Would you buy a voice-controlled DAW?

Audio Plugin Hosts and other audio software applications discussion
Post Reply New Topic
RELATED
PRODUCTS

Post

mgw38 wrote: Thu Aug 01, 2019 5:27 pm
yehboy1 wrote: Thu Aug 01, 2019 4:36 pm Assuming it works well and is straightforward, e.g., "Hey Ableton, bring up the bass to [X setting] and reduce the vocals to [Y setting]," would you be interested in it?
Have you ever tried talking to Alexa when the TV is on? It's like talking to a legally deaf person.
Are there illegaly deaf persons? :o
Music tech enthusiast
DAW, VST & hardware hoarder
My "music": https://soundcloud.com/antic604

Post


Post

I live in Holland and some words are pronounced similarly to “Siri” so the phones in the office are constantly going off. It’s amusing and annoying in equal measures.
"I was wondering if you'd like to try Magic Mushrooms"
"Oooh I dont know. Sounds a bit scary"
"It's not scary. You just lose a sense of who you are and all that sh!t"

Post

I was born in the South of England, but live in the US. My pronunciation of some words is very different from all American accents, so I sometimes run into problems with automated phone systems that expect you to speak.

A couple of years ago I was trying to get through to the right department at AT&T as I was having internet issues. The system asked me what service I was calling about, and it could not understand me when I said “internet”. I said it over and over, trying to enunciate it more clearly each time, and my daughter asked me if I wanted her to say it. It recognized her saying it first time...

Post

If it were able to detect every nuance in my voice .i.e I didn't have to keep within some specific range, then absolutely :tu:


Post

The DAW would be useless if you got laryngitis etc. I prefer iLok protection.
This is the same method MJ used when he was working on Anthony Marinelli's Thriller.

Post

I prefer doing things by hand, thanks.
My solo projects:
Hekkräiser (experimental) | MFG38 (electronic/soundtrack) | The Santtu Pesonen Project (metal/prog)

Post

Logic can take voice commands, but I've never tried it, so I don't know the extent of what it can do.
“The Generals sat, and the lines on the map, moved from side to side.”
― Pink Floyd

Post

Tell it to self destruct and see what happens. :P
This is the same method MJ used when he was working on Anthony Marinelli's Thriller.

Post

Aloysius wrote: Fri Aug 02, 2019 9:36 am The DAW would be useless if you got laryngitis
Or were wearing a gimp mask.

Just saying...

Post

Forgotten wrote: Sat Aug 03, 2019 12:47 pm
Aloysius wrote: Fri Aug 02, 2019 9:36 am The DAW would be useless if you got laryngitis
Or were wearing a gimp mask.

Just saying...
that's gonna ruin casual fridays at work.
:ud:

Post

So given that voice recognition works on the basis of constantly listening to all incoming audio, then trying to pattern-match particular sound sequences in the range of human speech to known voice commands, how exactly is that going to work when the program being controlled is constantly generating audio which may include sound sequences in the range of human speech?


Also; Do you know what an adverserial image is? Its an image deliberately created to confuse an AI into thinking its 'seeing' something else.
https://openai.com/blog/adversarial-example-research/

Next up : how to embed the adverserial audio sequence for 'delete my project' into your sound samples.
An idiot on Set Theory:
"In some cases there is an object called red that contains everything that is red. In much the same way a pot is a plate."

Post

antic604 wrote: Fri Aug 02, 2019 8:16 am
mgw38 wrote: Thu Aug 01, 2019 5:27 pm
yehboy1 wrote: Thu Aug 01, 2019 4:36 pm Assuming it works well and is straightforward, e.g., "Hey Ableton, bring up the bass to [X setting] and reduce the vocals to [Y setting]," would you be interested in it?
Have you ever tried talking to Alexa when the TV is on? It's like talking to a legally deaf person.
Are there illegaly deaf persons? :o
It’s Reese Witherspoon’s latest movie I believe.
"I was wondering if you'd like to try Magic Mushrooms"
"Oooh I dont know. Sounds a bit scary"
"It's not scary. You just lose a sense of who you are and all that sh!t"

Post

whyterabbyt wrote: Sat Aug 03, 2019 1:04 pm So given that voice recognition works on the basis of constantly listening to all incoming audio, then trying to pattern-match particular sound sequences in the range of human speech to known voice commands, how exactly is that going to work when the program being controlled is constantly generating audio which may include sound sequences in the range of human speech?


Also; Do you know what an adverserial image is? Its an image deliberately created to confuse an AI into thinking its 'seeing' something else.
https://openai.com/blog/adversarial-example-research/

Next up : how to embed the adverserial audio sequence for 'delete my project' into your sound samples.
now it's actually sounding cool.
i just say "begin" and it goes on forever constantly trying to make sense of itself, leading to unimagined sounds and melodies!
far out!
:ud:

Post

It will be real fun when editing spoken text, or a bad drummer...

Really sounds more like a nightmare. I once was on a session on the radio and they recorded a piano piece of mine. There was a tonmeister and a sound engineer (probably the union had that idea) The tonmeister was not allowed to touch the knobs and she told the engineer to push the equalizer and such. More than weird and useless, it was like being in an ironic movie. And that guy was a human, now imagine a machine... Though I bet the union would prevent that happening forever...

Post Reply

Return to “Hosts & Applications (Sequencers, DAWs, Audio Editors, etc.)”