Louder's review pretty much echoes all the forum criticisms:

https://www.loudersound.com/bands-artis … the-greats

Shave an hour off the setlist
Trim the covers
Focus on the lower-register stuff

I mean, for songs that are predominantly in the lower-register you've got:

It's So Easy
Mr Brownstone
Chinese Democracy
Patience
You're Crazy
Used To Love Her
Bad Obsession

just off the top of my head, based on the stuff Weiland could manage. They might require a bit of tweaking (Weiland tended to use the megaphone on It's So Easy to mask the fact that he couldn't do the outro screams...).

There are a few songs like Sorry and This I Love that are mostly lower-register with a few high bits; they could be tinkered with (or get some female backing singers or Melissa to bolster the higher-register stuff).

When Axl did those shows where he had a cold and had to drop the register, Nightrain and Estranged stood out as being really good, so whack them in too. For covers there's a whole load of stuff you could do, but Slither and Down On The Farm are obvious picks (doubly so if Axl actually sang Slither the way Weiland did...). Live and Let Die, too, if you cut out the big screams in the guitar section.

TBH most of the catalogue could be dropped into a lower register with a bit of effort (Yesterdays?). I think you've got to get WTTJ and Paradise City in the setlist, people will expect them regardless of what Axl sounds like. Probably November Rain, too.

There are a few setlist staples that you probably couldn't do, like YCBM and Civil War, but, well, trim the setlist to two hours and you'd need to lose some stuff anyway.

This is why I remain convinced that the only viable solution for him long term is to lean on AI. Build a model using their vast library of isolated studio vocals and live multitracks, apply it as a filter over his current vocals. He'd effectively be "playing" his own 80s-era vocals like a musical instrument with his current voice, and in theory it wouldn't sound artificial because the vocal tics and dynamics would be the same.

Yeah, it'd renege on the "No Trickery" thing, and people (not unjustly) have a gigantic hate-boner for AI right now, but, well... it wouldn't be doing anyone out of a job because the model would be trained on his own vocals.

No idea if it'd work in a live setting – I'm not sure it could be done in real-time (yet) – but it'd definitely work in studio.

And ultimately it's no worse than Mick Fleetwood (allegedly) hiding a session drummer behind the curtain when "playing" live.

I remain astonished that he hasn't started using AI as a fix for any vocal issues, in-studio at least. It'd be the work of an afternoon to train up an AI on all his isolated vocal takes from the 80s and 90s, and then they could slap it on his actual vocals as an effect. I mean, fucking YouTubers have already used the basic principle to whip up some fairly convincing stabs at the idea, and they're not working with isolated master tapes.

Sure, it'd be "trickery," but in a world where Roger Taylor and Mick Fleetwood are hiring session drummers to create the illusion that they can still bash away live, I imagine multiple aging singers are already considering it.

[/embed]
Huh, wonder if they're eying Watt as a possible producer for future GN'R material?