Vocsyn

Vocsyn

Messages with the tag vocsyn, topics related to vocal synthesis.

[74774649] 2023-06-21-15-33: wait, lenghts are more precise in the text: L/10, L/15, 2L/15, L/5, L/5, 2L/15, L/15, L/10.

[f32ba38d] 2023-01-31-09-21: Soundgen is an open-source toolbox for voice synthesis, manipulation, and analysis. https://cogsci.se/soundgen.html

[ff31fa10] 2023-01-31-09-21: phase aligned formant (PAF) generator. http://msp.ucsd.edu/techniques/latest/book-html/node96.html

[3bf1ae25] 2023-01-31-09-21: Vocoder Puccini. https://dittytoy.net/ditty/6f30b0885d

[80327d94] 2022-11-25-15-15: Imitation vs inspiration: a duality in the context of vocal synthesis.

[1905fdcb] 2022-11-25-15-15: Articularity Synthesis: why invest time in such an unrealistic speech synthesis technique, when there are plenty of more modern approaches available?

[80cca442] 2022-11-25-15-15: It comes down to philosophy: the voice is the seed of inspiration of music. Follow music through the perspective of th evoice.

[f93f821c] 2022-11-25-15-15: It isn't a high objective to directly imitate or reproduce the human voice on the computer. Humans are pretty good at that already, and to me it's the wrong wheel to re-invent.

[cdd52689] 2022-11-25-15-15: My singing synthesizers don't sing words, and I don't plan on adding them any time soon. It's not the point of it. The voice is simply the starting point in exploring lyrical control structures.

[37c269f4] 2022-11-25-15-15: Why artic synthesis over other techniques (formant, concatenative, deep learning): malleability and control. Sculptability is quite desirable. It is a technique that wants to be playfully poked at.

[10023e1c] 2022-11-25-15-15: The voice is a good instrument to model on a computer because it is a complex timbral instrument. There are many interesting challenges to solve regarding meaningful control.

[4f10380a] 2022-11-25-15-15: Similar algorithms I've found in the same vein of articulatory synthesis as malleable timbre sources: FM Synthesis, morphing wavetables. The waveguide model of the voice is still the most inspiring (perhaps because it feels so relatable?)

[d0a697c9] 2022-10-07-08-21: A Robotic Voice Simulator and the Interactive Training for Hearing-Impaired People https://www.hindawi.com/journals/bmri/2008/768232/

[c98bbe7c] 2022-07-02-08-43: Making the Master System a Master of Speech https://nicole.express/2022/let-me-show-you-the-sound-of-m y-master.html

[e3e74e5a] 2022-06-29-06-24: glance at lu/smith 1999 paper

[39a94323] 2022-05-04-09-46: HiFi-GAN is the underlying neural neural net used for speech synthesis in Larynx. https://github.com/jik876/hifi-gan

[931fab72] 2022-05-04-09-46: Larynx is an end-to-end text to speech system. Wonder how hard it is to set up? https://github.com/rhasspy/larynx.

[538a9df1] 2022-05-04-09-46: A blog post on the speech synthesis technology employed by Casio. Contains some useful links. http://sandsoftwaresound.net/casio-speech-synthesis-technology/.

[ca21be66] 2022-05-04-09-46: I believe this uses waveguides to produce cat sounds. Have not actually tried it, but I like the idea of it. https://experiments.withgoogle.com/feline-synth.

[d828cef8] 2022-05-04-09-46: An atari speech synthesizer for the AT2600. Could be worth studying and possibly adapting. https://github.com/rossumur/SAM2600.

[27c4edce] 2022-05-04-09-46: Software Automatic Mouth: another implementation of SAM (I believe it's the same), based on speech software published in 1982 by Don't Ask Software. https://github.com/s-macke/SAM

[598a4049] 2022-05-04-09-46: Chipseech by Plogue. A collection of vintage speech synthesizers. This product page has a bunch of useful information on it. https://plogue.com/products/chipspeech.html.

[1e8e708a] 2022-04-05-15-50: study gnuspeech TRM

[84c8c21e] 2022-02-24-11-08: this is apparently the reference used in praat's articulatory synthesizer. It seems very comprehensive and quite a bit more involved than the waveguides I've worked with.

[7f017a2b] 2022-02-23-14-58: this appears to be a CLM model of perry cook's vocal synthesizer: https://ccrma.stanford.edu/courses/220b-winter-2002/lectures/9/examples/singer.ins

[3a9a6026] 2022-02-23-14-58: vocsyn will be stuff related to vocal synthesis.

[c0eef67b] 2021-12-21-20-07: articulatory inversion: the process of taking acoustic data and converting it back into vocal tract shapes.