AI Tools We Actually Use in Post (And the Ones WeStill Treat Carefully) | Catch Lyt

Industry Insight            5 min read             

AI Tools We Actually Use in Post (And the Ones WeStill Treat Carefully) | Catch Lyt

​

AI is already in the edit suite. The useful question is what you let it touch.

​

The conversation has moved on from whether the tools exist. What matters is which parts of post get faster, and which parts still need a human who can tell when something feels off.

​

1. Use it on the mechanical work

Transcription, captioning, noise reduction, cleanup. Speed matters here and the creative risk is low. Drudgery out, attention left for the cut.

​

2. Keep taste on the emotional result

Pace, performance, colour, what stays in the film — that is still a person. The tool can draft. It should not sign off.

​

3. Be careful with generative B-roll

Audiences can often tell. When the artificial quality shows, it spends the trust the rest of the film may have built.

 

4. Be careful with AI voices on finals

Same problem. Fine for a scratch track. Risky as the voice the client ships.

Our principle is simple: let AI move faster on the chores. Keep human judgement on what feels authentic. The tools are getting better. Responsibility for the result still sits with the people using them.

 

 

 

 

​

Learning summary

1. AI is a tool, not a taste.

2. Use it for transcripts, captions, cleanup.

3. Don’t outsource the feeling of the film.

4. Generative B-roll still shows.

5. AI voices are safer as scratch.

6. Someone still has to say “that’s it.”

​

Useful reading

1. Cullen Kelly — colour and judgement: youtube.com/@CullenKelly

2. Sony Camera Channel: youtube.com/@imagingbysony