I use Cubase for audio and got a free upgrade to the basic version of a Steinberg product called Spectral Layers - I fiddled with it a bit and it seemed good. I convinced myself (with a job in mind,) the full version would be worth the upgrade. It did really well at removing an annoying bird in an outside recording, but I've just really impressed myself. I work with a female singer and she's always looking for new songs. I heard one on the radio - Emeli Sante, a singer that I'd totally missed - and I thought the one called Clowns would be good, so I started work on a version for tracks on stage - We'd have live bass and piano - and a string quartet on a track - so I recorded this all, to give to the singer to learn, but she struggled a bit with the timing - my version was am little slower and the arrangement a bit different. Normally, I'd sing a guide track, but this one is NOT a fella friendly track so I had the idea that maybe I could remove Emeli Sante's voice from her recording in Spectral Layers, and chop it up and pop it on my recording. Lots of software claims to remove vocals for people who want karaoke type tracks, but Spectral Layers removes vocals - but puts them in their own track, and the version I have can also find piano, bass and other stuff too.

My friend has the correct version to learn now - but I'm so pleased with how well, the vocal got extracted - I thought I'd share it here for the audio folk to pick at.

What you are hearing is just Emile Sante from her commercial recording, everything else was recorded here. There are plenty of possibilities - this videos where there is interference, or background noise, or other nuisance stuff - the best bit with my version is that you end up with the original track, but on multiple faders, so you can just reduce or enhance certain things in a mix - damn neat!

https://www.eastanglianradio.com/clownv6.mp3

It looks interesting Paul. I don’t work with singers as you do, so I won’t comment on your use except to say that to my ear it sounds good.

I do frequently use Izotope RX 7 Advanced to remove unwanted sounds. There’s now a version 8. It does a pretty good job. Are you familiar with Izotope, can you make a comparison?

Certainly Spectral Layers comes at a much lower price point.

I've never used the RX7 - but I'm not totally convinced that I have anywhere enough use for this feature - but I'm very happy how well it seems to work.

Here's a review of Spectral layers. Review: Spectralayers Pro 8 - YouTube

At about the 20 minute mark he compares the de noise with RX 7. In this test, spectral layers does do a better job. Unfortunate there isn't a free trial. The cost of a new license for the pro version is similar to what my upgrade price would be from RX 7 to RX 8

I am a SpectraLayers Pro 7 user and it is amazing, though some functions require some intensive learning, but the 'stems' and 'component' extraction process' are fairly easy and works well with typical pop/rock music. It ain't perfect by any means but it works resonantly well extracting 'drums, vocals, bass and keyboard' stems. Guitars and other instruments are extracted as an "Other" stem. Sensitivity can be tweaked on all.
The restoration tools can repair, attenuate or eliminate all kinds of annoyances, Some are automatic and some must be done manually... (very much like editing in Photoshop or other advanced digital photo editor.
iZ's RX Advanced is another popular option but it is 3x the cost of SLP.. It has more automated features but I do not know if RX Adv. can 'unbake the cake' (extract stems).
'Spleeter' (free) can allegedly extract stems too, but I am skeptical and have not tried it.

btw, I briefly used SpectraLayers 2 (which was originally published by SCS (Sony), but it did not float my boat. I have not used the low cost (non-pro) version of Steinberg's SL.. SLP-7 rocked my world.