• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

AI (Artificial Intelligence) and Mastering Your Music Files

HamboneF18

HamboneF18

Super Member
Just a thought….

I’m wondering if in the future, AI will define, descriminate, separate, instruments in music digital files, and allow adjusting those instruments, even voices, to cater to the listener’s tastes? Want high hats, cymbals louder? Turn it up! Want drums louder? Turn them up! Want placement of voices changed? Change it! Want the sax a touch louder? Turn it up!

AI is great at doing what the human brain does, only faster and in the digital domain.

Thoughts?
 
Register to hide this ad
HamboneF18 said:
Just a thought….

I’m wondering if in the future, AI will define, descriminate, separate, instruments in music digital files, and allow adjusting those instruments, even voices, to cater to the listener’s tastes? Want high hats, cymbals louder? Turn it up! Want drums louder? Turn them up! Want placement of voices changed? Change it! Want the sax a touch louder? Turn it up!

AI is great at doing what the human brain does, only faster and in the digital domain.

Thoughts?
Click to expand...

The technology already exists. The Revel Audio system in Lincolns, identifies different instruments, and directs them to different speakers so it is just a matter of incorporating that into editing software. So, my guess is that one day someone will.
 
HamboneF18 said:
Just a thought….

I’m wondering if in the future, AI will define, descriminate, separate, instruments in music digital files, and allow adjusting those instruments, even voices, to cater to the listener’s tastes? Want high hats, cymbals louder? Turn it up! Want drums louder? Turn them up! Want placement of voices changed? Change it! Want the sax a touch louder? Turn it up!

AI is great at doing what the human brain does, only faster and in the digital domain.

Thoughts?
Click to expand...
AI lacks taste, restraint, and proper ears, Automatic re-mastering is rarely if ever done right. Remixing and remastering done right need humans to do it. With proper tools. AI is a crapshoot.
 
KentTeffeteller said:
AI lacks taste, restraint, and proper ears, Automatic re-mastering is rarely if ever done right. Remixing and remastering done right need humans to do it. With proper tools. AI is a crapshoot.
Click to expand...

The OP is asking whether AI will eventually enable all of us to remix recordings to our own preferences. The result of that will probably be a lot of remixes that sound much worse than automatic remastering.
 
KentTeffeteller said:
AI lacks taste, restraint, and proper ears, Automatic re-mastering is rarely if ever done right. Remixing and remastering done right need humans to do it. With proper tools. AI is a crapshoot.
Click to expand...

Bingo! That’s what AI can do.

This is not an attempt to copyright infringe anything.

For decades there have been dynamic range expanders. Some were pretty good; most, not.

As an illustration of what I’m talking about, the human brain can listen to recorded music and recognize specific instruments. If you look at the recorded waveforms, say in Audacity or one of the many other apps, you can’t see a bass guitar in the waveforms, mixed in. But the human ear/brain can hear it and recognize it when it’s played back. This type of human brain analyses can be done through AI also. But NOT through DSP, EQ, “dynamic range expanders”, anything of the sort!

So, I’m imagining a system that can scan a music file, then the user selects the type of presentation one wants in his listening room, and selects “Outdoor Venue”, “Intimate”, “Chamber”, etc. When played back, the individual instruments jump to lifelike levels, with user-adjustable settings in addition to presets. It’s the same music, just catered to the listener’s desires. Only AI could do this.

So… If you want to listen to rock music, and have it sound like you’re sitting in front of the band in your living room, or in a concert hall, or in a more intimate jazz venue setting, you can.

Basically, unlock the recording.
 
I can't help wonder what performers in years past would say over this new technology. Recently I heard Ringo Star speaking about the work Giles Martin did on get back.... "Hey! You can hear my Drums now!".

Precisely ...

Garry
 
HamboneF18 said:
Just a thought….

I’m wondering if in the future, AI will define, descriminate, separate, instruments in music digital files, and allow adjusting those instruments, even voices, to cater to the listener’s tastes? Want high hats, cymbals louder? Turn it up! Want drums louder? Turn them up! Want placement of voices changed? Change it! Want the sax a touch louder? Turn it up!

AI is great at doing what the human brain does, only faster and in the digital domain.

Thoughts?
Click to expand...
read my last post on using ai to program multicell horn flares. AI made me a 3000+ line program to achieve this and it works. the problem is it is very computationally intensive and requires a lot of training on specific data and it has to be economically viable for a company to go through such a process to create an AI that can do this. i'm sure there will be artifacts and unwanted negative side effects (sonically speaking) of using an AI to do this however.
 
Will AI be able to make my small listening room into a stadium? I suspect not, but to me, the difference in acoustic spaces (and the volumes to fill them) are the biggest hurdle to believeable audio reproduction. I don't think I have golden ears, and I certainly can enjoy even very low end audio systems, but I have only once heard a recording played back that sounded real - that was at Take Five in New Haven in the early 80s and was a pair of big Vandersteens (1s?) on Curl electronics, playing acoustic jazz. Which is soft ball in terms of ease of dynamics and imaging, but you really felt like you could walk among the instruments (which, of course, you can't when at most performances). Everything else has sounded like a stereo - maybe a really good one - but a facsimile of the real thing. (including the sensation that you can locate instruments perfectly, which is the grail of much audiophilia, but you can't actually do this at a real performance...).
So what will AI do to solve this? Maybe it can, but it seems to me you simply can't make it sound like the BSO has somehow managed to fit into your listening room - even if your ears could take the proximity, your room's resonant modes will be there.
I suspect AI and headphone listening might be more likely to be convincing.
 
Yep, all you need is the appropriate software, capable of 24 track (or more) simultaneous recording, the computer to run it on and not drag, and that many inputs/mics.
 
Nat said:
the difference in acoustic spaces (and the volumes to fill them) are the biggest hurdle to believable audio reproduction.
Click to expand...
This has been my most salient conclusion, and it is far more difficult for me to arrive at a venue (listening space) that does this to please my ears than it is to chase the gear siren......it brings back the old adage that you can put some mediocre gear in an acoustically ideal space and it will sound amazing. Conversely, putting amazing gear in an acoustically challenged space will more than likely disappoint.

I wonder how AI will solve this challenge??? I know all kinds of software has been designed, some right into our preamps that lets us mimic a more ideal environment acoustically (Audyssey, Dirac e.g.) but I'm not overwhelmed by any I've heard. And then, there's the whole idea of keeping things analog......some prefer that, others don't care and it's all about what one's ears hear.

gkimeng said:
The OP is asking whether AI will eventually enable all of us to remix recordings to our own preferences. The result of that will probably be a lot of remixes that sound much worse than automatic remastering.
Click to expand...
Exactly, for every set of ears there may be a different remix......I've often thought about that, and why I so much prefer the work of certain mastering engineers (think Bob Ludwig, Bernie Grundman, George Marino to name a few of mine) and how perhaps they might hear similarly as I do, or moreover perhaps may prefer certain sounds similarly to me?

So, a timely question to be asking as it would seem we are on the cusp of many things in our world as we know it changing, and likely pretty dramatically at that! It may actually be a lot of fun to play around with various recordings and make changes to see how we like the results......I'm reminded of how this will put us squarely in the digital domain and how this may leave some who prefer keeping things analog still chasing, instead, the ideal room (listening space) and the gear to satisfy us therein. I dunno, will be interesting to see......(hear, that is ;)).
 
Fishead said:
you can put some mediocre gear in an acoustically ideal space and it will sound amazing. Conversely, putting amazing gear in an acoustically challenged space will more than likely disappoint.
Click to expand...

My man @tomlinmgt would agree, as do I. This important, physics-based fact needs to be canonical, but it sadly gets relegated to "oh, one day I'll fix all those acoustic issues I've spent thousands to hide, cover, and overpower."
 
Hey Tim, and ain't that the truth!!!

I have been thinking about adding a room on to the house that is a dedicated listening space and would probably already have done so but the details are mind-boggling! Living where I do, there is just no one around here that has any clue, in fact far less of a clue than I who know very little, so it has gone undone. I don't want to spend a ton of $$$ and wind up with an extra, over-size bedroom if you know what I mean. Every detail must be attended to, from the construction of the floor, walls, treatments etc. etc. that I'm just not capable of managing......heck, who knows, soon I may be able to tell some AI program what I want in terms of sound and be more likely to wind up with something that pleases me!!! This is an exciting time to be living in and I hope we will be reaping some fabulous benefits in the near future in terms of our "hobby....." Carry on, amigo.....best regards from the west!
 
Fishead said:
This has been my most salient conclusion, and it is far more difficult for me to arrive at a venue (listening space) that does this to please my ears than it is to chase the gear siren......it brings back the old adage that you can put some mediocre gear in an acoustically ideal space and it will sound amazing. Conversely, putting amazing gear in an acoustically challenged space will more than likely disappoint.

I wonder how AI will solve this challenge??? I know all kinds of software has been designed, some right into our preamps that lets us mimic a more ideal environment acoustically (Audyssey, Dirac e.g.) but I'm not overwhelmed by any I've heard. And then, there's the whole idea of keeping things analog......some prefer that, others don't care and it's all about what one's ears hear.


Exactly, for every set of ears there may be a different remix......I've often thought about that, and why I so much prefer the work of certain mastering engineers (think Bob Ludwig, Bernie Grundman, George Marino to name a few of mine) and how perhaps they might hear similarly as I do, or moreover perhaps may prefer certain sounds similarly to me?

So, a timely question to be asking as it would seem we are on the cusp of many things in our world as we know it changing, and likely pretty dramatically at that! It may actually be a lot of fun to play around with various recordings and make changes to see how we like the results......I'm reminded of how this will put us squarely in the digital domain and how this may leave some who prefer keeping things analog still chasing, instead, the ideal room (listening space) and the gear to satisfy us therein. I dunno, will be interesting to see......(hear, that is ;)).
Click to expand...
I've always thought the thing to have is the digital file of all the tracks as they were fed to the studio mixer, open in a DAW on your computer and have at it, mix, master, eq., affects, you're in the mix position, have fun, use A. I., or not, all the tracks are there. I'd buy that as a download or disc, DVD, bluray.
 
jsisk said:
The technology already exists. The Revel Audio system in Lincolns, identifies different instruments, and directs them to different speakers so it is just a matter of incorporating that into editing software. So, my guess is that one day someone will.
Click to expand...
For sure! This is why a lot of ppl complain that today's music (really since the 2010s) is very homogenized and pasteurized. Processed through sound boards and computers to get the "right" sound that will sell.

dog - really - UGH.gif
 
spicer said:
I've always thought the thing to have is the digital file of all the tracks as they were fed to the studio mixer, open in a DAW on your computer and have at it, mix, master, eq., affects, you're in the mix position, have fun, use A. I., or not, all the tracks are there. I'd buy that as a download or disc, DVD, bluray.
Click to expand...
This already exists. There are guys like Beato that show this all the time to talk about various songs. The AI divides a song into all the recognizable tracks. You can then do whatever you want to them. I don't know the name of the software though. There's also a guy on FB that uses it to analyse songs.
 
You must log in or register to reply here.
Back
Top Bottom