People decry the fact that computer generated music almost sounds like it is performed by real singers.
What if the issue is that music has become so transformed by the application of technology that real singers sound like they are computer generated?
All this machinery making modern music can still be open hearted
Not so coldly charted, it’s really just a question of your honesty
I know I switch radio stations in my car the moment I hear a modern song in which the singer’s voice has been so ‘processed’ that it lacks any real depth of feeling, emotion or humanity. I simply recoil from the plastic noise of it all. No wonder AI can sound so convincing if people’s ears are accustomed to hearing modern mass market music that is itself made by machinery.
Nearly all of the youtube channels I follow post live performances or home recordings. I have thought that modern music made in big studios does sound unnatural and is not to my liking.
It could also be that producers are trying to show off all their bells and whistles. Just because you can add 15 layers of effects doesn’t mean you should.
Some good, solid hand-made music is what floats my boat since ages. With all the post-recording efforts, especially things like auto-tune, done to loads of nowadays’ productions, the last bit of a natural, human feeling is getting erased or killed. It sounds all too clean and tidy for my ears. I want to listen to humans making music, conveying their feelings. Often the true beauty lies in the imperfection, be it the few ms behind the beat to feel laid back, the vibrato being dramatic, a slight dissonance to have the little bit of friction to keep up the tension…
This topic is not new and as you said, did not only come up with AI itself. We have a German Songwriter, Reinhard Mey, who released this song (Ein Stück Musik von Hand gemacht) back in 1986
It’s German, I know, but it fits the topic really well.
But to finally get to the point of my ramble: you might be on to something here, Richard. Nowadays mainstream music with all its overproduction shaped the listener’s hearing habits to be accustomed to songs sounding robotic, thus, the difference to generated music feels all too small to be noticed at all for some.
Disclaimer: I don’t say, post-production is bad. No it’s not, but it should not be overdone.
You are opening to a big can of WORMS here IMHO. I have been into HiFi for many years, whatever HiFi means and that is quite debatable as well. But as many may be aware there has always been a long debate in HiFi circles of what is actually needed to reproduce what is recorded as made at the time of recording. In a way this is quite farcical because unless you are in the same environment (room, theatre, recording studio etc) you cannot hope to achieve a replica rendition, unless you emulate that in some fancy way. I highlight this because I think it has a relevance to what and how music is produced today, like it or not. what you hear is not necessarily how it sounded when recorded.
The issue is and has been that over at least the past 60 years the ability to control and modify how music is produced and recorded has grown exponentially. By that I mean the tools and capability to filter, add effects, modulate, re-tune notes/chords, overlay etc is so great that virtually anything in an engineers and producers mind can be done. Early examples would be Pink Floyds - ‘Dark Side of The Moon’ and The Beatles - ’ Sgt Peppers’ albums, along with a load of others, if you compare live performances of these to what was laid down in the studio then they sound quite different generally. Nowadays a computer can do whatever you want it to change how something sounds and is recorded, whether it originates from a real source instrument or real persons vocal.
a horrible example as we move forward in time is the horrific ‘Wall of Sound’ which became a thing in the late 90’s and 00’s. The music industry boosting low bass volumes to ridiculous levels, destroying dynamic range, all so that what a younger generation listening on in ear buds thought is sounded it had bass, the result being awful!
I agree with you a great deal of current modern music, pop and in some cases rock is so contrived and modified it probably bears little relationship to what the artist originally started out with. With the advent of AI unfortunately the can of musical WORMs just gets bigger sadly.
It’s the commerciality that matters to the Music Industry and what sells, not its content or how it was produced. This has probably always been the case but today we are beginning to see extremes of this.
I hope there is a backlash against this general trend and that ‘Real’ music produced by real musicians will be seen to be highly valuable and that authenticity and the ability to perform real live reproduction of what was recorded becomes important to maintain the integrity of it.
I get what you mean it’s like our ears have been trained to expect that overly processed sound, so AI just blends right in. Real, raw vocals with emotion really do stand out now, and that’s what I miss in modern tracks.
RUSH! (my fav) Agree with you here, I also think the volume of songs being cranked out has also made people passive listeners because of the endless nature of playlists. Albums don’t seem to have the same meaning anymore, as when you used to sit and actively listen to one.
The fact that AI is already able to sound like over-engeneered music now means that shortly (as in, next year) it will probably be able to sound like real people making music live somewhere, to the point where it will be indistinguishable to most ears.
Eventually, most digital content will be AI generated and tailored to the individual. The question will be whether the individual will want to consume it. Probably most people will. Human generated content and art will become niche and expensive, probably less appealing to most content-hungry humans than AI.
I know it sounds like a dystopian sci-fi future but it’s closer to reality than would be thinkable just a couple of years ago.
Yeah … my ears object to humans sounding like machines. If AI makes music containing vocal noises that sound the same, my ears will object the same.
@AJSki2fly I understand your hi fi comments. Apocryphal tales would lead us to think much modern music is produced to sound optimal through smartphone speakers or earbuds rather than decent loudspeakers.
@erionstrings If AI begins to generate vocals with passion, soul and expression it will be a challenging advancement for actual singers.
@brianlarsen Your 1st album title - Dystopian Future.
ESPN has been running a commercial for their pro golf coverage that has some song that’s obnoxiously auto-tuned. I can’t get to the mute button on my remote fast enough when it comes on.
There is no right or wrong when it comes to music, there is only perspective and/or preference. It has nothing to do with convincing anyone of anything. Theories and preconceived notions go out the window if something sounds good. We might want to ask ourselves if we are open to liking something that could catch us off guard, which may be outside of our comfort zone. Do we only like things that are somehow perfect? Self-imposed boundaries and limitations can stifle a listener’s enjoyment and a player’s confidence and growth.
@AJSki2fly, I’m pretty sure you are referring to the ‘Loudness War’ here which was aimed at making a song louder than others and make it standout on the radio. The ‘Wall of Sound’ - which is also terrible - was a production style created by Phil Spector (very influential but terrible person) during the 60s. I don’t enjoy listening to his productions. But well-regarded artists like John Lennon, George Harrison and Brian Wilson love it. Crazy stuff. But then again, they also looked into innovating. So I wouldn’t be surprised if they’d use voice designers or auto-tunes if those existed back then.
I don’t know @Richard_close2u , I have mixed feelings about this. Assuming we are discussing popular music (R&B/Rock/POP/etc…), I’d say that producers and performers were pushing the envelope and using the most of the processing power they had to produce records since mass record production became a thing (the 50s???).
One thing I personally believe is that artists will use every tool they have to express themselves or make the sound they hear in their head come out of the speakers. Many times this means recording techniques, post-processing, etc…
Some artists use these tools to create enjoyable art. Others will create rubbish. That was true and will remain true forever.
Things that were considered unconventional in the past are taken for granted today.
Without people hacking their amps in the 50s, we wouldn’t have a distortion and fuzz guitar pedals.
Without Les Paul’s recording experiments, we wouldn’t have overdubs as we know them today.
Without overdubs and track bouncing, we wouldn’t have Bohemian Rhapsody’s amazing orchestral part.
Without auto-tune, we wouldn’t have Cher (Just kidding )
What I’m trying to say is that this has been going on for ages. In many cases, a band’s live performance is very different from their recordings. And even this is changing now that it’s so much easier to get the same sound and effect on the road as in the studio.
And finally, AI.
I’m by no means a defender of AI. Far from that. It’s being shoved down my throat at work and I hate that. But to me. It is just another tool. It can be useful or a waste of time. All depends on the task at hand and how it’s used. Just like any other tool.
As to whether it’s moral to use LLMs (because it’s trained on stolen work and environmental cost), that’s for each person to decide.
@nadimz
I do actually enjoy Cher’s song Believe. I think it was a positive and creative use of auto-tune.
What my ears tell me is that auto-tune is now so over saturated on so many songs the radio stations throw out that I simply hate the homogeneous sterile sound. When AI imitates that I hate the sound too, naturally. If all AI can do is mimic the lowest common denominator of over-produced conveyor belt music then it has nothing interesting to offer me.
My ears are entrenched in the mid 60s to late 70s sounds and selective bands after that.
The homogeneous sounds being put out by the so called main stream media by the 80s led me to stop listening to Radio 1 or the so called top 40. It went in a direction I did not desire to follow, so I pursued more niche “markets”. AI is taking that same route. That’s fine and dandy as long as the “artist” tags it as AI, fobbing it off as a real human artist or a band is what I find abhorrent and offensive, as it is fraudulent.
I also cannot understand folks who take their own productions and drop it into some AI product to make it better. Would it not be more productive to learn the mixing and mastering skills (ala our old friend Roman) to take it to another level, rather than just cut paste render ? Everyone is looking for a quick win as always.
No shortcuts, no quick wins just hard graft and accumulated knowledge skilfully applied but when folk are accustomed to a 30 sec TikTok blip, guess hard work aint on the table !
reminds me of my fellow countryman, Gerry Cinnamon. He did it the old school way. Recorded evrerything at home and spent many hours learning to mix and master tracks. Created his owm albums. Sold them outside his pub gigs etc. Now he’s sellimg out stadiums and still under his own record label.
I was reading today on the BBC news the following which was quite interesting BBC News - My songs get tens of thousands of likes but each Spotify stream earns me just £0.003
Every Filipino band that did USO shows in Viet Nam played In-A-Gadda-Da-Vida, including the rest of the band walking off the stage while the drummer did the long solo.
)
)
