How to Extract Voice only?

[alcazar]

Hello,
Please, I want to know how can I extract only voice without all others “sound”… sample joined
Thanks.

I’m under Audacity 3.3.3

This is a stereo recording sample

[Trebor]

@alcazar

I want to know how can I extract only voice

Not treatable in native Audacity. May be incurable.
Could try this free AI service to reduce the reverberation

Enhance Speech from Adobe | Free AI filter for cleaning up spoken audio

Don’t get your hopes up.

[alcazar]

I tried Adobe AI… the result is catastrophic; It seems like we are hearing another language different from the original!!
Please, is there one or more effects or filters in Audacity 3.3.3 that can improve the voice better than the original as much as possible? Please tell me the procedure.
Many thanks.

[Trebor]

@alcazar

… is there one or more effects or filters in Audacity 3.3.3 that can improve the voice

Sorry Nope.
There are de-reverb plugin$,
but they are of little use when the reverb is that bad.

[Meem]

Try this:
GOYO Voice Separator

GOYO Voice Separator | De-noise, De-Reverb, De-Voice - All in One Plugin

and Accusonus ERA-Bundle 6.2.0
https://testblog.music-society.de/wp-content/uploads/2022/03/ERA_Bundle_v6.2.00-VoiceChanger_v1.3.10-WIN.zip

Both are free.

[Trebor]

@Meem

Try this:
GOYO Voice Separator

Goyo plugin for Audacity

If you run the “.exe” it installs a VST3 plugin, which should work in Audacity 3 as a realtime effect. But in reality, (as with many VST3 plugins), it crashes Audacity when you try to use it. It’s a no Goyo. [No Goyo in Audacity 332]

[Meem]

I am working with many Realtime Effects normally without any problem.
For GOYO and some AI VST3 plugins, they need a decent hardware.

Anyway, try to reset your Audacity configuration and start again.

[kozikowski]

Some of the rescues are exotic, some too expensive, some both. There are AI tools that try to simulate the voice by getting exposed to enough of it. Those can be expensive because the publishers and developers know what they have.

The Hollywood solution is find an actor with a similar voice and re-record it in a studio. Matching the existing video is called “Looping.”
Also expensive.

This is a similar problem to someone trying to record a concert from the audience. “How do I get rid of the audience noise and echo?”

You usually don’t. Most “Help Me Clean Up” posts can’t be done.

Koz

[Trebor]

@Meem

I am working with many Realtime Effects normally without any problem.
For GOYO and some AI VST3 plugins, they need a decent hardware.

Are you running GOYO in Audacity?. If so, what OS are you using.

[Meem]

I tried it on Windows 10.
If you have a problem with it in Audacity as a realtime effect, try to use it as a direct effect. Otherwise, I recommend using it with Ripper or other DAW software.

The main two issues with this plugin are:
1- Using very high CPU and system resurces.
2- Require internet connection to send some user and usage data, as they mentioned in PRIVACY POLICY, which is a real privacy concern!

[christop]

@kozikowski

The Hollywood solution is find an actor with a similar voice and re-record it in a studio. Matching the existing video is called “Looping.”

Also called “automated dialogue recording” (ADR). I believe the term “looping” comes from when film was cut and spliced into a loop and then played back on a projector so actors could synchronize their new voice performance with the video (and the best take could later be used to replace the original voice recording).

There’s a lot that can go into audio production to make it sound good and natural.