Getting Started with TensorShots

From setting up the project to exporting the final video, this article walks you through each step of the process and shows you how to customise your ad along the way. Everything covered here is also explained in this video.

When you first open TensorShots, you will land on the Project Selection screen. This is where all of your projects are stored.

If you do not have any projects yet, click the “New Project” button to get started.

After creating a project, you will be taken to the Project Setup screen. Here, you have two options:

The onboarding form collects the information TensorShots needs to generate your ad. You will be asked to provide:

The more detailed your input, the better the AI-generated ad will be.

Choose your desired duration and format. For example, you might select a 10-second video in a 9x16 format. Upload at least one product image. If you have specific models you want to appear in the ad, you can add them here as well.

If you do not have these, you can skip this step and move on to creating a hook.

The hook is the first sentence of your ad. It sets the tone and direction for the entire video.

You can preview different hooks and listen to how they sound before selecting one. Each hook is editable, so you can adjust the wording as needed.

You can also preview and select different speakers from the panel on the right side. Once you are happy with the hook and speaker, continue to the next step.

Based on your selected hook, TensorShots will generate a full voiceover script for your ad.

Similar to hook creation, you can preview how the script sounds and edit the text.

Once the script sounds right, continue to storyboard generation.

TensorShots will now process all of the information you provided and generate a complete video ad. This may take a few minutes.

When finished, the ad will appear as a storyboard.

The storyboard is made up of shots. Each shot includes:

Shots are the core building blocks of your ad. You can move them along the timeline, delete them, or create new ones using the plus icon.

Before generating a video for a shot, edit the first frame by clicking on it.

This opens the Image Generator, where you can:

TensorPix Quick Image Gen Guide

To reference an image in the prompt, either tag it using Alt + G or click the Use in prompt button.

You can manually type instructions, such as changing the camera angle, or select from preset styles. Preset styles control framing, angles, and lighting.

Generate the image and select the version you want to use as the first frame by hovering over the image and clicking the “as First frame” button.

Once the first frame is ready, open the Generated Videos tab and generate the video.

Before generating, make sure the video duration is the same length or longer than the voiceover. You can always crop the video afterwards.

While the video is generating, you can continue editing other shots.

When the video is ready, it will automatically be applied to the shot.

If the video is longer than the voiceover, you can crop it to match the voiceover length.

You can move the crop selector and voiceover directly on the timeline. Once everything is aligned, click Render Shot to render the updated version.

This section explains how to change speakers and keep your audio perfectly synced with the video.

You can change the speaker in two ways:

At any point, you can change the speaker for a shot by clicking the speaker icon. This opens the speaker selection panel, where you can preview different voices and choose the one that best fits your ad.

Once selected, the new speaker will be applied to the voiceover, which will cause the audio to become unsynced.

To change the speaker for the entire ad, open the Settings tab. In the Default Voice section, click the current voice and select a new one.

If you edit the voiceover text or change the speaker after a shot has already been rendered, TensorShots will display an error indicating that the audio is no longer synced with the video.

To fix the sync issue:

Once syncing is complete, preview the shot and adjust the crop selectors if needed.

You can also adjust the voiceover speed if needed. Use the speed control to make the voice faster or slower.

After editing all shots, you can preview the entire ad by clicking the “Preview” button in the top right corner or by going to the Export Menu and clicking the “Export Preview” button. Preview will render a lower-quality version for faster review.

In the Export Menu, you can:

When you are satisfied with the preview, export the final video by clicking the blue “Export Final Cut” button.

TensorShots will render a full HD version of your ad, ready to be used on social media and other platforms.