Re: [stella] Berzerk Voice

Subject: Re: [stella] Berzerk Voice
From: "Clay Halliwell" <clay.h@xxxxxxxxxxxxxxxx>
Date: Wed, 14 Aug 2002 22:47:31 -0500
----- Original Message -----
From: "timchrissnider" <timchrissnider@xxxxxxxxxxx>
To: <stella@xxxxxxxxxxx>
Sent: Wednesday, August 14, 2002 8:12 PM
Subject: Re: [stella] Your Homebrews At CGE ?

> Here ya go for the details:

Unfortunately, very little in the way of details there. I want sample rates,
bit depths, etc, darnit! :)  Any chance of getting the author to post here
explaining how he did it?

In a curious bit of synchronicity, I had already been thinking about voice
on the 2600. It began lately while playing Castle Wolfenstein on Atari800Win
(finally got around to installing it). The speech in that game is crude at
best, composed of 1-bit samples at a fairly low sampling rate. Yet, the
resultant scratchy sounds and the context of the game provide just enough
detail for our brains to think they're hearing words like "Halt!" and

So, with the 2600 in mind, I came up with this wacky idea-- What if you took
a digital voice file, broke it up into jiffy-sized (1/60th second) chunks,
determined the average frequency and amplitude of each chunk, and then had a
2600 play back this data (at one sample per frame). With clever bitbashing
you could pack all the data for each chunk into a single byte, yielding 60
bytes per second of simulated digital audio.

I'd really like to know what this would sound like, or if it's already been
done. The hardest part of testing this technique would be generating the
data. Parsing a WAV is easy, but determing the average frequency of each
chunk would require a Fourier transform or similarly hairy algorithm.

Anyone interested?


Archives (includes files) at
Unsub & more at

Current Thread