Voices as instruments, instruments as voices
« previous post | next post »
Yesterday I pointed out the trombonish glissando in Bobby Vinton's "Blue Velvet"; today, during my morning ablutions, on the radio I heard a jazz singer do a whole song sounding like a musical instrument. I don't think there was any digital or electronic assistance, just his natually endowed voice.
One thing that always blows my mind is hearing Pink Floyd's guitar riffs that sound like a human voice, with the crowd going nuts in the background. This is due to David Gilmour's masterful use of effects like the Talk Box, which channels guitar sound through his mouth to form vowel sounds (like on "Pigs (Three Different Ones)"), alongside techniques like thumb-picking for vocal-like pitch bends, heavy use of Leslie speakers for swirling textures, and "wet/dry signal" setups with Uni-Vibes and delays, creating expressive, singing tones that mimic vocal phrasing. (AIO)
When I was in high school and played in various bands, brass instrumentalists would skillfully use different types of mutes to manipulate the sound stream originating from the buzzing of their lips and the tubing of their horn. In fact, before valves were added to the French horn, hornists had to fashion their melodies from a combination of embouchure control and literal manipulation of their hand inside the bell of the instrument. There are many astonishingly good French hornists, but the only two modern exponents of the instrument I know who could play the "natural" (i.e., without valves) horn were Hermann Baumann and Anthony Halstead).
The French horn is directly descended from large, circular hunting horns (cor de chasse) used in France. A French horn has a amazingly long length of coiled tubing, typically ranging from 12 to over 20 feet, depending on whether it's a single horn (around 12-13 ft for the F side) or a more common double horn, which combines both F and B♭ tubing for a total length that can reach 22 feet or more. I played a double horn (you have to generate a lot of wind to push your lip buzz through all that tubing), but started out on a much simpler, cheaper single horn, and had to learn to do the transpositions mentally /automatically. Not easy.
About thirty years ago, I went with my sister Heidi to visit a friend of hers in a Seattle suburb. His whole living room was full of electronic keyboards, amplifiers, speakers, headsets, and what-not. I'm sure that it all must have cost many thousands of dollars. He was not a professional, but created his music for his own pleasure and the entertainment of his guests such as Heidi and me.
It was like an electronic version of a late medieval organ, which has a vast array of "voices" (sounds) mimicking orchestral instruments, human voices, or creating unique tones, achieved by different pipe shapes (flue/reed), materials, and meticulous adjustments for brightness, darkness, or specific timbres like flutes, strings, brass, and winds (trumpets, oboes), and even unique effects like celestial harps or vox humanas. These voices are selected via stops on the keyboards, allowing a single organ to sound like an entire orchestra or choir in a majestic setting.
Last summer, I made a pilgrimage to Salt Lake City to hear the Mormon Tabernacle Choir, but also to attend a concert played on their magnificent pipe organ. Even the building was constructed to optimize the acoustics of all the reeds, pipes, and pedals.
Let's see where we are now in voice-instrument interaction:
This plugin transforms your voice into ANY instrument
You will scarcely believe this dazzling demonstration, all done by one man with a plugin.
Add in some AI:
Turn Your Voice Into Any Instrument with AI (Tutorial)
Using your body to make a full range of musical effects.
—–
Imitating Musical Instruments with the Human Voice
if instruments were voices and voices were instruments …
How to use Vocals like Instruments
The Voice As A Music Instrument!
Talk Box (blasting sound into your mouth) — watch how they explain the production of consonants and vowels
Can a voice sound like a stringed instrument? String Voice …
Watch this band with only their vocals as musical instruments.
—–
There are countless other presentations online of what all the electronic, digital paraphernalia can do FOR YOU. You control it. It is à votre service.
Upon initial encounter with all this intelligent equipage, it starts to get depressing, at least for me. I feel like the machines are starting to take over; what's left for me, the human?
In the end, though, I realize that I can tell the machine what to do.
The machine may be smart, but I can tell the machine what to do and what not to do. Above all, I can turn the machine on and off.
Selected readings
- "French Horn Church" (9/13/24)
- "Nasality" (8/18/23)
- "AI generated vocal model: Chinese popular ballad, Sandee Chan" (3/30/25)
- ‘We own control of our voice’: Taiwan singer Sandee Chan says don’t fear the machines as she reveals new song was ‘sung’ by artificial intelligence
Chan tells the Post that artificial technology will never replace the job of a music producer
The singer revealed the secret behind her new song, ‘Teach me the ways to be your lover’, one week after its release
Yuanyue Dang, SCMP (3/30/23) - "The Oldest (Known) Song Ever" (1025/25)
Julian said,
December 7, 2025 @ 10:11 pm
Look up "Javier Bonet natural horn" on YouTube. Great stuff.
Julian said,
December 7, 2025 @ 10:21 pm
" Above all, I can turn the machine on and off."
In one of my shelf of pop science/ pop maths/ pop psychology books, the author (Steven
Pinker? Douglas Hofstadter?) recalled his fascination when, as a child, he was given a wind-up toy which went through an elaborate routine with the sole purpose of turning itself off.
bks said,
December 8, 2025 @ 5:34 am
Julian:
https://en.wikipedia.org/wiki/Useless_machine
Matthew J. McIrvin said,
December 8, 2025 @ 9:42 pm
Musicians have been struggling with these questions at least since the beginning of recorded music, and musical automata like the player piano and music box. Music started dealing with the encroachment of machines on the live musician's turf much earlier than most other arts did.
Yet we still seem to want to see and hear live musicians, even if their live performances are less technically perfect than a studio recording. In popular music, that roughness can be part of what distinguishes an in-person performance, along with the simple fact that the performer is responding to a crowd.
JPL said,
December 9, 2025 @ 12:06 am
Thank you very much, Matthew. Here's an example of what you're talking about.
For a jazz musician, this is what it's all about. I hope you all know the song; The violin seems to be able to go beyond what even what the greatest singers can emotionally express.
https://www.youtube.com/watch?v=4Sx0SBWjGQQ
Here's Sarah Vaughan singing the song (also a recording of a live performance). BTW, her production on the final "you" is like a trumpet's half-valve, followed by a slur.
https://www.youtube.com/watch?v=gUxMWO7AdmM
J. Lin - Video AI Engineer said,
January 1, 2026 @ 10:05 pm
Mark raises crucial points about the accelerating challenge of detecting deepfakes. As someone deeply involved in the practical application of generative video AI, I see this as the defining technological double-edged sword of our time.
The underlying models driving realistic deepfakes—such as advanced diffusion-based inpainting and temporal coherency algorithms—are largely the same ones we use for legitimate, positive purposes.
For instance, our team utilizes these exact techniques to help creators cleanly restore video assets by removing unwanted artifacts or overlays. You can see a practical breakdown of how this generative approach works in a positive context here: https://blog.videowatermarkremove.com/remove-capcut-watermark-ai
The difference lies entirely in intent. This is why the industry needs to support initiatives like the C2PA standards (https://c2pa.org/) for content provenance to combat malicious use, without demonizing the underlying technology that also empowers legitimate creativity.