Wanting to make a clone of the Bell labs Voder which is a 1939 speech synthesizer

[Michael Weaser]

I have been wanting to build a clone of the voder. I know mostly how the Voder works , since it is based off Homer Dudleys' Vocoder design in which who is also the designer of the Voder. The circuitry of the voder is simply the 2nd half of a 10 filter vocoder , the decoder part , hooked up to a resistive keyboard . As you pressed down on the key to a particular filter the output would get louder and louder.

To play vocal sounds, you had to be really precise on how you pressed the keys. The keys had to be pressed at the correct resistive levels , and with exact timing to have it say words and sentences.

What I need help with , that isn't in the patent information ( which there is a pdf uploaded )

1 . figuring out how the voice changing works , which is probably done by modifying and adding signals to the the 2 sound generators or modifying the output of all the filters .

2. figuring out the 3 extra plosive keys and what filters each key controls , originally plosive sounds where created by swiping keys really fast to make plosive sounds, eventually automatic circuitry got added that does this for you.


The Voder is going to hard to learn to play , as there is no information on how words than turned into sentences are played on the voder on the internet. There is information on how individual sounds are made , and that's it. A guy named Stanley S. A. Watkins is the one that figured out how to play the voder.

Technically there is a manual that exists that Stanley Watkins wrote on how it is played , but only 2 people other than bell labs have it , one of those people is actually Stanley Watkins daughter, and the other guy , is another person who designed a clone of the Voder for bell labs, Bell labs doesn't want the documents released or put on the internet for some reason. I have talked to the daughter of Stanley Watkins about these issues.

A handful of ladies where trained to be able to play the Voder accurately , only a few passed.

 

[BradtheRad]

I suspect a major ingredient is to generate waveforms for the vowel sounds. In an electronic circuit you would need to store only one cycle each for a-e-i-o-u (short and long), then repeat that waveform for as long as a key is pressed. Either that or add the right combination of overtones.

By speaking into a microphone you can see waveforms on an oscilloscope. Even more convenient is a program called Audacity, for digital sound processing. It's free.

The foot pedal probably contains a potentiometer (variable resistor) for changing the frequency (pitch) of the oscillator.

Rather than a keyboard with resistive keys, I think the fingers can press simple switches. Conceivably you might use an ordinary computer keyboard.

It's tricky to create consonants (plosives, dentals, sibilants). With Audacity It's possible to capture one occurrence of mouthing a consonant, then store it electronically.

The voder sounds like it has a surprising amount of personality, from seeing this audio-video demonstration (2/3 of the way down the webpage):

 

[Michael Weaser]

I suspect a major ingredient is to generate waveforms for the vowel sounds. In an electronic circuit you would need to store only one cycle each for a-e-i-o-u (short and long), then repeat that waveform for as long as a key is pressed. Either that or add the right combination of overtones.

By speaking into a microphone you can see waveforms on an oscilloscope. Even more convenient is a program called Audacity, for digital sound processing. It's free.

The foot pedal probably contains a potentiometer (variable resistor) for changing the frequency (pitch) of the oscillator.

Rather than a keyboard with resistive keys, I think the fingers can press simple switches. Conceivably you might use an ordinary computer keyboard.

It's tricky to create consonants (plosives, dentals, sibilants). With Audacity It's possible to capture one occurrence of mouthing a consonant, then store it electronically.

The voder sounds like it has a surprising amount of personality, from seeing this audio-video demonstration (2/3 of the way down the webpage):

Yeah you actually can't actually play the voder with a single resistance key, because you can't combine keys together that well. Someone already did create a emulator for the voder and it doesn't work right, and I've contacted the person about it , they are saying the keys needs to be resisitive , they contacted the only person so far to have made a voder clone , and they had information from bell labs about it being a requirement. look up voder2017 on github.

But though there is a another voder emulator that weirdly does seem to work fine using just a computer keyboard, called the gmoe/voder on github. also available on http://griffinmoe.com/voder/ , So its confusing why the other emulator doesn't work correctly.

looking at the patent the keys do have a variable resistance that changes as you press down on the keys.: http://patents.google.com/patent/US2121142A/

 

[BradtheRad]

Okay, I suppose resistive keys provide crossfade from one phoneme to the next. It's more listenable than abrupt transitions.

However I doubt it's desirable to play two sounds at once. I picture an electronic circuit which fades out the present sound as soon as another keypress is detected, and quickly fades in the new sound. The overlap is maybe 1/10 of a second.

The more expensive electronic pianos have velocity-sensitive keys. They contain spring tension. I guess they're resistive. However I've never heard of an easy way to fabricate such a mechanism. It might be done by a potentiometer which has a spring attached to reset its position.

 

[Michael Weaser]

Okay, I suppose resistive keys provide crossfade from one phoneme to the next. It's more listenable than abrupt transitions.

However I doubt it's desirable to play two sounds at once. I picture an electronic circuit which fades out the present sound as soon as another keypress is detected, and quickly fades in the new sound. The overlap is maybe 1/10 of a second.

The more expensive electronic pianos have velocity-sensitive keys. They contain spring tension. I guess they're resistive. However I've never heard of an easy way to fabricate such a mechanism. It might be done by a potentiometer which has a spring attached to reset its position.

That is actually what I was going to do , someone make a key that uses a potentiometer and spring , so as the key is pressed the sound gets louder and can spring back up to its original state.

 

[BradtheRad]

To make it easier for a finger to push down, there's the idea of placing a cadmium-sulfide photocell near an led, so the key moves between them as you press it. The CdS cell is about a quarter inch diameter, so it doesn't cut in and out abruptly.

I saw the above method described in a project for a guitar effects pedal. Whereas, I bought a pedal that operated via the spinning potentiometer. Attached was a cap driven by hinged arms. The mechanism was troublesome because the cap constantly came detached from the potentiometer.