To use the Video Podcast feature, simply upload an audio file (such as .mp3 or .wav) or provide a URL from platforms like YouTube or TikTok, then select a scene and two characters for your podcast. You can also upload a PDF file or enter the topic you wish to discuss, and VisionStory will generate a podcast script with a host and guest for you. VisionStory will then automatically create a storyboard with intelligent shot selections based on your audio or script, and you can customise the shots, voices, and characters. Once you’re happy with the setup, click "Generate" to create your video podcast.
Anyone can upload a podcast audio file to generate a storyboard with AI-powered speakers and camera shots, but you will need a Pro Plan or higher subscription to generate the final podcast video.
Not at all! VisionStory’s AI handles the majority of the process for you. The system automatically segments your audio, assigns camera shots, and creates a storyboard for your video podcast. You can make adjustments such as choosing voices and shot types, but there’s no need for advanced video editing skills.
At present, all generated video podcasts are limited to 10 minutes in length, regardless of your subscription tier. Please be aware that longer or more complex videos will use more credits.
Yes! You can select characters from your previously uploaded image library or upload completely new ones. VisionStory's AI will automatically place these characters into your chosen scene, creating a realistic and engaging video podcast environment.
Yes, you can change the voice of each speaker in the original audio. Once the storyboard has been generated, you can choose different AI voices for each character to suit the tone and style you prefer.
There are three main shot types: single-person close-up, single-person mid-shot, and two-person shot. To change a shot, simply click on the relevant segment in the storyboard and select your preferred shot type from the available options. You can adjust the shot to focus on one speaker or display both speakers interacting.
Once the storyboard has been generated, you cannot change the characters themselves. However, you can swap the dialogue between the two speakers, so their appearances remain the same but their voices and lines are exchanged.
Don’t worry! As long as you haven’t started the final video generation, you can make changes at any point during the storyboard stage. Your edits are saved automatically, so there’s no risk of losing your progress.
If you do not have an existing podcast audio file, you can use tools such as Google's NotebookLM to generate dialogues from text. VisionStory also allows you to generate a podcast script from text files, a URL link containing the content you wish to discuss, or even simply from the podcast topic itself.
Yes, you can upload your own custom background scene for the video podcast. VisionStory will position your characters within the uploaded scene, enabling you to create a fully personalised setting.
You can easily switch between 16:9 (landscape) and 9:16 (portrait) aspect ratios by clicking the toggle button at the top of the storyboard page. This enables you to adjust the video format for different platforms with a single click.
There is no set limit to the number of video podcasts you can create. However, please be aware that each video will use credits according to its length and complexity, so your ability to create podcasts depends on your available credits within your subscription plan.
We do not provide a preview of the final video, but you can review and edit the storyboard before generating your video. VisionStory’s AI guarantees high-quality results, and the finished video will accurately reflect your storyboard with professional precision.
At present, if the speaker identification is incorrect, there is no option to manually correct it. This usually occurs when two people are speaking simultaneously. To minimise this issue, we recommend using audio where only one person speaks at a time. We are working to improve this feature in future updates.
