you don't
Learn how to transcribe audio to text for free in Audacity using OpenVINO AI and Whisper model with step-by-step installation and usage guide.
Key Takeaways
- OpenVINO AI plugins enable free, offline audio-to-text transcription in Audacity using Whisper.
- Multiple Whisper models are available, balancing accuracy and storage requirements.
- Transcriptions can be exported as subtitle files for easy integration with video or text projects.
- Installation requires downloading large model files and enabling plugins in Audacity preferences.
- The tool supports multiple platforms (Windows, Mac, Linux) and hardware optimizations.
Summary
- Introduction to using Audacity with OpenVINO AI for free audio-to-text transcription.
- Overview of Audacity versions and download options including Muse Hub and GitHub.
- Explanation of OpenVINO AI plugins supporting Whisper transcription and other AI audio effects.
- Details on exporting transcriptions as subtitle files like SRT, VTT, or TXT.
- Guide on installation process of OpenVINO plugins and model selection based on storage and accuracy needs.
- Tips for best transcription results including noise suppression and language settings.
- Compatibility and hardware recommendations for optimal performance.
- Step-by-step walkthrough of enabling and using the transcription feature in Audacity.
- Demonstration of exporting label tracks and subtitle files from transcribed audio.
- Summary of use cases such as podcast notes, subtitles, interviews, and accessibility.
Chapters
- 00:00Introduction to Audacity and OpenVINO AI transcription
- 00:51Downloading Audacity and installation options
- 01:51Overview of Whisper transcription and export options
- 02:51Transcription modes, languages, and model accuracy
- 03:44Tips for best transcription results and hardware optimization
- 04:38Accessing OpenVINO plugins on GitHub and latest releases
- 05:30Downloading and installing OpenVINO plugins for Windows
- 06:15Installation process and model selection
- 07:43Completing installation and enabling plugins in Audacity
- 10:10Using OpenVINO Whisper transcription in Audacity
Full Transcript — Download SRT & Markdown
Speaker A
know it, it's more or less a free recording tool for audio. You can use it to edit podcasts and all that.
Speaker A
Now, it does support free AI transcription via Whisper using OpenVINO. So, I've downloaded and installed Audacity, and a couple of the few things that you'll get is that you get our free OpenVINO AI tools for noise removal, audio to text,
Speaker A
super resolution, music separation, and all that. Now, I'll take you to a couple of pages that I'll leave in the description below for you to follow along. Now, this is more or less the home page for Audacity,
Speaker A
and you can download Audacity, and this is the Muse Hub installation, and if you want to download without Muse Hub, the one that's from GitHub, you click on this one. Now, that is if you don't have Audacity downloaded and installed. And
Speaker A
I'm running the version 3.7.7. Yes, 3.7.7. And now, let's focus on what has brought us here. It's the plugins page, and the reason why I'm going to leave these links below is because sometimes getting to the section
Speaker A
that you want is a little bit difficult. So, for example, this is the OpenVINO plugins page, and it's more or less AI effects for Audacity. You can download for Windows, you can download for macOS. And if I click on the drop
Speaker A
down for macOS, you see the different options for OpenVINO that are available. Obviously, the Intel and the Apple Silicon are there. For Windows, we've got this one. So, what are they available for? Music separation, noise suppression, music generation,
Speaker A
Whisper transcription. This is what is of importance to us to be able to transcribe audio to text using OpenAI's Whisper model. You can export the resulting label track as a subtitle via File > Export > Other > Export Label. So, it's important that we
Speaker A
can always export as a text file or even as an SRT subtitle file, which is really, really awesome. And there's some documentation on how the free AI audio transcription supports conversion of speech to text offline using Whisper.
Speaker A
What it does. It's free for Windows, Mac, and Linux. Let's actually close this pop-up. And you can see the steps to follow to do the installation and all that. I think this one is a little bit wrong because
Speaker A
that's not kind of how it's working. But let's see what you can do is what is the AI transcription, obviously the installation and all that you need to do. And also the transcription settings, the Whisper models that are available,
Speaker A
the mode, transcription or translate, and also source languages and all that. You can also see how good the different options are based on the model you select, what's best for, the accuracy, and the typical speed via the CPU. And also what are the
Speaker A
common use cases like for podcast show notes, subtitles and captions, interview notes, lecture and meetings, accessibility, translation and localization, journalism and research.
Speaker A
So, I know that's a lot, and you can also see some tips for best results. Start with base or small to preview speed, then rerun the final model, which is good. Clean up recording fast with AI noise suppression. Whisper transcribe
Speaker A
the noisy speech far more accurately. That is good to know. You can select the region you want to transcribe. Whisper processes whatever is selected. That's good to know. Set the source language manually for short clips. Auto detect needs a few seconds of audio.
Speaker A
So, we can also use initial prompt to lock in proper spelling for names, brands, and technical jargon. On Intel Core Ultra laptops, select NPU device, it's faster and cooler than CPU for larger models. Export the label
Speaker A
track, File > Export > Other > Export Labels SRT, VTT, or TXT. So, it's good that we have all that knowledge. And as you can see, the frequently asked questions: Is AI transcription really free? Yes, it is.
Speaker A
Does it work offline? What's the accuracy? That is on English, on the medium and light models, it's very accurate.
Speaker A
And you can see all the different options that are available for us. So, we've covered what needs to be done, the free AI transcription tools. So, we have two options to download. We can either download via this option, or you can
Speaker A
also go to GitHub to the OpenVINO plugins AI Audacity. And it's a set of AI-enabled effects, generators, etc. for all that. And you can see the latest version of this, and I can click on the releases section here.
Speaker A
And please, it's compatible with the latest 3.7.3, means currently running 3.7.1 or 3.7.2 plugins installed, feel free to simply install 3.7.3 and should be good to go. So, we are using Audacity version 3.7.7.
Speaker A
So, there's a new feature, audio super resolution and all that. And you can see the different changes. You can actually download the Windows installer.
Speaker A
And the plugin release is only compatible with 3.7, release for Windows, so make sure you have this installed and all that. Supported hardware, you can see that. And then you can see the assets here. So, I'm going
Speaker A
to use this one and click on OpenVINO plugins from the drop-down for Windows. I'm on Windows.
Speaker A
And it's going to download onto our computer. And once it downloads, we're going to install it. So, let's navigate to the installation folder right here. Let's actually refresh this.
Speaker A
And here we have it, Audacity win version 3.7, right there, point one. And just run as administrator.
Speaker A
Yes, I want. Let's actually minimize everything else. Actually, I can close Audacity. And then I'll click on Next right here. Welcome to the OpenVINO AI plugins for Audacity setup wizard.
Speaker A
Click on Next. Click on Next. And I want it to go to C:\Program Files\Audacity.
Speaker A
Click on Next. And from here, they do have the recommended models. You can click on the drop-down, install all models, install no models, custom installation. Now, I'm going to go with what they've selected. But as you can
Speaker A
see, they're about 4 GB. But instead of going with the base, let me go with the small.
Speaker A
So, that means about 4.48. Please take note if you decide to go with the large V3, they're going to need much more space. For example, if I click on this, going to need 9.74 GB, and I don't have that. So, I'll
Speaker A
click on Next. And what happens next is that the download is going to begin for the different models for Audacity right there. So, it's going to do this for a while, like for the
Speaker A
different options that you selected. And so, depending on your internet connection, this may take a while because as we've seen, the files are a little bit huge. They're about 4.07 GB depending on the selections you've made.
Speaker A
Mine came to about 4.04 GB more or less. So, we're going to give this a while. And once it's done, we'll navigate and see how it actually works.
Speaker A
Once the OpenVINO AI plugins for Audacity are done downloading, you're going to get the option more or less ready to install. Setup is now ready to begin installing OpenVINO AI plugins for Audacity on your computer. Click Install
Speaker A
to continue the installation or click Back if you want to review or change any settings. So, you can see the destination, the setup type is custom more or less. Get these different models available and then just click on Install.
Speaker A
So, there we go. Everything has been installed for us to use in our projects in Audacity. So, it's extracting the files.
Speaker A
And before long, we'll be done and ready to actually see if it works as advertised.
Speaker A
There we go. Extracting files. And you can see should almost be done in like a couple of seconds here.
Speaker A
But yeah, it's good to have a free AI transcription interface more or less or workaround.
Speaker A
When you're done with your projects in Audacity, you can just transcribe them for show notes, for subtitles and all that in Audacity. So, let's just give it a couple of se
Speaker A
see if it works as advertised. So there we go. Let's just wait a little bit more.
Speaker A
So there we go. The installation is complete. Completing the OpenVIVO AI plugins for Audacity setup wizard.
Speaker A
Installation complete. To enable use of these plugins now open Audacity and navigate to preferences, modules, set mode OpenVIVO module to enabled, close and then reopen Audacity. That's okay and you can read the full documentation. I'll click on finish.
Speaker A
Let's open Audacity. Right there. And once we open Audacity, we're going to get the welcome splash screen and then go to edit. I'm on Windows preferences and then modules and we're looking for mode open Vivo needs to be enabled.
Speaker A
Right there and then click okay. We're going to close and then we're going to open Audacity again.
Speaker A
So once you open Audacity, we can just import an audio file. Let's just import an audio file or video just here drag and drop it.
Speaker A
And it's right there. In this tutorial, I'll show you how to use So that's okay. Now let's look for open Vivo's AI tool.
Speaker A
I do believe somewhere right here it mentioned that it's in go to analyze open Vivo transcription.
Speaker A
If you go to analyze, it's not there. It's under effect. You go to need to go to effect and I do believe they need to update this particular page.
Speaker A
Right there because it's under effect. Open Vivo AI effects and then you can see open Vivo AI music generation, music separation noise suppression, super resolution and then open Vivo whisper transcription.
Speaker A
I'll click on that. Now select the audio for open Vivo whisper transcription to use for example control A to select all then try again. Okay.
Speaker A
Click and then control A. We've selected all. As we looked at as we were reading some of the write-up, you notice that you can also make a selection of a section to have the transcription for that. So it's good that you know that.
Speaker A
Let's do control A and then go to effect. Open Vivo AI effects, open Vivo whisper transcription.
Speaker A
And once you do that, you're going to get this new dialog box open Vivo whisper transcription presets and settings. I can click on that. You You see save preset, factory presets. You can import, export, and also about. Uh type built-in, and then
Speaker A
you can see the different options now. We go from OpenVINO inference device, CPU right there. What are the device details?
Speaker A
That is more or less my computer uh CPU. And then Whisper model, what do I have?
Speaker A
I have the small one downloaded. What more do I want to transcribe or translate? Transcribe.
Speaker A
I can also click on this checkbox for advanced options. Initial prompt, maximum segment length, beam size, best of. I'll just leave that off.
Speaker A
And then the language, source language, this one is in English. I'll um once I'm done, I'll click on apply.
Speaker A
Once I do that, uh it's going to load transcription and then uh run Whisper transcription using OpenVINO.
Speaker A
And it's done. Now, this was a short file, it's about 13 seconds, and if I play it, In this tutorial, I'll show you how to use Descript's eye contact feature to simulate eye contact with the camera even when you're reading from a script.
Speaker A
Welcome to this video. My name is David, and I hope you're well wherever you are around the world.
Speaker A
So, the transcription is done. It's looking good uh more or less now. Let's export this, and it's super simple. If you go to file, export other, actually, let me just click on this uh this tab for the transcription, click on
Speaker A
file, export other, export labels, and we've got this particular option, transcription small, so you can title it the way you want, and then you can do the drop down, and then let's say SRT.
Speaker A
Let's just click on save. Now, let's navigate to the Audacity folder. Let's refresh this.
Speaker A
Right there, and then look for this transcription small.srt. Just double click on it, and there we have uh our subtitles available. Now, as you will notice, this text is too much to be for subtitle. It's going to be really
Speaker A
huge. So, my thought process is that you may need Let's actually try it out.
Speaker A
Uh let's just uh remove this. And then control A, effect, open Vevo whisper. Let's go for advanced.
Speaker A
Maximum segment length, let's see. I don't know what really this is, but can we try and see?
Speaker A
Uh Let's go for the maximum. Does it have that option to actually let us know?
Speaker A
Uh what are the transcription settings? Uh initial trim to steer the uh controls how long each label can be.
Speaker A
Uh shorter values help and uh so okay. Okay, let's say we go with something like Let's say we go with something like eight.
Speaker A
I don't know what that really is, but uh let's click on apply and see what we get. Once we do that, uh because as uh mentioned, yes. So, that more or less looks like that. Let's actually see if when we export file,
Speaker A
export other labels, uh and then we navigate to the Audacity folder, this particular one. Okay, comes in another text.
Speaker A
Uh okay, notepad. That is what we have. Uh but let's actually go for file, export labels, and let's go for SRT.
Speaker A
Save. Yes, replace it. Let's open this one. It's more or less like one word uh or two words more or less, uh right there.
Speaker A
Uh but, it needs some little understanding to do, especially how to export to SRT, uh based on uh what we've seen here, uh in uh the uh more or less the advanced effects. Uh So, I forgot to actually set that.
Speaker A
This particular section needs uh for one to understand what best works where. And also, I wish they had a section where the punctuation does not uh kind of come, uh let's say, uh isn't is like uh more or less logical, so that it can actually
Speaker A
break uh apart from uh the uh punctuations available, but if we're looking at free transcription in Audacity, that's what we get. And that is how you can use OpenVivo AI plugins to automatically transcribe audio or video in Audacity for free. That's it for this
Speaker A
video. I hope it's of value to you. Thanks for watching.
Topics:AudacityOpenVINOWhisper transcriptionAI audio transcriptionfree transcriptionaudio to textpodcast transcriptionsubtitle exportoffline transcriptionOpenVINO plugins
