Section outline

  • General

    Go to section General

    •      

    • Create AI Voices with Clipchamp

      Using AI-generated text-to-speech voices in Clipchamp is a game-changer for course creators. Hiring professional voice talent can be expensive, making AI voices a cost-effective alternative. While AI voices may not be perfect, their affordability often outweighs concerns about their artificial nature. In some cases, they can even sound better than human voices. Although Clipchamp’s AI voices are quite good, there are a few ways to enhance their quality.

      Steps to Generate AI Voices in Clipchamp

      1. Create a New Video

        • Open Clipchamp and start a new project.

      Generate AI Voiceovers

      • Click on Record & Create, then select Text to Speech.

      • Enter your desired text at (1).

      • Choose a voice type at (2).

      • Click Save (3) or Preview to listen before finalizing.

      Managing Voice Clips in the Timeline

      • Once saved, the voice clip appears in the timeline at (1).

      • To add a new segment, click Text to Speech again.

      • Position the playhead at (2) where you want the new clip to appear.

      • Click Play (3) to review how the dialogue flows.

      Exporting Your Voice Clips

      • Once your narration is complete, click Export to save the project.

      • Clipchamp exports audio as an MP4 file, which isn’t ideal for audio editing in Audacity.

      • In the next section, we’ll explore how to convert and enhance your AI-generated voices for better results.

    • Use Audacity to Extract the Sound File

      Audacity is a free and easy-to-install audio editing tool. To download it, visit www.audacityteam.org and look for the download section.

      Since Audacity works with MP3 and other audio formats but not MP4, we need to extract the sound from our exported text-to-speech file. To do this, we must install the FFmpeg plugin for Audacity.

      Steps to Install FFmpeg for Audacity

      1. Download the FFmpeg Plugin

        • On the Audacity website, locate the FFmpeg plugin and download it.

      1. Extract and Store the Files

        • Extract the downloaded files and place them in an easy-to-find location.

      2. Enable FFmpeg in Audacity

        • Open Audacity and go to Edit > Preferences > Libraries.

        • If Audacity does not automatically detect FFmpeg, manually navigate to the folder where you extracted it.

      Extracting Audio from an MP4 File

      Once FFmpeg is installed and working, you can simply drag and drop any MP4 video file into Audacity to extract its audio track—including the voiceover video you created in Clipchamp.

    • Use ChatGPT to Improve a Sound File

      If you're new to Audacity, watching a few YouTube tutorials can be a great way to get started. Begin with this one.

        .

      While ChatGPT cannot hear sound, it can analyze digital audio data to identify potential issues and suggest fixes. This allows it to detect problems such as background noise, clipping, distortion, or poor frequency balance—and recommend adjustments accordingly.

      Steps to Get AI-Generated Audio Fixes

      1. Copy a Short Audio Segment

        • In Audacity, select a short section of your sound file.

        • Copy its waveform (or export a small sample as text data if needed).

      2. Ask ChatGPT for Analysis

        • Paste the data into ChatGPT.

        • Request an analysis and improvement suggestions.

      Here’s an example of the response I received:



    • Enhance the Audio in Audacity

      Now that we have the voice file in Audacity, we can clean it up and improve its quality. Here’s the correct order of operations for the best results:

      Noise Reduction

      Background noise is a problem and has to be removed as much as possible as early as possible. Before Audacity can reduce noise, it needs to learn what the noise sounds like.

      1. Find a section of your recording where only noise is present (e.g., a pause with background hum).

      2. Highlight this section by clicking and dragging over it.

      3. Go to Effect (1) > Noise Reduction.(2)

      4. Click Get Noise Profile(3)—this tells Audacity what to remove.

       

      Now that Audacity knows what the noise sounds like, you can remove it from the whole track.

      1. Select the entire track (Ctrl + A).

      2. Go to Effect > Noise Reduction again.

      3. Adjust the following settings:

        • Noise Reduction (dB): (Start with 6–12 dB. Higher values remove more noise but may distort speech.)

        • Sensitivity: (3–6 is a good starting point. Too high, and it removes parts of the voice.)

        • Frequency Smoothing (Hz): (0–3 is best for speech; higher values may make the voice sound unnatural.)

      4. Click Preview to hear the effect before applying it.

      5. Click OK to apply noise reduction.

      6.  

      Step 3: Fine-Tune If Needed

      If your audio still has noise:

      • Redo Noise Reduction with slightly different settings.

      • If speech sounds too robotic, lower the Noise Reduction (dB) value.

      • If some noise remains, increase Sensitivity slightly.

       

      Normalize the Audio

      • Go to Effects > Normalize.

      • This evens out the volume level across the entire recording.

      Apply the Filter Curve EQ

      • Go to Effect > Filter Curve EQ > Presets & Settings > Import.

      • Instead of manually adjusting frequencies, you can ask ChatGPT to suggest an EQ profile for your voice.

      • Applying the right EQ settings enhances clarity and reduces muddiness.

      This is a text version of a filter EQ Setting.

      Apply Compression

      • Go to Effect > Compressor and adjust the settings to smooth out volume fluctuations.

      • Compression reduces differences between loud and soft parts, making the audio more balanced.

       Apply a Final Normalize Effect (Optional)

      • This ensures the overall volume is at an appropriate level before exporting.

      Remove Unwanted Sounds (Breaths & Mouth Clicks)

      • Identify faint breathing or unwanted mouth sounds in the waveform.

      • Select them and use Effects > Amplify, applying negative gain to reduce volume.

      • Alternatively, use Noise Reduction for a cleaner sound.

      Export the Final Audio

      Once you're happy with the improvements:

      1. Go to File > Export.

      2. Choose MP3 or WAV format.

      3. Save and use the cleaned-up audio in your videos or courses.

      Use ChatGPT to Optimize Your Sound

      If you're unfamiliar with Audacity, watching a YouTube tutorial is a great starting point.

      Even though ChatGPT cannot hear sounds, it can analyze digital audio properties. If you copy a short segment of your waveform data (text representation) and paste it into ChatGPT, it can suggest improvements based on typical human perception of audio clarity, frequencies, and noise reduction techniques.

      Final Thoughts

      With this workflow, you can turn AI-generated voices (and others) into professional-quality audio without expensive equipment. Even if you have a cheap microphone, following these steps can significantly enhance sound quality and make your content more engaging. (I have a very cheep mic. I f you think I should buy a more expensive one please leave a comment in the forum below...)

  • New section

    Go to section New section
    Activities: 0

  • New section

    Go to section New section
    Activities: 0

  • New section

    Go to section New section
    Activities: 0

  • New section

    Go to section New section
    Activities: 0