To create LexiCast episodes there are 3 processes:

 

Step 1) creating the podcasts in Notebook LM  

Step 2) making appropriate edits in Canva

Step 3) Compiling Final Episode



Step 2: Edits in Canva 


In Step 1 you would have created the draft audio for the podcast episode, however Notebook LM often makes mistakes in its pronunciation.


Follow these steps for BOTH the Short Free Audio and the Paid CPD Audio:


  1. Listen through the podcast in full, listening out for errors.
    1. Eg. 'NSW' instead of 'New South Wales' or 'A C C C' instead of 'A triple C'
  2. If there are any errors note down the timestamps 
  3. Once listened through update Columns E-F of the Podcast Completion Sheet


If there is an error:

1. Import the audio to Canva

  1. Download audio from the google drive
  2. Create a new canva design in this folder and title it '[episode name] draft audio - [free/cpd]'
  3. In the design upload the mp3 audio


2. Make appropriate edits with voice cloning app

  1. Find the part that is incorrect and split the audio around this word
    • To do this make sure the cursor is at the right time, right click, and split audio.   
      • Eg. if it mispronounced 'New South Wales' then you would split right before and after this word.
  2. Scroll on the left hand side to apps and search 'AI Voice Cloning'
  3. Select choose file and Upload either 'Female Voice' or 'Male Voice' depending on which speaker spoke the word incorrectly
    • These voice snippets can be found in:

              →NAS FOLDER: \\data\lawcpd-staff\Products\LexiCast Podcast\Voicetemplates

  4. next to 'Script' type the sentence around the word you want to replace - we use the whole sentence only to ensure flow. 
  5. Generate this and insert it over the word you are replacing. Crop using the side bars around the audio, to ensure you are only including what needs to be replaced. 
    1. It may take a few tries to generate a voice that sounds good. If it works better, replace the whole sentence of the original audio.
    2. If you are having issues with the generated audios, use pitch editors to adjust the audio. To do this you will need to:
      1. create a new canva design
      2. copy and paste the voice cloned snippet and download as an mp4. Then se the pitch editor to 
      3. Use the transmitter to convert the mp4 to an mp3 as the pitch editor will need an MP3 file to make edits
      4. Use the Pitch Editor to adjust speed and pitch to make it as close to possible to the voice in the episode audio that you are trying to match
      5. download the finalised audio and copy and paste it into the original canva design


3. Saving the edited audio

  1. Once all errors are fixed, download the audio as an mp4
  2. Save this file into the appropriate file in the Google Drive.
  3. Update columns G-J the Podcast Completion Sheet with the progress you have made
    1. In Columns G and H (Edits) Add the timestamps of where the edits were made Hyperlink the name (Column B) with the URL of the completed audio location in the google drive. 


NOTE: These Steps must be completed for both the Free and CPD Episodes


The next step is Compiling the Final Episode - Follow this Solution