Using Synthesia to generate videos

This page has 0 threads | Add post

Blogs

Finding our blogs useful? Please consider helping us to share more!

Using Synthesia to generate videos
Synthesia is the market-leader in video generation software (think PowerPoint presentations with movement and avatars and you'll be on the right lines). How good is it?

Posted by Andy Brown on 31 March 2025

In this blog

Starting the process
Choosing an avatar
Adding or editing speech
Background scenes and music
Generating your video
Judging the final results
Pricing
Conclusions

Synthesia is a tool which allows you to generate professional-looking customised corporate videos in minutes (well, that's the theory - this blog assesses to what extent this works in practice). Synthesia was founded in 2017, and is based in London.

Starting the process

I initially used an introductory wizard to explain to Synthesia that I wanted to create a short video explaining the benefits of attending our two-day Introduction to SQL course. Synthesia took about a minute to generate a template of what my video might look like (I must admit I interrupted the process before it had finished):

The initial template generated by Synthesia, with no input from me.

It's time now to customise my video!

Choosing an avatar

Your video can be presented by over 125 avatars (it's even possible to create an avatar of yourself). Here I'm going to choose Francesca:

The free plan includes 9 AI avatars; for £14 a month you get 125 or more, while for £49 a month you can create 5 avatars of your own.

Weirdly, the avatar you choose is shown alongside the initial one:

I now need to select and delete the first avatar.

I've now tidied up my avatars and deleted a few auto-generated scenes to get this:

I need to keep this short because you only get 3 minutes of video in the free trial!

Adding or editing speech

This is delightfully easy - you just click on the scene you want to edit, and type in what you want your avatar to read out:

Synthesia supports over 120 languages, so it's likely that it will speak yours. Particularly if you're reading this (English language) blog ...

Background scenes and music

It's time now to add some background, and music:

You can use these two options to customise how your video will be constructed.

I'm going for a city skyline:

I've only applied this to my final scene, so the video may be a bit jerky!

As for music, let's go for a track called Unnecessary Behavior:

Notice that you can upload your own backgrounds and music.

Generating your video

It's time now to generate my video:

This is annoying - I wish Synthesia had told me this at the time.

So it's bye-bye to Francesca and hello again to Alex:

Back to the original status quo.

I can now generate my video:

Note that I can't remove the Synthesia watermark without ... upgrading to the paid version.

I've used up nearly half of my allotted free video time:

Spotting a theme yet?

I now have to wait, although I can do other things at the same time:

It did say that it would take 10 minutes to generate.

After much less than 10 minutes I get my final video:

You'll be able to judge for yourself how successful this was below!

Judging the final results

I can now create an embed code:

With the paid version, you won't by now be surprised to know, I could generate an MP4 file.

And here's what my video looks like:

Pricing

The inescapable conclusion of my mini-review above is that you'll need at least the first level of paid version. Here are the main choices:

There's also a team-level Enterprise choice - see the full list of options.

The pricing seems reasonable to me, although it's clever of Synthesia to bill yearly (they realise that otherwise everyone would spend a month producing their videos and then cancel their subscription).

Conclusions

Here - at least IMHO - are the good things about Synthesia:

Feature	My thoughts
The avatar	My avatar's movement and lip synching feel almost 100% natural.
The speech	Synthesia has rendered my speech almost perfectly, with good expression and intonation.
The effects	I feel if I spent some time on combining background media and music with sensible avatar positioning and judicious text I could generate a video which could almost pass for one created manually.

And the bad? It's that word almost in each of the comments above. Even if I manage to lose the American accent and persuade my avatar to pronounce SQL consistently (I counted 3 different variations) my video will still scream "AI-generated".

For many users tools like Synthesia will be perfect, generating promotional or explanatory videos in any language the user prefers in minutes, but this feels like it's a long way from being an acceptable stand-in for professional corporate videos.

Some other pages relevant to the above blog include:

Artificial Intelligence training

This blog has 0 threads

Add a new post

ALL BLOGS

ARTIFICIAL INTELLIGENCE BLOGS

ARTIFICIAL INTELLIGENCE (AI) BLOGS