nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

Removing music from basketball arena

EFL

I would like to use Scene Rebalance to remove the music that is heard playing inside a basketball arena. We do not have the rights to the music, but I still want to hear the sqeaking shoes and grunts from the players. The current version of Scene Rebalance is tuned for music in the foreground and catagorizes the arena background music as "Effects".

Would it be possible to have different AI models that are trained on different environments? Much like how Topaz Video has different AI models for different types of video? Or does someone have another workaround?

I tried feeding the Effects stem back into Scene Rebalance with a gain boost, but this did not result in anything useable.

Find more posts tagged with

Feature Request

Accepted answers

All comments

PK The DJ

I assume as it's in an arena, it's swamped in reverb?

You could try removing the reverb first, and see what's left of the music. The old school way of removing the music would be by mixing an out of phase copy of the music in, to cancel it. Is it one track or many different tracks?

EFL

There is a live DJ mixing a new track every 2 minutes for 2 hours straight. When I recieve the audio it is a 5.1 mix with the play by play narration in the center channel. I deliver a stereo mix.

I can use voice isolation for the center channel to remove the music, but it's the Stereo LtRt channels that have the squeaks and grunts I want to keep.

PK The DJ

Do they have to be the original squeaks and grunts, or could you just add some new ones? 🤔

EFL

There is no time. It is 2 hours of basketball and it re-airs the next day.

PK The DJ

Ah, maybe that should have been mentioned in the first post. It does now look like you're asking iZotope to add new AI models to RX just for you in less than 24 hours. 😮

If it's the narration that's most important, you have that - all that's needed is a little "basketball game" audio atmosphere. Something that could be found in a sound effect library.

PoorFellow

https://community.native-instruments.com/discussion/51171/removing-music-from-basketball-arena

Would it be possible to have different AI models that are trained on different environments? Much like how Topaz Video has different AI models for different types of video? Or does someone have another workaround?

@Jeremy_NI , this sounds as something that you could have an opinion on (?)

https://community.native-instruments.com/discussion/comment/264688#Comment_264688

If it's the narration that's most important, you have that - all that's needed is a little "basketball game" audio atmosphere. Something that could be found in a sound effect library.

On that note : Stadium Instruments from Sample Logic

EFL

@PK The DJ May I suggest that if you don't have something useful to contribute to the conversation that you refrain from commenting at all. 1) I didn't ask for this to happen in 24 hours and 2) There is no generic sound effect that can replace the sounds of a basketball game when the focus of the broadcast is the actual basketball game.

Having different AI models trained on different content is not an unreasonable request and would benefit everyone. It would also give RX12 the competitive edge over other similair tools, such as Adobe Podcast AI and DaVinci Resolve's AI Music Remixer, which are both in a much lower price tier.

PK The DJ

I've been making suggestions trying to help you. There's not much I can do if you think they're not useful suggestions.

No you didn't ask for it happen in 24 hours using that exact phrase, and I never said you did - but what you did is ask for different AI models, and that didn't you have time to add squeaks and grunts "because it re-airs the next day".

I did ask whether they needed to be the originals, which you didn't answer - but IMO if they're literally just squeaks and grunts, why can't they be replaced with foley?

I did actually check online before suggesting using foley, and found multiple recordings of basketball games in progress, so yes there are "generic" sound effects.

If you were just asking for the AI models and not wanting a solution for this particular recording, it would have been nice if you'd mentioned that in one of your replies.

Quick links

Unanswered questions